Texts of the Tibetan Buddhist Tradition Rdzogs Chen Sems Sde: a Historical, Literary and Textual Study with Critical Edition and Translation of the Tibetan Texts

Home , Direct Introduction, Samantabhadra, Semde, Tibetan Buddhist canon, Vairotsana, Vajrasattva

University of Calgary PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2019-08-28 The “Five Early” (sNga lnga) Texts of the Tibetan Buddhist Tradition rDzogs chen Sems sde: A Historical, Literary and Textual Study with Critical Edition and Translation of the Tibetan Texts

Derbac, Mihai

Derbac, M. (2019). The “Five Early” (sNga lnga) Texts of the Tibetan Buddhist Tradition rDzogs chen Sems sde: A Historical, Literary and Textual Study with Critical Edition and Translation of the Tibetan Texts (Unpublished doctoral thesis). University of Calgary, Calgary, AB. http://hdl.handle.net/1880/110861 doctoral thesis

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY

The “Five Early” (sNga lnga) Texts of the Tibetan Buddhist Tradition rDzogs chen

Sems sde: A Historical, Literary and Textual Study with Critical Edition and

Translation of the Tibetan Texts

Mihai Derbac

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

GRADUATE PROGRAM IN RELIGIOUS STUDIES

CALGARY, ALBERTA

AUGUST, 2019

ABSTRACT

The objective of this dissertation is to critically examine the formative period (8th to 9th century CE) of the rDzogs chen Sems sde tradition. Primarily associated with the rNying ma school, the rDzogs chen teachings are considered by many Tibetans, such as the present-day Dalai Lama Tenzin Gyatso and Namkhai

Norbu, to be among the most advanced forms of Buddhist teachings. rDzogs chen posits pristine awareness, the unconditioned and “ever-present” state, as the direct path to the realization of perfection.

This study is based on a historico-critical analysis and text-critical examination of five classical rDzogs chen Sems sde texts—Rig pa’i khu byug (The

Cuckoo’s Song of Awareness); rTsal chen sprug pa (The Great Potentiality);

Khyung chen lding ba (The Great Garuda in Flight); rDo la gser zhun (The

Refining of Gold from Ore); and Mi nub pa’i rgyal mtshan (The Un-waning Victory

Banner)—and a relatively obscure historical source, the tenth century bSam gtan mig sgron (Torch of the Eye of Meditation). I utilize a two-fold approach for this project that incorporates diachronic (historical) as well as synchronic (textual) examinations. The critical analysis of the historical development will examine the formative phases of the rDzogs chen Sems sde tradition, including evidence of lineage history and textual transmission. I also establish critical editions and translate the five early Sems sde texts.

iii

ACKNOWLEDGEMENTS

First and foremost, I would like to express my deepest gratitude to my supervisor, Professor James Apple. He has been my academic advisor from the very beginning, and his guidance and encouragement throughout the whole writing process of the dissertation have been indispensable. I am greatly indebted to him as

I benefited from his vast erudition as much as I benefitted from his high standard of scholarship. Some graduate students are lucky if they end up working with a good scholar; I have been fortunate to work with several great scholars, and Professor

Apple is one of them. The experience that I gained working as his Research

Assistant (on the critical edition of the Avaivartikacakra Sūtra) and his Teaching

Assistant is invaluable. Thank you!

The late Professor Leslie Kawamura was my co-supervisor, and only people who got to know him can understand how hard it is for me to express my profound gratitude. Professor Kawamura was not only an academic, a great scholar, but also a reverend, an ordained Buddhist minister. For him, the “academic study” was as important as the “personal experience,” and to this end, he invited me to “dig” deeper, and challenged me to practice a form of scholarship that would bridge the gap between the scholars and the practitioners. It is hard to overestimate the profound influence that he has had on my work. Thank you!

I am most grateful to the members of my dissertation committee for their support, comments, and suggestions: Professor Emeritus Eliezer Segal and

Professor Christopher Framarin (members of the supervisory committee), Professor

Anne Moore (the internal/external examiner), Professor George Keyworth, of the

University of Saskatchewan (the external examiner), and Professor Joy Palacios

(the neutral chair). I had the good fortune to study Textual Criticism with Professor

Segal, and working on the stemmatic and phylogenetic analyses of the rDzog chen

Sems sde texts has been one of my most rewarding experience as a graduate student. As most great scholars think alike, I had only to benefit from each and every member of the dissertation committee. Thank you!

I am also most grateful to my professors, Professor Emerita Eva Neumaier,

Professor Emeritus Douglas Shantz, and Professor Emerita Morny Joy, as well as to many scholars that I have been influenced by (some that I met, and others that I wish I had a chance to meet), such as, Namkhai Norbu, James Valby, Samten

Karmay, Jonathan Silk, Janet Gyatso, David Germano, Carmen Meinert, Sam van

Schaik, Anne Klein, Rob Mayer, Cathy Cantwell, Dan Martin, Karen Liljenberg,

Wendy Adamek, Richard Mahoney, Virginia Tumasz, Willi Braun, Tinu Ruparell,

Elizabeth Rohlman, and Rachel Blake (to name just a few). Thank you!

Last, but not least, I am most grateful to my family and friends for their support and encouragement. They have taught me that love is priceless, “It opens your mind, and sets you free!” Thank you!

TABLE OF CONTENTS

Abstract ...... ii Acknowledgements ...... iii Table of Contents ...... v Sigla of Texts, Collections, and Editions ...... vii

Chapter 1. Introduction ...... 1 1.1 Methodological considerations ...... 5 1.2 Review of existing research ...... 9

Chapter 2. Lineage History and Historical Overview of the Early Period of the Sems sde Tradition ...... 15 2.1 The nature of the rDzogs chen Sems sde ...... 16 2.2 The mythic history of the rDzogs chen tradition ...... 20 2.2.1 dGa’ rab rdo rje ...... 22 2.2.1 Mañjuśrīmitra ...... 24 2.2.1 Śrī siṃha ...... 26 2.3 The historical period (8th century) ...... 27 2.4 The “origin” of the rDzogs chen Sems sde ...... 33

Chapter 3. Oral and Textual Transmission of the Sems sde Teachings ...... 38 3.1 Oral transmission ...... 46 3.2 Textual transmission ...... 58

Chapter 4. Methods and Principles of Textual Criticism ...... 69 4.1 The Classical Method of Textual Criticism ...... 71 4.1.1 Recensio ...... 73 4.1.2 Emendatio ...... 82 4.2 The Claremont Profile Method ...... 87 4.3 The Coherence-Based Genealogical Method ...... 88 4.4 The Phylogenetic Method ...... 90

Chapter 5. Phylogenetic Analysis of the bSam gtan mig sgron’s Sems sde Quotations ...... 97 5.1 The bSam gtan mig sgron text ...... 99 5.2 Methodology ...... 102 5.3 Data set ...... 104 5.3.1 Collections and Editions Used ...... 104 5.3.2 The Five Early Seme sde Texts ...... 108 5.4 Analysis of data ...... 114 5.5 Conclusion ...... 118 5.6 Collated bSam gtan mig sgron’s Sems sde quotations ...... 121

Chapter 6. Introduction to the Critical Editions ...... 136 6.1 “Recensional” and “transmissional” readings ...... 138 6.2 Explanation of the layout ...... 143

Chapter 7. Critical Editions ...... 147 7.1 Rig pa’i khu byug ...... 147 7.2 rTsal chen sprugs pa ...... 148 7.3 Khyung chen lding ba ...... 151 7.4 rDo la gser zhun ...... 162 7.5 Mi nub pa’i rgyal mtshan ...... 166

Chapter 8. Translations ...... 186 8.1 The Cuckoo’s Song of Awareness (Rig pa’i khu byug) ...... 201 8.2 The Great Potentiality (rTsal chen sprugs pa) ...... 202 8.3 The Great Garuda in Flight (Khyung chen lding ba) ...... 204 8.4 The Refining of Gold from Ore (rDo la gser zhun) ...... 212 8.5 The Un-waning Victory Banner (Mi nub pa’i rgyal mtshan) ...... 216

Chapter 9. Conclusion ...... 230

Bibliography ...... 236 Primary Tibetan Sources ...... 236 Editions of the Bai ro’i rgyud ’bum, rNying ma rgyud ’bum, and bKa’ ’gyur . 237 Secondary Tibetan Sources ...... 239 Non-Tibetan References ...... 240

vii

SIGLA OF TEXTS, COLLECTIONS, AND EDITIONS

Texts “as”

IT Sems sde texts as “independent” texts SM Sems sde texts as “quotations” in bSam gtan mig sgron KG Sems sde texts as “chapters” in Kun byed rgyal po

Collections

K bKa’ ’gyur NGB rNying ma rgyud ’bum Bg Bai ro rgyud ’bum

Editions

Bg Bai ro rgyud ’bum (Bg) - IT TkK gTing skyes (NGB) - IT

Dg sDe dge (NGB) - KG Tb mTshams brag (NGB) - KG Tk gTing skyes (NGB) - KG

Ba Basgo (K) - KG C Co ne (K) - KG D sDe dge (K) - KG H lHa sa (K) - KG J ’Jang sa tham (Lithang) (K) - KG K Kangxi (Peking) (K) - KG N sNar thang (K) - KG Q Qianlong (Taiwan) (K) - KG

INTRODUCTION 1

CHAPTER 1

Introduction

The rDzogs chen Sems sde tradition (The Nature of Mind Tradition of the

Great Perfection—hereafter cited as rDzogs chen Sems sde) is a Tibetan Buddhist tradition. Practical in its concerns, it is centered on the experience of awareness— awareness understood as a natural state of being in the present moment. The earliest rDzogs chen Sems sde teachings are contained within a group known as the sNga lnga (Five Early) texts: Rig pa’i khu byug (The Cuckoo’s Song of Awareness); rTsal chen sprug pa (The Great Potentiality); Khyung chen lding ba (The Great

Garuda in Flight); rDo la gser zhun (The Refining of Gold from Ore); and Mi nub pa’i rgyal mtshan (The Un-waning Victory Banner). The sNga lnga texts are not descriptive in their approach (telling us what is the case) or prescriptive (telling us what we ought to do), but rather suggestive. They hint and allude, inspire and evoke a particular experience (on suggestive language see DeAngelis 2008; Odin

1986; and Wang 2000). They provide what could be called a rDzogs chen “vision” of reality through heuristic devices, such as metaphoric expressions and apophatic discourse. This vision refers to reality as-it-is, a reality that is not filtered through human prejudices, habits, and preconceptions: a reality that reveals itself spontaneously in a rich variety of possibilities. Awareness, in this context, is not a

INTRODUCTION 2 matter of some altered states of consciousness, but of being present to the wonders that life and the world have to offer. A close analysis reveals that all rDzogs chen texts have been conceived in such a manner that they not only say something but do something as well. A text’s “disclosure” provides illumination and inspiration for a spontaneous insight, a direct experience of awareness.

The objective of the dissertation is to critically examine the formative period (8th to 9th century CE) of the rDzogs chen Sems sde tradition, and provide a strong foundation for the study of rDzogs chen based on the historico-critical

(chapters 2 and 3), text-critical (chapters 4, 5, 6 and 7), and literary (chapter 8) analyses. Its primary aim is to call into question (and challenge) the “ahistorical” origin of the tradition and the “mythological” accounts of textual transmission and translation, and argue that the rDzogs chen Sems sde is a Tibetan tradition. The ahistorical and mythological “claims” are based on the traditional legendary accounts designed to provide some form of legitimacy to an emerging tradition.

They date to a period when texts that were produced in India, and by the Indian masters, were considered as “authentic,” and thus that were produced in Tibet, and by the Tibetans, as “spurious.” I argue that, perhaps, it is time to move away from the legendary accounts and acknowledge that the rDzogs chen tradition is a Tibetan tradition that originated in Tibet. The historical and textual evidence, as well as the historico-critical and the text-critical analyses, suggest that the tradition originated in Tibet itself. It co-emerged with the Tibetan Ch’an and the Tibetan Mahāyoga,

INTRODUCTION 3 based on the oral teachings that were circulating in Tibet in the late eighth and early ninth century and a relationship of mutual interaction and interdependence, between the three traditions.

The dissertation, therefore, aims to establish (1) that the rDzogs chen Sems sde is a Tibetan tradition that originated in Tibet (based on historico-critical examination), (2) that the Sems sde texts are ancient texts (based on text-critical and phylogenetic analyses), and (3) that the Sems sde texts are Tibetan compositions (based on literary analysis). It is the first study to present a critical edition of the five “earlier” Sems sde texts, and provide an English translation based on critically edited texts. It also is the first study to apply phylogenetic analysis and map stemmatic relationships across several canonical collections of

Tibetan Buddhism, such as Bai ro’i rgyud ’bum (The Collected Tantras of

Vairocana), rNying ma’i rgyud ’bum (The Hundred Thousand Tantras of the rNying ma School), and bKa’ ’gyur (The Translated Words [of the Buddha]). As the sNga lnga texts are examined in terms of critical examination, the study deepens our understanding of the history and the teachings of the rDzogs chen

Sems sde, and in so doing makes a significant contribution to the body of knowledge about this ancient tradition.

Chapters 2 and 3 of the dissertation focus on historical analysis—a historical study of the texts and their tradition within a particular cultural context.

Sems sde has been viewed as one of the three series of rDzogs chen: Sems sde (the

INTRODUCTION 4

Mind Series), kLong sde (the Space Series) and Man ngag sde (the Esoteric

Instruction Series). However, based on its own lineage and the collection of texts that are distinct from the other two series, I argue that Sems sde can be distinguished and treated as a tradition in its own right. Chapter 2 provides a historical overview of the early period of the Sems sde tradition and discusses the lineage of its teachings as conveyed through the hagiographies of the five early patriarchs: dGa’ rab rdo rje, Mañjuśrīmitra, Śrī Siṃha, Vimalamitra, and

Vairocana. Chapter 3 examines the oral and the textual transmission of the Sems sde teachings in general and the five early texts (sNga lnga) in particular.

Chapters 4 to 8 of the dissertation focus on text-critical analysis, critical editing, and translation to English of sNga lnga texts. Both Tibetan and Western scholars have raised serious questions about these texts. Ngag dbang bLo bzang rgya mtsho (1617-1682), the Fifth Dalai Lama, has asserted that the five texts are not ancient, as the Sems sde tradition is claiming, but rather later “extractions” from another text, the Kun byed rgyal po (Karmay 1988: 207). Neumaier-Dargyay

(1998), a present-day scholar, while not questioning the original texts, has suggested that the ancient texts may have undergone significant changes during their transmission, and that extant texts may not be the same as the ancient ones.

These are some of the questions that I attempt to address in chapter 5 by providing a phylogenetic analysis of the bSam gtan mig sgron’s Sems sde quotations, and a description of stemmatic relationships between and among manuscripts, as found in

INTRODUCTION 5

Bai ro’i rgyud ’bum, rNying ma’i rgyud ’bum, and bKa’ ’gyur. In chapter 4, I provide an overview of the main methods and principles of Textual Criticism, whereas in chapter 6 I provide an introduction to the critical edition of the texts and describe recensional and transmissional readings that can be found in the manuscript witnesses. Furthermore, in chapter 7 I establish the critical editions, and in chapter 8 (based on these editions) I translate the texts into English.

1.1 Methodological considerations

My intent in this dissertation is to separate the historical, literary, and the textual, and apply methods that are well suited to the task at hand. The guiding principle is that methodologies are constantly being refined, developed, and changed in order to better illuminate and disclose new and alternative “readings.”

Due to their limitations, some methods may not be adequate for some studies and, hence, may produce unsatisfactory results. The “danger” with any method is that it can trap the one who follows it within its narrow confines and possibilities, as no method is universally applicable. By consistently implementing some of the major tenets of the various methodological agendas (such as post-modernist), a critic may try (in vain I may add) to implement a method that has not yet produced, and shows no hint that it will ever produce, anything of lasting value to a particular field of study.

INTRODUCTION 6

In any research project, many issues and difficulties will arise that a scholar has to deal with. Historical and cultural gaps, as well as scholar’s own presuppositions, may come into play at one time or another. They can affect the way in which a study is conceived and presented, in terms of its structure, content, and conclusions. At the present time, most prevalent western scholarly methodologies assume there is no text of timeless value, and no scholar who approaches a text is free from presuppositions (and can never be). Accordingly, the contexts of the reader and those of the text are more important than the text itself.

We, as readers, it is claimed, are shaped by the traditions from which we come, and which we have learned to “read,” as much as by our views of the roles that texts are supposed to play. In this context, it is always about meaning and the best ways of how to get hold of it (in contrast to a personal “direct experience”), whether the meaning is in the text or is assigned to the text by the reader or the community of readers to which he (or she) belongs. My point is that even if there were a text of timeless value, we would not necessarily see it accurately because of our conditioning to look for meaning and context. Our principles and methods allow us to see; however, these same principles and methods limit what we may actually see

(and experience) without having to rely on the view or perspective of a particular principle or method.

Chapters 2 and 3 of the dissertation rely on historical-critical methodology.

The method demands a properly historical understanding of the past and should

INTRODUCTION 7 allow me to question and challenge the traditional claim of an “ahistorical” origin and development by situating the rDzogs chen Sems sde tradition within its proper historical context. The traditional accounts assert that Vajrapāṇi, who functions here as a Buddha, revealed the authoritative Sems sde texts to semi-mythic figures and that, eventually, these texts became part of the human world. The method of

Historical Criticism, by seeking to understand people, texts, and events in history, in their historical and cultural context, serves the scholars’ need for valid and reliable evidence.

Methodologically, Historical Criticism is one of several historical-critical research methodologies. Thus, Form Criticism is used to uncover oral traditions that precede the written texts. It investigates diverse forms and structures and seeks to reconstruct the “life setting” of these traditions. Source Criticism seeks to detect and, in some cases, reconstruct the documentary sources used by various authors in the composition of their works; whereas Redaction Criticism has as its purpose to discern the process carried out by a redactor of selecting, organizing, interpreting and combining diverse oral and written traditions into a final textual form. Textual

Criticism studies the subsequent transmission of the text and has as its aim to reconstruct the original text of the document (see chapter 4).

The main methodology of the dissertation is that of Textual Criticism, as most chapters depend on the conclusions of the text-critical analyses. Textual

Criticism studies the transmission of a text, with the goal of establishing the earliest

INTRODUCTION 8

“attainable” text, that is, to produce a text as close as possible to the original (Maas

1958: 1). Whenever a manuscript is copied mistakes are made, and the result is that no two extant texts are exactly the same. Each form of corruption presents a special challenge for the textual critic, and each must be approached on its own terms. This has led to the development of several methods and principles, and these methods are discussed in chapter 4 of the dissertation and applied in chapters 5 and 7.

Textual Criticism allows me to perform stemmatic and phylogenetic analyses, and establish critical editions of the Sems sde texts.

In chapter 8, the dissertation employs the method of Literary Analysis and focuses on one particular element (literary device), metre, in order to establish that the rDzogs chen Sems sde texts are Tibetan compositions. The texts are literary works in a metrical verse form (with the exception of the rDola gser zhun written in a versified prose), and they belong to a rich Buddhist tradition of versification that for centuries was using verse as a literary medium to transmit the Buddhist teachings. What distinguishes the Sems sde texts from other Tibetan translations is that they are based on the Tibetan metrical system. The Sanskrit metrical system is a multisyllabic system that is based on the Sanskrit syntax and script; it is incompatible with the Tibetan language, which is primarily bisyllabic, and cannot be adapted to it. The five Sem sde texts are nothing like the Tibetan translations, and for a good reason: they are not translations, but rather Tibetan compositions, written in Tibetan and by the Tibetans.

INTRODUCTION 9

1.2 Review of existing research

English-language translations of the sNga lnga texts

The five Sems sde texts (sNga lnga) have been translated into English as a group three times, once as a scholarly work and twice as non-academic translations.

Neumaier translated the first part (root tantra, chs. 1-57) of the Kun byed rgyal po in 1992 as The Sovereign All-Creating Mind—The Motherly Buddha. The translation was based on the gTing skyes and the mTshams brag editions of the rNying ma’i rgyud ’bum. In 1999, Norbu and Clemente published The Supreme

Source: The Kunjed Gyalpo. This text was originally published as La Supreme

Sorgente in 1997—an Italian translation of select excerpts from the Tibetan—and was re-translated from Italian into English by Lukianowicz. The text does not mention which particular Tibetan edition the original translation was based on and the English-language passages appear to be select, although they do include the five

Sems sde texts. The translated text, however, does include some insightful oral teachings given by Norbu at the end of 1977 and the beginning of 1978 on the Kun byed rgyal po.

In 2006, Dowman published a translation of the five works as independent texts (not as chapters of the Kun byed rgyal po), under the title Eye of the Storm:

Vairotsana’s Five Original Transmissions. From a scholarly perspective, this translation is rather unsuccessful. Although it is readable and contains some informative comments, it lacks any explicit text-critical or historical

INTRODUCTION 10 contextualization. It is unclear whether the translations of texts as chapters read differently from the translations of the same texts as independent ones. One could expect the situational context, the fact that they are embedded in a larger text, to play a crucial role in the way they are translated. This may or may not be the case.

What is more significant is that during the process of incorporation, the editors added a short introduction and a conclusion to the texts, and this could definitely affect the way in which they are translated.

Of the five Sems sde texts, only the first, Rig pa’i khu byug, can be said to be relatively well known, as it has been translated into English several times. Its reputation was greatly enhanced when at the beginning of the twentieth century

Stein and Pelliot discovered a copy belonging, most probably, to the ninth century, in the caves of Dunhuang in northern China (the cave was sealed sometime at the end of the tenth and the beginning of the eleventh century). The text and its commentary were subsequently translated and discussed by Karmay (1988: 42-59) in his well-known philosophically oriented account of rDzogs chen, The Great

Perfection (rDzogs Chen): A Philosophical and Meditative Teaching in Tibetan

Buddhism. Some of the other well-known translations are by Karmay (1985: 281),

Norbu (1986: xv), Norbu (1989: 48), and Reynolds (1996: 232-33).

In 1987, Norbu and Lipman translated Byang chub sems bsgom pa’i rdo la gser zhun in Primordial Experience: An Introduction to rDzogs-chen Meditation.

This text is attributed by the rDzogs chen tradition to Mañjuśrīmitra (an early

INTRODUCTION 11 rDzogs chen patriarch from India), and Norbu and Lipman claim that this text is the rDo la gser zhun, one of the five texts under consideration. Preliminary text-critical research suggests, however, that this text is a philosophical treatise that is based on another text, a root text, and is not the text that can be found in the Kun byed rgyal po. Furthermore, as scriptures, none of the Sems sde texts are either commentaries or treatises, or are attributed to an author.

Secondary Literature on rDzogs chen Sems sde

Despite the importance attached to sNga lnga texts, until very recently there has been a surprising dearth of academic studies on them or the rDzogs chen Sems sde tradition as a whole. Most scholars acknowledge that these texts are some of the earliest rDzogs chen codices, and, as such, are seminal to the entire rDzogs chen literature. Notwithstanding, such assessments occur only sporadically and in obscure comments. Thus, Germano, in his article entitled “Architecture and

Absence in the Secret Tantric History of the Great Perfection (rdzogs-chen)”

(1995), suggests that Sems sde was originally an independent movement of

“unknown indigenous elements” defined by the “rhetorical rejection” of tantric beliefs and practices (204-205).

The same view was upheld earlier on by Karmay (1988) and more recently by Meinert (2002). In his discussion of the rDzogs chen, as presented in the Man ngag lta ba’i phreng ba (a late eighth-century text), Karmay argues that the text’s

INTRODUCTION 12 view is a syncretic teaching mainly drawn from the controversial Guhyagarbha- tantra and tinged with “thought originating in Sems sde” texts (152). Meinert

(2002: 297-304), in her analysis of two Dunhuang manuscripts, IOL Tib J 689/1 and Pelliot tibétain 699, concludes Pelliot tibétain 699 belongs to a syncretic movement between Chinese Ch’an and rDzogs chen (not-yet-systematized rDzogs chen, Sems sde).

Unlike Karmay, Germano, and Meinert, van Schaik, in his article, “The

Early Days of the Great Perfection” (2004), proposes that Sems sde was originally nothing more than a way of interpreting the Mahāyoga tantras, one that tantric scholars employed in their exegetical works when commenting on the tantric texts.

At the time when the commentaries were being produced (9th century), texts were also being written that resembled “the early Sems sde texts” (178, 200). Scholars, such as gNyan dPal dbyangs, wrote treatises on Mahāyoga, with an emphasis on non-duality and spontaneous presence, without directly elaborating on Mahāyoga rituals. Thus, according to him, it would seem that “two types of composition occurred within Māyājāla [tantric] commentarial traditions” (201). Perhaps a better explanation for the existence of the two types of composition would be to assume, as most academic scholars have done all along, that tantric exegetes wrote the commentaries on tantras, while Sems sde scholar/practitioners wrote the early texts, based on the teachings of their own tradition, the Sems sde.

INTRODUCTION 13

The most significant contemporary study of the sNga lnga texts, in terms of the questions that it raises and the problems that it poses, is by Eva Neumaier: “The bSam gtan mig sgron and its rDzogs chen Quotations: A Study in the Production of

Tibetan Texts,” a paper presented in 1998, at the proceedings of the 8th seminar of the International Association for Tibetan Studies in Bloomington, Indiana. gNubs chen Sangs rgyas ye shes’ text, the bSam gtan mig sgron (early 10th century), is the earliest known account of the rDzogs chen tradition, and it includes several quotes from the earliest Sems sde texts. Out of five sNga lnga texts, four are quoted (Rig pa’i khu byug, rTsal chen sprugs pa, Khyung chen lding ba, and Mi nub pa’i rgyal mtshan), and two of four are quoted extensively—Khyung chen is quoted 18 times, and 48 lines out of its total of 108 are quoted, while Mi nub pa’i rgyal mtshan is cited 34 times, and 108 lines out of its total of 219 are quoted. Neumaier’s study is the study that brought these facts to the attention of scholars.

The main question that the Neumaier’s study raises is whether the extant

Sems sde texts available to us in several editions are “the same” ones that the author of the bSam gtan mig sgron, gNubs chen Sangs rgyas ye shes, had access to in the tenth century. Neumaier suggests that that may not be the case as the textual bodies attached to individual titles “may have undergone significant change in the course of history” (2). The only way to answer the question properly is to establish critical editions of the five Sems sde texts, and in the process collate the bSam gtan mig

INTRODUCTION 14 sgron quotes against the “similar” text, as found in other extant editions, such as the rNying ma’i rgyud ’bum, the Bai ro’i rgyud ’bum, and the bKa’ ’gyur.

Karen Liljenberg, at the University of London (School of Oriental and

African Studies, UK), has completed her dissertation on the “Thirteen Later” (Phyi bcu gsum) texts of the Tibetan Buddhist tradition rDzogs chen Sems sde (2012)— texts attributed as translations to Vimalamitra (8th century). Dylan Esler, a doctoral student at the Université Catholique de Louvain (Belgium), is translating the bSam gtan mig sgron in English. In 2010, Baroetto translated chapter seven of the bSam gtan mig sgron, the one that deals primarily with rDzogs chen, in Italian. These studies are complemented by the articles on rDzogs chen that have been published by Liljenberg (2012, 2018), Esler (2012), van Schaik (2012), Valby (2012), and

Wilkinson (2012). The present dissertation aims to contribute to this dialogue, and, thus, enhance our understanding of the ancient Tibetan Buddhist tradition, the rDzogs chen Sems sde.

HISTORICAL OVERVIEW 15

CHAPTER 2

Lineage History and Historical Overview of the Early Period of

the Sems sde Tradition

The earliest rDzogs chen texts are those that today are known as Sems sde,

“the Mind Series” texts. The Sems sde itself refers to both a group of texts as well as the early form of rDzogs chen discourse characterized by a constant rhetorical denial of the validity and relevance of rituals and contemplative techniques. Later developments moved away from this central theme and gradually incorporated complex doctrines and ritual-contemplative practices. David Germano (2005) has coined two useful phrases to distinguish between the two forms of discourse: (i)

“pristine” rDzogs chen, which literally consists of “aphoristic poetry” with “terse experiential descriptions” lacking any detailed outline of practice; and (ii)

“visionary” or “tantric” rDzogs chen, related to wider, normative tantric discourse and praxis.

As early as the ninth century, there was a recognizable form of the pristine rDzogs chen, yet the term sems sde, denoting a category, does not appear in any pre-eleventh-century texts (van Schaik 2004: 167, n. 6). In the eleventh century emerged a new tradition, tantric rDzogs chen. The contents and practices of these two forms of rDzogs chen often differed greatly and were highly critical of each

HISTORICAL OVERVIEW 16 other. In this context, tantric rDzogs chen rhetorically understood itself as the pinnacle of Buddhist teachings and positioned itself as such in contrast to other traditions. It developed a hierarchical classificatory scheme, rDzogs chen sde sum

(the Three sections of the Great Perfection), which described the Sems sde (Mind

Series) and the kLong sde (Space Series) as inferior to the Man ngag sde (Esoteric

Precept Series) (Anspal 2005: 37). It is not yet clear to what extent the retroactive categories are mere “artificial labels,” or whether they are labels firmly grounded in the literature and denote traditions’ own self-representation (Germano 2005: 8).

What is unquestionable, though, is that each series refers to a specific group of texts and a particular form of discourse.

2.1 The nature of the rDzogs chen Sems sde

The rDzogs chen Sems sde is a loose category that covers the majority of the Sems sde texts written prior to the eleventh century. These texts, it could be said, are “bound together” by motifs and terminology characteristic of the early period, that is, the pristine ritual-free form of discourse. The largest and most well known of these texts is the Kun byed rgyal po’i mdo, which incorporates the sNga lnga texts as chapters (Neumaier-Dargyay 1992). In this text, the “mind of perfect purity” (byang chub sems) is personified as the teacher Kun byed rgyal po (the All-

Creating Sovereign). Sems dpa’ rdo rje, the main character of the text, is in dialogue with Kun byed rgyal po, and as such in dialogue with his own mind—the

HISTORICAL OVERVIEW 17 mind of perfect purity. The aim of the text is to introduce the true nature of the mind, that is, the “self-originated pristine awareness” (rang byung ye shes) of the mind of perfect purity, to the disciple Sems dpa’ rdo rje (and thus to the reader) who must recognize or “discover” it in a direct, non-conceptual manner.

In his teachings, Kun byed rgyal po directs and guides Sems dpa’ rdo rje.

He urges him to “listen” and to “direct his mind” to the instructions that he is about to receive. He also points out to him that he, Sems dpa’, needs to understand and will be “led” to understand. The term “understand,” in this context, is used in a practical, experiential sense, rather than a theoretical, intellectual way, as Kun byed rgyal po warns Sems dpa’ that if he does not understand the real meaning of pristine awareness, he will never meet his own mind of perfect purity, “regardless of how many words he has learned” (Neumaier-Dargyay 1992: 59). The meaning of byang chub sems (Sanskrit bodhicitta), when used in technical terms, is quite different in the Sems sde from the way it is understood in Mahāyāna and Vajrayāna forms of Buddhism. In the Mahāyāna, byang chub sems is understood as the “mind of enlightenment” that is cultivated through the practice of compassion and emptiness. In the Vajrayāna, it connotes the symbolic union of compassion and emptiness (sexual union of man and woman) and refers to the male and female sexual fluids in the body. In the Sems sde, byang means “pure” or “purified”, chub means “perfect” or “perfected”, and sems literally means “mind”: the mind of

HISTORICAL OVERVIEW 18 perfect purity—the true nature of the mind (See Neumaier-Dargyay 1992: 28, and

Norbu and Clemente 1999: 58).

Kun byed rgyal po “reveals” to Sems dpa’ rdo rje that he, as the mind of perfect purity, is no other than Sems dpa’s own mind free of subject and object and, therefore, non-conceptual. The very nature of this mind, “pristine awareness,” is all-encompassing like the sky. It is beyond striving and achieving, spontaneously self-perfected. The mind of perfect purity is the root of all instructions and the climax of all teachings; yet, there is no path that leads towards it and no doctrine that could get hold of it. Kun byed rgyal po entreats Sems dpa’ to get rid of his desires for achievement and result—desires that only obscure the mind of perfect purity. Doctrine and practice, pursuit and abandonment, striving and achievement are nothing but errors and obscurations with regard to “the mind of perfect purity.”

The Sems sde teachings do not impose a new, conditioned structure on the mind, in terms of doctrinal views, but rather reveal what is already present—the mind of perfect purity that all human beings are endowed with. “Pristine awareness” and “the mind of perfect purity” are metaphorical expressions that aim to “illuminate” and directly introduce one to the nature of the mind, to his or her own natural perfection. These concepts do not denote a particular, objectified form of reality and should not be reified. It would seem that the sole requisite for attaining the Sems sde “vision” is a wide-open mind. There is nothing in the tradition that can be grasped or conceptualized, cultivated or practiced.

HISTORICAL OVERVIEW 19

The sNga lnga texts point out that the nature of reality is non-dual, construct-free, and already perfect as it is. The sphere of the here-and-know is timeless; it transcends the three times (past, present, and future) and is not bound by concepts such as right or wrong, good or bad. The mind of perfect purity dwells in the self-originated pristine awareness. It cannot be changed, transformed, or accessed through exertion, discipline or meditation. Nor can it be captured or expressed in words. “Pristine awareness” is free of all speculative thinking and one who dwells in it abides in a sky-like state. Discursive ways of thinking and the presumption that a realization can be obtained through analysis, as well as the conceptual, progressive, and goal-oriented dharma practice, are all errors that should be avoided. They preclude a spontaneous realization. The Kun byed rgyal po points out that the true nature of reality cannot be found by seeking it; it is already present, therefore, no doctrine is to be contemplated, no vows to be observed, no path to be proceeded on, and no bodhisattva stages to be practiced

(Neumaier-Dargyay 1992: 73).

Based on the extant texts, it would seem that early Sems sde “practitioners” did envision some form of “practice” even though tantric and normative meditational techniques were rejected without reservation, as errors and obscurations. The texts seem to have played a central role in these practices as they displaced, at least to some extent, the actual teacher(s). Their role was to

“introduce” a disciple to his (or her) own mind, “the mind of perfect purity.” Once

HISTORICAL OVERVIEW 20 the true nature was realized, ones’ own “pristine awareness” was to become the teacher. Those, in the Kun byed rgyal po text, Sems dpa’ rdo rje’s teacher is his own “pristine awareness” personified as Kun byed rgyal po (the All-Creating

Sovereign), the Samantabhadra Buddha.

According to Germano (1995), it is not clear to what extent and in what practices, if any, the early practitioners may have been involved in. However, he

“envisions” five possibilities: 1) “semantic contextualization,” simply reinterprets visualization practices of the deity yoga (practices that they were already involved in); 2) “calming technique,” emphasizes calming, meditational techniques that serve to integrate and universalize; 3) “formless meditation,” focuses on sessions of sitting meditation (some strand of Chan) devoid of any specific technique; 4)

“poetically thematized meditation,” creates new practices in which one employs oral instructions or texts to engage in guided reveries and poetic contemplation; and

5) “non-meditation,” simply dictates doing as one pleases and using the teachings to justify one’s lifestyle (228-229).

2.2 The mythic history of the rDzogs chen tradition

According to the traditional accounts, rDzogs chen teachings were

“introduced” in Tibet at the time of the first diffusion of Buddhism, the reign of

Tibetan king Khri srong lde btsan (755-797). The teachings were transmitted to

Tibetans by a succession of highly revered Indian masters. The supreme Buddha,

HISTORICAL OVERVIEW 21

Samantabhadra (Kun tu zang po), revealed the texts to Vajrasattva, who, in turn, transmitted them to a human master dGa’ rab rdo rje. The lineage-lists thus include both mythical as well as historical figures.

A tenth-century Sems sde text, Pan sgrub rnams kyi thugs bcud, gives a list of 25 teachers, starting with Samantabhadra and ending with Vairocana. The lineage-list in this particular text is as follows: Kun tu bzang po (Samantabhadra); rDo rje sems dpa’ (Vajrasattva); dGa’ rab rdo rje; ’Jam dpal bshes gnyen

(Mañjuśrīmitra); rGyal po ’Da’ he na ta; Sras thu bo Ha ti (Rājahasti); Sras mo Pa ra ni; rGyal po Yon tan lag gi bu mo Gnod sbyin mo byang chub; rMad ’tshong ma

Par na; Kha che’i mKhan po Rab snang; U rgyan gyi mKhan po Ma ha ra tsa (King

Indrabhūti); Sras mo Go ma de byi (Princess Gomadevī); A rya A lo ke; Khyi’i rgyal po Gu gu ra tsa (Kukkurāja); Drang srong Ba sha ti (Bhāṣita); rMad ’tshong ma Bdag nyid ma; Na ga ’dzu na (Nāgārjuna); Gu gu ra tsa phyi ma (the later

Kukkurāja); ’Jam dpal bshes gnyen phyi ma (the later Mañjuśrīmitra); Lha’i mkhan po Ma ha ra (Devarāja); Bud dha kug ta (Buddhagupta); Shri Sing nga (Śrī

Siṃha);1 dGe slong ma Kun dga’ ma; Bye ma la mu tra (Vimalamitra); and ’Phags pa Bai ro tsa na (Vairocana) (Kapstein 2008: 279-280). The list provides only brief remarks on the circumstances of transmission, and may have been the basis for later texts, such as the biography of Vairocana, ’Dra ’bag chen mo (The Great

1 The linage-lists are either “long” or “short,” and the difference is that in the “short” list the first Mañjuśrīmitra transmits the teachings directly to Śrī Siṃha. See Norbu and Clemente 1999: 27; and Valby 1983.

HISTORICAL OVERVIEW 22

Image). Of the 25 masters, the most important are dGa’ rab rdo rje, Mañjuśrīmitra,

Śrī Siṃha, Vimalamitra, and Vairocana. At present, there is no conclusive evidence that would confirm the claim that rDzogs chen teachings did, in fact, originate in

India and, as Achard (2013) rightly points out in his essay, “The Tibetan Tradition of the Great Perfection,” historians “are still at pains to describe the early history of rDzogs chen in any satisfactory way” (2).

2.2.1 dGa’ rab rdo rje

dGa’ rab rdo rje (7th century?), the first human master and as such the founder of the rDzogs chen, codified the teachings into a written form. He is included in the lineages of all three traditions (Sems sde, kLong sde and Man ngag sde). A legendary figure shrouded in mystery, he is best known by his Tibetan name, dGa’ rab rdo rje. Modern scholars have re-Sanskritized his name as either

Surativajra, Pramuditavajra, Ānandavajra, or Prahevajra.2 There is no certainty as to when he actually lived. In the most elaborate account of his life, that given by kLong chen pa (1308-1364) in his work rDzogs pa chen po snying tig gi lo rgyus chen mo, he was born sometime in the year 55 to 60 of the common era (Valby

1983). Hanson-Barber (1986) considers these early dates as unreasonable. Based on

2 Based on the Mahāvyutpatti (8th century), Hanson-Barber (1986) suggests that the previous attempts by Suzuky and Guenther to re-Sanskritize dGa’ rab rdo rje’s name as Surativajra and Pramuditavajra are incorrect, and argues instead for the Ānandavajra. Reynolds (1996), with some reservations, adopts the Sanskrit name Prahevajra.

HISTORICAL OVERVIEW 23 a standard of thirty-five-year chronological difference between student and teacher, he suggests that perhaps a more acceptable date would be 550 CE.3

The ’Dra ’bag chen mo mentions that dGa’ rab rdo rje was born to a

Buddhist nun (bhikṣunī) named Bharaṇī, but was without a father. He was born in

Dhanakośa, a region in Oḍḍiyāna, where his grandfather Dhahenatalo ruled as king.

One morning, while Bharaṇī was bathing in a nearby lake, a golden swan

(Vajrapāṇi) approached her and pecked at her lightly three times. A year later she gave birth to a boy who recited a rDzogs chen text as soon as he was born.4 It would seem that he spent most of his life in solitude, high in the mountains or in the charnel grounds and cemeteries. He received teachings directly from

Vajrasattva and was able to defeat numerous paṇḍitas (scholars) in various debates.

It was during one of such debates that he met Mañjuśrīmitra, his best-known disciple.

The accounts on the manner in which the “final transmission” took place, between the master dGa’ rab rdo rje and the disciple Mañjuśrīmitra, differ considerably in the three rDzogs chen traditions. In the Sems sde, the final transmission is expressed as a “song of realization.” After he has transmitted all the scriptures, and conferred upon Mañjuśrīmitra the complete empowerments, dGa’

3 Reynolds (1996) is critical of Hanson-Barber’s methodology and views the so-called “standard of thirty-five years” as “arbitrary” and “dubious” (208-211). 4 The text that dGa’ rab rdo rje recited was rDo rje sems dpa’ nam mkha’ che (The Total Space of Vajrasattva). See Palmo 2004: 14. In the Man ngag sde tradition Bharaṇī is known as Kudharma and she becomes pregnant when she dreams that a white man comes and touches the crown of her head, three times, with a crystal vessel (Valby 1983: 10-11).

HISTORICAL OVERVIEW 24 rab rdo rje summarized for him the essence of his teaching. Mañjuśrīmitra, in return, acknowledged and expressed his own realization (Palmo 2004: 46-47). In the kLong sde, the transmission is expressed in a “symbolic form.” dGa’ rab rdo rje placed a vajra on Mañjuśrīmitra’s head and uttered the syllables A HA HO YE, to indicate the ultimate reality of his mind (Anspal 2005: 28).5 In the Man ngag sde, the transmission takes the form of a “treasure chest.” At his death, dGa’ rab rdo rje ascended to the sky where he was seen sitting in a circle of radiant light. Out of the sphere, he stretched out his arm and placed in Mañjuśrīmitra’s hand a jeweled box

(the size of a thumb) containing his last testament (Valby 1983: 18-19). In later times, all major masters of the three rDzogs chen traditions followed these

“patterns” for final transmission.

2.2.2 Mañjuśrīmitra

Mañjuśrīmitra (’Jam dpal bshes gnyen), a renowned Indian Buddhist scholar in the middle of the eighth century, was a bhikṣu (monk) associated with the great Nālanda monastery. One of his most important texts, still extant, is a commentary on a very popular canonical work that centers on Mañjuśrī, the

Mañjuśrīnāmasaṃgīti (Davidson 1981). Together with another six respected scholars, he was sent to debate with dGa’ rab rdo rje and refute his rDzogs chen

5 Anspal’s comments are based on Kun bzang rdo rje’s sNyan brgyud rdo rje zam pa, a twelfth century text in the kLong sde tradition.

HISTORICAL OVERVIEW 25 view of non-action (beyond cause and effect), a view that was said to transcend those of Mahāyāna and Vajrayāna. Unable to defeat him, the other scholars returned home, but Mañjuśrīmitra decided to remain and became dGa’ rab rdo rje’s disciple (Palmo 2004: 46-47).

The ’Dra ’bag chen mo (Palmo 2004: 47) mentions that Mañjuśrīmitra wrote a text, rDo la gser zhun, after he attained a perfect understanding of the rDzogs chen teachings. This text is also known as Byang chub kyi sems sgom pa

(Meditation on the Enlightened Mind), and a text with a similar title is ascribed to

Mañjuśrīmitra in the lDan dkar ma (lHan dkar ma) catalogue (no. 610) (Lalou

1953: 354). According to Herrmann-Pfandt (2002), the catalogue is a list of translated texts made in 812 but “added to later,” as it contains titles translated after

830 (135). The lDan dkar ma is the earliest surviving, first-hand datable source that cites a Sems sde text—one that most probably already existed in the eighth century.

Some scholars consider Byang chub sems sgom pa (still extant) to be one of the

“five earliest” texts (See Norbu and Lipman 1987: 6-11). However, text-critical preliminary research suggests that this is not the case, as the text is not a “root” text. The text is an exegetical work about the Victorious Ones (rgyal ba), described in third person, and their realization. It is expository in nature, its aim is to reaffirm what has already been said and proclaimed. The author points out that he has validated for himself the “excellent path” taught by the Victorious Ones (Norbu and Lipman 1987: 68). Mañjuśrīmitra, the ascribed writer, may or may not have

HISTORICAL OVERVIEW 26 been the actual author of the text, as the tradition claims and the catalogue suggests, but it is doubtful that he was dGa’ rab rdo rje’s disciple. The tradition, it seems, associated the legendary dGa’ rab rdo rje with the respected Mañjuśrīmitra (who represented a “solid orthodoxy”) to provide some semblance of legitimacy to an otherwise obscure tradition in a period of development.

2.2.3 Śrī siṃha

Śrī siṃha (8th century), like his predecessors, is another legendary master associated by tradition with the transmission of the rDzogs chen teachings to Tibet.

He was, and still is, a prominent figure in all three traditions. Most accounts seem to agree that he studied with Mañjuśrīmitra, and only the ’Dra ’bag chen mo is somewhat inconsistent in this respect. In the “general lineage” (the long list that cites twenty-five masters), his master is given as Buddhagupta, whereas in the

“special lineage” (the shortlist) the master is Mañjuśrīmitra—the shortlist excludes everyone who is in between Mañjuśrīmitra and Śrī siṃha (Palmo 2004: 48-70, 80).

Śrī siṃha was born in China in the town of So sha’i ling.6 Starting at the age of fifteen, he studied for three years all five branches of learning (grammar, astrology, literature, logic, and Dharma) with his Chinese teacher, Hatibhala. After that time, he traveled to Ri bo rtse lnga, in eastern China, where he continued his

6 Germano (2002: 239) suggests that Śrī siṃha, an Indian, may have been associated with China by later accounts (12th century) to express the “international nature” of rDzogs chen, and thus to legitimize it.

HISTORICAL OVERVIEW 27 studies for seven more years. During this time he devoted himself to various outer and inner aspects of Vajrayāna under a Chinese master named Bitekiti. Later on, he became a monk, observing the vows for thirty years before he traveled to India. In

India, he met his master Mañjuśrīmitra; he studied with him for another twenty-five years. He received the full transmission and became proficient in meditation (Valby

1983: 22-24). After Mañjuśrīmitra passed away, he retired to charnel grounds where he first taught Jñānasūtra, then Vimalamitra, and finally Vairocana.

2.3 The historical period (8th century)

According to the traditional historiography, all rDzogs chen teachings, whether Sems sde, kLong sde or Man ngag sde, were introduced in Tibet during the eighth century by Vairocana, Vimalamitra, and Padmasambhava. The Sems sde teachings fall under the rubric of “pristine” rDzogs chen; they are represented by the pre-tenth-century texts, such as the Kun byed rgyal po’i mdo, and a group of

“eighteen Sems sde texts” (Sems sde bco brgyad): five are called the “five earlier translations” (sNga ’gyur lnga) of Vairocana, and thirteen the “thirteen later translations” (Phyi ’gyur bcu gsum) of Vimalamitra. The kLong sde teachings are ascribed to Vairocana and include both “pristine” and “tantric” forms of rDzogs chen. The Man ngag sde teachings are associated with Vimalamitra and

Padmasambhava and are primarily “tantric” in their form. Most of the available

HISTORICAL OVERVIEW 28 textual material concerning the kLong sde and the Man ngag sde traditions was composed during the eleventh to the fourteenth centuries.

The Sems sde and the kLong sde texts are considered bka’ ma (transmitted teachings), in contrast to the Man ngag sde texts considered gter ma (rediscovered teachings). bKa’ ma texts are those that have been passed on throughout centuries, from a master to a disciple and in an unbroken succession. gTer ma texts are those concealed in the past by a renowned teacher, most often by an eighth-century master, such as Padmasambhava or Vimalamitra, and later rediscovered by a gter ston (a treasure revealer). The gter ma texts are ahistorical in the sense that their transmission is not concerned with history or historical development. Thus, the ultimate source of a gter ma is a Buddha (Samantabhadra, Vajradhara or

Vajrasattva), that resides in a timeless realm, a Buddha-field. The Buddha transmits the text to a master, such as Padmasambhava, who codifies the text into a written form and conceals it as a “treasure” to be rediscovered sometime in the future. A gter ston, based on the prophecy made by the concealer as well as his own vision and spiritual realization, reveals this cryptic text, then renders it into an intelligible form. Furthermore, the text can be re-concealed again and later rediscovered by yet another gter ston. In this process, a text “rediscovered” in the nineteenth century could be regarded either as an authentic thirteenth-century text, when it was

HISTORICAL OVERVIEW 29 discovered for the first time, or as a dynastic period text (8th century), when it was concealed for the first time.7

Padmasambhava (8th century) was a famous magician and spiritual tantric master from Oḍḍiyāna. Apart from the legends that have developed about him, he has to this day remained an enigma. The later historical accounts mention he took part in the foundation of the first great Buddhist monastery (bSam yas), and played a crucial role in hiding treasure texts throughout Tibet that were to be revealed by future generations. Allegedly, he hid the mKha’ ’gro snying thig text in a secret place and taught some of its teachings to a small circle of twenty-five disciples, including the king Khri srong lde btsan (reigned 755-797) and the scholar

Vairocana.8 The text was later rediscovered by the gter ton Padma las ’brel (fl.

13th/14th c.) and has ever since been considered an eighth-century Man ngag sde work. The only extant rDzogs chen text directly attributed to Padmasambhava is the Man ngag lta ba’i phreng ba. It is a fundamental text in the Man ngag sde tradition; its ascription to Padmasambhava dates back to the tenth century, but the actual author of the text remains unknown (Karmay 1988: 137-174).

7 gTer ma, lit. “treasure,” may refer to a specific teaching or to a physical object, such as a jewel or an icon. On gter ma see Dargyay 1977: 62-210; Tulku Thondup 1986; and Gyatso 1986. The first gter ston seems to have been Sangs rgyas bla ma (ca. 1000-1080). Nyang ral Nyi ma ’od zer together with Gu ru Chos kyi dbang phyug (1212-1270), rDo rje gling pa (1346-1405), O rgyan Padma ling pa (1459-1521), and ’Jam dbyangs mKhyen brtse’i dbang po (1820-1892) are considered as the most prominent, and are known as the Five Discoverer Kings. See Tulku Thondup 1986: 72. 8 For a list of Padmasambhava’s disciples, see Tulku Thondup 1986: 231-234.

HISTORICAL OVERVIEW 30

According to the dBa’ bzhed chronicle, one of the most celebrated texts of the Tibetan historiography that dates back to the dynastic period, Padmasambhava played only a minor role in the early dissemination of Buddhism in Tibet. King

Khri srong lde btsan sought to establish Buddhism in his country; to that end, he invited Śāntarakṣita (ca. 723-787) from Zahor to Tibet. Śāntarakṣita arrived and began to teach, but there were fierce storms and floods, fires, epidemics and famines. These were interpreted as signs that the local demons (wicked gods, nāgas, and evil spirits) were displeased. To placate the ministers, Śāntarakṣita moved to Nepal but suggested to the king he should invite the tantric master

Padmasambhava to Tibet. Padmasambhava was invited. Once there, he was able to subdue the demonic forces with his mantric abilities. This, in turn, allowed

Śāntarakṣita to return and establish a monastic order in Tibet. However,

Padmasambhava, unsuccessful in pleasing either the king or the ministers, who suspected he might try to seize the power of the kingdom, was asked to leave the country (Wangdu and Diemberger 2000: 40-59).

Vimalamitra (8th/9th century) is linked to both the Sems sde and the Man ngag sde traditions. He, like Vairocana, is consistently associated with the early

Sems sde. This involvement seems to suggest that by this time, in the late eighth and early ninth century, Sems sde tradition was emerging as a distinct movement.

Some scholars (Garson 2004) argue that there may have been two Vimalamitras: one who studied with Buddhaguhya, commented on, and promoted the Mahāyoga

HISTORICAL OVERVIEW 31 tantras; and another who studied with Śrī siṃha, “translated” Sems sde texts, and was involved in the dissemination of the Man ngag sde texts by concealing them, with the intent they are revealed at a later time (176-187).

In one of the best-known stories of Vimalamitra’s life, he was born in the western Indian town of gLang po’i tshal; he later lived and studied with five- hundred other Buddhist scholars in Bodhgaya. One day, he had a vision of

Vajrasattva, who directed him to go to China and study with Śrī siṃha. He left for

China; for twenty years he received instructions from Śrī siṃha. Satisfied, he returned to India. There, he met a learned scholar, Jñānasūtra (Ye shes mdo), and informed him about Śrī siṃha. Jñānasūtra also went to China where he studied with

Śrī siṃha for sixteen years. From the master he received the instructions as well as the actual texts, later to share them with Vimalamitra (Valby 1983: 28-41).

Vimalamitra was invited to Tibet by king Khri srong lde btsan (around

795). He remained in Tibet for thirteen years. During this time, he devoted himself to the translation of various texts on rDzogs chen; he also instructed some of his disciples, such as the king and his minister Myang Ting nge ’dzin, in the Man ngag sde teachings of the Seventeen Tantras. After Vimalamitra left Tibet for China,

Myang gave the exposition of the Seventeen Tantras to ’Brom Rin chen ’bar, but concealed the texts in the temple of Zhva. More than two centuries later, towards the end of the eleventh century, the monk lDang ma lHun rgyal (the caretaker of the temple) discovered the hidden texts; he then showed them to lCe btsun Seng ge

HISTORICAL OVERVIEW 32 dbang phyug, who edited them. Another collection of five texts, Bi ma snying thig, also attributed to Vimalamitra, was later rediscovered by lCe btsun in mChims phu

(Karmay 1988: 209-211). These collections, together with mKha’ ’gro snying thig, that is assigned to Padmasambhava, are fundamental to the Man ngag sde tradition.

Vairocana (8th/9th century) was a Tibetan, one of the first of his contemporaries to be ordained by Śāntarakṣita at bSam yas. He is also considered to be the one who introduced rDzogs chen into Tibet from India. However, above all he is known to have been a great Lo tsa ba, that is, a translator of Buddhist canonical works. He collaborated closely with Indian paṇḍitas. Vairocana’s life, like that of all other founding patriarchs of rDzogs chen, is shrouded in mystery and legend. Practically nothing is known about the historical Vairocana.

In the ’Dra ’bag chen mo, his alleged biography, it is mentioned that he was born into the Ba gor clan. As a young boy, he studied Sanskrit and was later was chosen by king Khri srong lde btsan to go to India in search of Buddhist teachings.

Vairocana, in the company of another young monk called Legs grub, managed to reach India after several “trials” (harassment from malevolent spirits, robbers, wild beasts, and other similar events). In central India, the two Tibetans were told that the most learned master was Śrī siṃha, who resided in a place called Dhahena.

However, before their arrival, there was a dispute over the rDzogs chen teachings.

All paṇḍitas had suffered bad dreams; they interpreted this as a bad omen for the disclosure of the teachings and the rDzogs chen texts were hidden away. A

HISTORICAL OVERVIEW 33 proclamation was made stating that whoever reveals the teachings will be punished.

After meeting Śrī siṃha and removing the hidden teachings, Vairocana and Legs grub studied Sutras and Tantras during the day while at night learning rDzogs chen.

As Vairocana remained to deepen his understanding, Legs grub returned to Tibet but was killed on his way home.

Vairocana, after his safe return, translated the first five of the eighteen Sems sde texts, the sNga gyur lnga: Rig pa’i khu byug (The Cuckoo’s Song of

Awareness); rTsal chen sprugs pa (The Great Potentiality); Khyung chen lding ba

(The Great Garuda in Flight); rDo la gser zhun (The Refining of Gold from Ore); and Mi nub pa’i rgyal mtshan (The Unwaning Victory Banner). In the meantime, jealous Indian scholars spread the rumor that what he had attained in India were magic spells, not the teachings of the rDzogs chen. He was exiled to Tsha ba rong and returned to bSam yas (and continue his work as a Lo tsa ba) only after spending several years in Eastern Tibet (Palmo 2004: 92-239).

2.4. The “origin” of the rDzogs chen Sems sde

The “origin” of the rDzogs chen Sems sde tradition has proved very difficult to establish for modern scholarship. The polemic of whether rDzogs chen originated in India or in China is as ancient, it would seem, as the tradition itself, and modern arguments follow a pattern already established by the ancient Tibetan traditions. Indologists and Sinologists vie for the coveted yet elusive title of “the

HISTORICAL OVERVIEW 34 origin”—the one and only source that has led to the emergence of rDzogs chen teachings. The polemics often turn in heated debates as scholars become critical of the opposing views and suggestions. They aim to establish genealogical connections and provide counter-arguments that are somehow “more evident” and seem “more compelling” than the competing arguments.9

Germano (1995) suggests that present-day scholarship on the early rDzogs chen must move beyond an “either/or” type of discourse, one that posits the rDzogs chen as “a survival of Ch’an in Tibet” or “disavows any relationship whatsoever” between the two traditions (217). Yet Germano himself falls into the trap of the

“origin” when he states that even the earliest rDzogs chen tradition (Sems sde, presumably) was “clearly, [and] profoundly tantric in character.” This, apparently, indicates that “its principal roots” were non-Ch’an in origin (217). Germano does offer a caveat, however, as he points out that this is true about rDzogs chen “by at least the latter half of the tenth century” (217). It would seem that the confusion arises because scholars tend to conflate the three traditions of rDzogs chen (Sems sde, kLong sde, and Man ngag sde).

The term rdzogs chen was used for the first time in the Guhyagarbha-tantra

(four times), a tantra that may have been circulating in India by the mid-eighth

9 For a Sinologist position, see Tucci 1958; and Dargyay 1977. For an Indologist position, see especially the work of van Schaik (2004, 2012), who tends to be biased in favor of Indian sources and dismisses any roll that Chinese Buddhist traditions may have played in the emergence and/or the development of the rDzogs chen teachings.

HISTORICAL OVERVIEW 35 century. The term was used to indicate the high level of realization that results from

“the sexual yoga of the perfection stage.” By the late eighth century, the term was used in tantric commentaries as the culmination of the development and the perfection stage: that is, the realization that “all phenomena have been the pure state of the deity” from the very beginning (van Schaik 2004: 178-79). The first datable use of the term rdzogs chen as a vehicle (theg pa) is in bSam gtan mig sgron, by gNubs chen (10th century), where it refers to Sems sde texts. According to van Schaik, the same scholars who were writing exegetical works on the tantric texts were also writing the early, pristine rDzogs chen Sems sde texts, and these activities, apparently, were not considered to be “inconsistent” (200).

The problem with van Schaik’s assumption, that “two [totally] different kinds of texts” were being written within “the Māyājāla [tantric] commentarial traditions” (201) by the same authors, is that all early Sems sde texts declare tantric beliefs, rituals, and practices as unnecessary and detrimental to a spontaneous presence. They deny the validity and relevance of normative tantric principles that define Mahāyoga. In these early texts, there is no meditative cultivation, no preserving of commitments, no path to traverse, and no esoteric precepts. Sems sde’s antinomian rhetoric could potentially threaten not only Mahāyoga tradition’s logical integrity but also its very existence and continuity (if taken literally as practical instructions and not only as philosophical speculations), and the question is then: would a tantric practitioner intentionally denigrate the tradition that he (or

HISTORICAL OVERVIEW 36 she) belongs to, the lineage, and his (or her) samaya (vows and precepts)? That does not seem to have been the case as most of the commentaries were primarily concerned with “interpreting” specific tantric practices, such as the ritual practice of killing, “liberating” (sgrol ba), and sexual union, “unifying” (sbyor ba)— practices that are to be found in the Guhyagarbha-tantra. When tantric exegetes did appeal to Sems sde teachings, they used them to provide an interpretive framework for the Mahāyoga rituals and practices, that is, a philosophical speculation; however, this “speculation” had no function of its own and no direct bearing on the actual practices.10 Nevertheless, van Schaik may still be correct if he restricts the influence of these texts to the proto-tantric form of rDzogs chen, that later came to be known as the Man ngag sde tradition.

The historical and extant textual evidence would seem to suggest that the rDzogs chen Sems sde tradition originated in Tibet itself. The Tibetan creative genius, in the eighth century, gave rise to the emergence of the pristine rDzogs chen, based on the interaction with the tantric and the Ch’an forms of Buddhist teachings. The Ch’an form of Buddhism that was influential in Tibet was that taught by the master Heshang Moheyan. Sometime during the eighth century, as the Tibetan historians tell us, a controversy arose in Tibet over two approaches to

10 This seems to have been a common practice in all tantric tradition, not just in Mahāyoga. Takahashi (2009) points out that in most Indian and much of Tibetan tantric literature, doctrine “merely serves to support the import and function of the rites, without an independent function as a sotereological tool” (274).

HISTORICAL OVERVIEW 37 enlightenment; it was to be settled in a debate. The first position was that of the

Chinese side, represented by the master Heshang Moheyan: his was “the simultaneous approach,” in that there was no need for gradual, lengthy practices as enlightenment was spontaneous; and the second position was that of the Indian side, represented by the scholar Kamalaśīla: this was “the gradual approach,” advocating that practices were indispensable. According to the Tibetan version of the story, Moheyan was defeated, and his method was subsequently rejected

(Gomez 1983b; and Ruegg 1992). Whatever may have happened at the debate, or whether it even took place or not, is debatable (see Demieville 1952, especially

169-178; Imaeda 1975; and Guenther1983). However, it is undeniable that Ch’an teachings, just as well as the tantric teachings, had a significant and lasting impact on the Tibetan form of Buddhism, the rDzogs chen Sems sde. Chapter 3 will focus on the oral and textual transmission of the Sems sde teachings, and attempt to establish that the rDzogs chen Sems sde tradition is a Tibetan tradition that originated in Tibet.

ORAL AND TEXTUAL TRANSMISSION 38

CHAPTER 3

Oral and Textual Transmission of the Sems sde Teachings

The rDzogs chen Sems sde tradition surfaced in Tibet during the reign of king Khri srong lde btsan (755-797). Most scholars tend to agree that its inception was in the eighth century; however, the search for an origin has proved futile. As the details of its emergence have been lost in the mists of time, the early history of

Sems sde remains unclear and subject to controversy, while the definitive source of the Sems sde teachings, the “primary impetus” and “original genesis,” is still as elusive as ever, in spite of the fact that several modern academic works have been published on rDzogs chen. Sems sde’s controversial nature primarily stemmed from its questionable claims of being “largely of Indic origin,” its “creative innovations,” and its “strikingly antinomian” language (Germano 2007: 293).

The “inception” of the Sems sde tradition (if there ever was one) may have been under largely unknown circumstances, however, as Germano (2007) points out, its subsequent development was clearly a Tibetan phenomenon “drawing on diverse” Buddhist traditions (293). It is well documented that during this time, both

Ch’an and Mahāyoga teachings were active in Tibet; nevertheless, these teachings were not Chinese and Indian, per se, but rather Tibetan. These teachings belonged to the “new” traditions of Ch’an and Mahāyoga that were emerging in Tibet. As

ORAL AND TEXTUAL TRANSMISSION 39 such, the “primary impetus” for the emergence of the rDzogs chen Sems sde tradition may not have been the Chinese and the Indian textual sources (as some scholars suggest), but the teachings of the Tibetan Ch’an and the Tibetan

Mahāyoga. In this context, as all these traditions were emerging approximately at the same time, perhaps, it would be better to focus on the co-emergence of these traditions, instead of focusing on “genealogical connections” (which tradition evolved from which one). This was an “interlinked set of movements” (Germano

2007: 293), and the actual contemporary relationship between these traditions was probably one of “dynamic mutual interaction and interdependence” (Liljenberg

2012a: 29)—taking place in both oral and written form.

Some present-day scholars, such as Studstill (2005), have identified five possible scenarios for the historical emergence of rDzogs chen: (1) rDzogs chen evolved from Indian Buddhist tantra; (2) rDzogs chen evolved from Indian

Buddhist tantra strongly influenced by Ch’an; (3) rDzogs chen is a syncretism of tantra and Ch’an; (4) rDzogs chen evolved from Ch’an strongly influenced by

Indian Buddhist tantra; and (5) rDzogs chen evolved from Ch’an. Based on the fact that early rDzogs chen texts contain both tantric and Ch’an elements, Studstill considers that (1) and (5) can be rejected, as they reflect nothing more than a sectarian “wishful thinking,” and any of the remaining three possibilities (two through four) is as valid as any other (2005, 146).

ORAL AND TEXTUAL TRANSMISSION 40

Studstill’s suggestion has some merit, especially if we focus exclusively on

“genealogical connections,” and restrict our analysis to Chinese Ch’an texts and

Indian Mahāyoga tantras. Still, there is a major problem with this approach as we have very few Chinese and Indian textual sources, from the time in which they are supposed to have been exerting their influence on rDzogs chen (the eighth and early ninth centuries), and even some of these are of a questionable nature. Thus, the primary textual material for the Indian Mahāyoga is the Guhyagarbha Tantra and a few related tantras of the Māyājāla class. However, Guhyagarbha has been the focus of much controversy. The eleventh-century translator, ’Gos khug pa lHa btsas, questioned its authenticity and suggested that it was composed by rMa Rin chen mchog (8th century)—a Tibetan translator of many tantric works belonging to the Māyājāla class, including the Guhyagarbha (Wangchuk 2002: 275-77). The tantra is extant in three versions (short, middle, and long), but only in the Tibetan language—there is no Chinese translation, and no Sanskrit original has been found.

Furthermore, the great majority of its associated literature exists only in Tibetan

(Takahashi 2009: 167-68). Spar khab, an early Guhyagarbha commentary ascribed to the Indian scholar Vilāsavajra (8th century), was also considered to be apocryphal. This text was apparently composed by Zur chen Shakya ’byung gnas

(1002-62) (Wangchuk 2002: 274). Chinese Ch’an textual sources do not fare much better than the Indian Mahāyoga sources. For example, even though the name of the Ch’an master Heshang Moheyan is “the most prominent one” in the Tibetan

ORAL AND TEXTUAL TRANSMISSION 41

Ch’an literature, we have no Chinese text attributed to him, but numerous Tibetan texts (Broughton 1983: 16, 52 n. 9).

Most extant Tibetan textual materials belong to the “new” traditions that were emerging and developing in Tibet, during the late eighth and early ninth centuries: rDzogs chen Sems sde, Tibetan Ch’an and Tibetan Mahāyoga. This would suggest that scholars interested in the early days of rDzogs chen would be better off if they concentrated on the “new” traditions, instead of focusing on the

Chinese and the Indian traditions. In the Ch’an context, Ueyama (1983) points out that Ch’an that had formed in China was “uniquely transformed” once it was in

Tibet, and reconstructing the history of the Tibetan Ch’an is an important and

“unavoidable” task for the study of early Tibetan Buddhism (343-44; see also Ying

2010). The same seems to have been true in regard to Mahāyoga; according to

Takahashi (2009), once in Tibet, Mahāyoga developed its own “unique Tibetan characteristics.” It moved away from tantric rituals and practices and focused instead on philosophical speculations, thus, giving rise to what might be called a

“philosophical Mahāyoga” (10).

Two of the most important textual sources for the study of the early Tibetan

Ch’an, and its interaction with the Sems sde tradition and the Tibetan Mahāyoga, are the Dunhuang manuscript Pelliot tibétain 116 (Bibliothèque nationale de

France) and gNubs chen’s bSam gtan mig sgron. The Dunhuang manuscript is a collection of nine texts that includes translations from Sanskrit and Chinese, and

ORAL AND TEXTUAL TRANSMISSION 42

Tibetan compositions. It also includes sayings of Heshang Moheyan, as well as that of several Ch’an masters (on Moheyan see Gomez 1983a). Text number VI is an interesting document as it includes quotations from eight masters in the following order: 1) na gar ju na (Nāgārjuna), 2) bo de dar ma ta la (Bodhidharmatāra), 3) bu cu (Wuzhu), 4) bdud ’dul gyi snying po (Hsiang-mo Tsang), 5) a rtan hwer

(Artanhwer), 6) ’gva lun (Wolun), 7) ma ha yan (Moheyan), and arya de ba

(Āryadeva) (Ueyama 1983: 331). The masters are not quoted in chronological order but they seem to have been important for the early Tibetan Ch’an practitioners.

Nāgārjuna and his disciple Āryadeva are the two well known Madhyamaka exegets,

Bodhidharmatāra (also known as Bodhidharma) is the first patriarch of Chinese

Ch’an, and Hsiang-mo Tsang is the teacher of Moheyan. Less well known are the

Ch’an masters, Wolun (see Meinert 2006), Wuzhu (see Adamek 2007), and A rtan hwer; however, their teachings (along with Moheyan’s) may have played an important role in the emergence of the Tibetan Ch’an. A rtan hwer is listed in the

Dunhuang manuscript Pelliot tibétain 996 in the lineage of Nam ka’i snying po. He was an Indian master who was the teacher of Bu sing ha shang (a Chinese master based in Dunhuang), Bu sing was the teacher of Man ha shang, and Man was the teacher of Nam ka’i snying po (a Tibetan from Amdo Tibet) (van Schaik 2015:

163-74).

The bSam gtan mig sgron presents four different approaches to realization, arranged hierarchically from the least to the most effective, that is, from the gradual

ORAL AND TEXTUAL TRANSMISSION 43 to the spontaneous. The first and least effective is the “gradual approach” of the

Mādhyamika as taught by Kamalaśīla (chapter four, 65-118); next is the “sudden approach” of the Ch’an as taught by Heshang Moheyan (chapter five, 118-186); better than that is the “non-dual approach” of the Mahāyoga as taught by gNyan dPal dbyangs (chapter six, 186-290); and the best is the “spontaneous approach” of the rDzogs chen as found in the Sems sde texts (chapter seven, 290-494) (see

Tanaka and Robertson 1992: 74). Each of these approaches is presented as distinct in terms of its authoritative teachings, texts, and lineage of teachers. To exemplify these views, gNubs chen quotes extensively from canonical scriptures, texts that are now only preserved in Dunhuang manuscripts, and oral teachings of the long- departed masters. As such, this text is an indispensable source for any study of the co-emergence of the three Tibetan Buddhist traditions: Sems sde, Ch’an, and

Mahāyoga.

For the Tibetan Mahāyoga tradition, the most important textual materials are the Guhyagarbha Tantra, the Dunhuang commentaries and ritual manuals on

Guhyagarbha (and other tantric texts), such as IOL Tib J 332/1 and IOL Tib J

331/2 (British Library), the bSam gtan mig sgron, and the extant texts of gNyan dPal dbyangs (9th century). The Tibetan Mahāyoga developed two forms, or streams, of tantric literature: one that was concerned with “mainstream” tantric rituals and practices—tantric yoga (peaceful and wrathful deity propitiation, sexual practice, ritual subjugation and killing, and so forth); and another one that was

ORAL AND TEXTUAL TRANSMISSION 44 concerned with philosophical speculation—philosophical perspectives that framed and gave meaning to the rites and practices. Most Mahāyoga documents from

Dunhuang address rituals in practical terms, while dPal dbyangs’ texts primarily focus on philosophical speculation (Takahashi 2009: 85).

The fact that dPal dbyangs’ works have been preserved and quoted extensively suggests that his teachings were greatly valued and widely disseminated. Thus, one of his better known and most popular texts, rDo rje sems dpa’i zhus lan (Vajrasattva Questions and Answers), an explanation of Mahāyoga, is extant in three Dunhuang versions: Pelliot tibétain 837, Pelliot tibétain 819

(fragmentary), and IOL Tib J 470. The other six texts that are known as sGron ma drug (Six Lamps) are preserved in the canonical collection, bsTan ’gyur. These texts are: Thugs kyi sgron ma (The Lamp of the Mind), lTa ba yang dag sgron ma

(The Lamp of the Correct View), mTha yi mun sel sgron ma (The Lamp

Illuminating the Extremes), Thabs shes sgron ma (The Lamp of Method and

Wisdom), bsGom thabs kyi sgron ma (The Lamp of the Method of Meditation), and lTa ba rin chen sgron ma (The Lamp of the Precious View). Furthermore, gNubs chen, in the bSam gtan mig sgron, mentions dPal dbyangs as an “authority on

Mahāyoga” more than any other figure, and in the chapter on Mahāyoga quotes twelve times from his works (Takahashi 2009: 18, 261).

In his texts, dPal dbyangs did omit “normative” tantric rituals and practices that are concerned with wrathful and sexualized deity yoga in favor of a

ORAL AND TEXTUAL TRANSMISSION 45 philosophical approach, yet, the texts were clearly written for those who are involved in tantric practices. The rDo rje sems dpa’i zhus lan touches on such topics as the importance of mantra and mudrā, the use of sacramental substances, the necessity of keeping one’s commitments (samaya), the role of ritual and meditative techniques, and other such tantric subjects. dPal dbyangs’ aim in the text, therefore, is not to reject these tantric elements, but to “re-conceptualize” them from a new philosophical perspective; a perspective that is based on the theme of spontaneous emergence of the tantric accomplishments. In his philosophical discussions, he applies the principle of spontaneous accomplishment to the system of tantric rites and practices, and, thus, his treatises point to the spontaneous states of realization. In that sense, his assertions greatly resemble that of Sems sde texts

(as well as Ch’an texts). However, his philosophical speculations are grounded in the tantric tradition, and, as such, effort, rituals, and practices are necessary as they lead to realization, even though, ultimately, they “do not cause” awakening or accomplishments (Takahashi 2009: 99).

Based on his extant texts, it is safe to say that the self-professing

Mahāyogin, dPal dbyangs, did not intend to create a new higher form of Mahāyoga, or form a Tibetan school, vehicle, or method, nor did he attempt to identify with the nascent Sems sde tradition. Rather, he employed philosophical assertions to strengthen and extend the philosophical limits of the traditional Mahāyoga perspectives. His innovative type of Mahāyoga does have an “affinity” with both

ORAL AND TEXTUAL TRANSMISSION 46

Sems sde and Tibetan Ch’an, to a lesser or a greater extent; all three traditions propound some form of sudden enlightenment, espouse a non-dual view of reality, and give primacy to the non-conceptual. As the three traditions co-emerged and interacted with each other, it is easy to identify and highlight similarities between them. They shared a common stock of terminology and perspectives that could be drawn upon by the Sems sde, Ch’an, and Mahāyoga masters; while the teachings of these emerging traditions developed simultaneously, perhaps even through the work of common practitioners, and contributed equally to their development. In other words, the teachings mutually influenced each other, and this fact was greatly facilitated by the fact that they were transmitted not only in the written form but also in the oral form. The “boundaries” between the Sems sde, Ch’an, and

Mahāyoga traditions were not as fixed, and the distinctions between their teachings were not as firm, as some scholars would suggest, and the later traditions would make us believe.

3.1 Oral transmission

One of the most pressing questions that scholars involved in the study of textual sources need to address, sooner or later, is the composition of the extant texts, that is, how were the texts composed in the first place? The question is most often left out, and the role of an oral tradition is ignored, however, it is highly likely, almost inevitable, that there must have been oral traditions pre-dating the

ORAL AND TEXTUAL TRANSMISSION 47 earliest texts. As Wynne (2004) points out, “no one can deny” that at the beginning of most Buddhist textual compositions “there must have been a period of free literary transmission” (120). Davidson (2002) also suggests that:

While the precise formative process needs to be considered scripture by scripture, the mechanism [he] believes was most influential was a procedure of diverse short explanations in the pedagogical environment, where masters imparted their understanding to disciples. These instructions were compounded into short scriptures, which were in turn included into longer tantras as chapters or parts of chapters. (205)

This process of textual composition continued to develop until it reached the point when the teachings compiled into short texts—and frequently called “adamantine phrases/verses (rdo rje’i tshig rkang)—challenged or even replaced the canonical texts as the most important sources of instruction (212-13).

Oral instructions, in one form or another, have always been important in all

Tibetan Buddhist traditions. For example, in the Sems sde tradition, the “pointing- out instruction” (ngo sprod) is considered as the most important; a direct introduction to the nature of mind, when the master identifies “pristine awareness,” and points it out to the student so that he (or she) may realize it. In the Tibetan context, there are several forms of instruction (khrid) and the two most important are: the “textual instruction or commentary” (gzhung khrid), and the “advisory speech” (gdams ngag). Oral textual commentary, on the one hand, consists in a series of discussions between a student and a teacher on a particular text, and is

ORAL AND TEXTUAL TRANSMISSION 48 based on a philosophical analysis that replicates, questions, and expands the text.

Just like a written commentary, it may also include: “word commentary” (tshig

’grel), that comments on every word of a text; “meaning commentary” (don ’grel), that expands on a text’s central issues; and “commentary on the difficult points”

(dga’ ’grel), that focuses on the most problematic points of a text (Klein 1994: 2-

6). Advisory speech, on the other hand, includes reflections that are independent of any specific text. Its defining characteristic is supposed to be its effectiveness, as it is able to eliminate any doubt. An important form of advisory speech is “direct speech” (man ngag), also known as “esoteric instruction,” i.e., advice that is derived from a hand-on, direct experience (6-8).

It would seem that in most Buddhist traditions, the early period was characterized by a relatively open oral composition, followed by an oral transmission that took either a relatively fixed or an open form, depending on what was being transmitted and why. Wynne (2004), by looking at the oral transmission of early Buddhist teachings and literature, suggests that the material that was being transmitted, Prātimokṣa (monastic rules) and Buddha-vacana (the words of the

Buddha), which conferred authority and legitimacy to the early Buddhist community, were hardly “the sort of materials suitable for improvisation” (108).

Thus, dharma and vinaya were transmitted verbatim and involved word-for-word repetition, as the early Buddhists were “more concerned with preserving important teachings” than “with taking into consideration the needs of their audience” (114).

ORAL AND TEXTUAL TRANSMISSION 49

That does not seem to have been the case with the three Tibetan traditions

(Sems sde, Ch’an, and Mahāyoga) that were primarily concerned with a practical realization of a sudden enlightenment (instead of preserving a teaching), whether through a tantric ritual or practice, a Ch’an meditation, or a Sems sde spontaneous realization. The “Three Statements” of dGa’ rab rdo rje—(1) “direct introduction,”

(2) “not remaining in doubt,” and (3) “continuing in that state”—exemplify well this point as they are considered to represent the essence of the rDzogs chen teachings (Norbu 2006: 110). The first statement, “direct introduction,” is indispensable, as one cannot attain realization without it, however, each master is free to introduce the nature of the mind as he (or she) finds fit, and depending on his (or her) own realization. The second statement, “not remaining in doubt,” and the third one, “continuing in that state,” are directly related to the student’s personal, direct experience—that the student has the experience, and that he (or she) has integrated that experience in his (or her) own everyday activities.

The role of Buddhist masters has always been that of guides, to lead students through practical instructions. Some of their teachings may have been based on textual sources, as Ch’an masters do quote from sūtras and tantric masters from tantras, however, most of the teachings were probably based on an oral form.

They were part of oral traditions, and their teachings were produced, preserved, and transmitted orally. From the extant texts of the Sems sde, Ch’an, and Mahāyoga traditions, it is evident that in the period when the traditions emerged, the diffuse

ORAL AND TEXTUAL TRANSMISSION 50 body of teachings must have undergone major developments, both in number and in content, until certain parts have reached their fixed, written form. How many of these teachings remained unwritten and thus disappeared once the oral traditions died out, we will never know. Furthermore, how many were later lost in a manuscript form we can only guess from the textual sources that were recovered from Dunhuang. This is not to say that the teachings were all similar, or that there were no differences between them, on the contrary, the mutual interaction and interdependence allowed for a great deal of diversity. Being oral, it also allowed for flexibility within the framework of generally accepted attitudes and concepts that were upheld by each tradition.

Open oral traditions may have existed side by side with, the relatively fixed in character, written traditions. Teachings, by definition, are oral-literary activities, and Sems sde, Chan, and Mahāyoga masters may have used both oral and written materials as a starting point for further instruction and teaching. These masters were not just passing on established traditions, but, while teaching, were also involved in the process of creation and the development of new forms of instructions. In this way, the “new” traditions emerged and developed through a dynamic process that did not depend as much on individual authors, as it did on mutual interaction. The traditions, therefore, may have emerged in the very process of instruction when new forms of practice were adopted, and new concepts, terms, and phrases were created. Once the “building blocks” were created, new

ORAL AND TEXTUAL TRANSMISSION 51 developments would emerge based on these teachings. In the beginning, it is most likely that these blocks—brief statements and set phrases, principles, and precepts—“were not fixed” in content and details, and “circulated freely” (Wynne

2004: 121-22). With the passage of time, these materials may have become available to practitioners in both written and oral form, and, as such, they were the collective property that was taken up, changed, and used according to the needs of each individual.

Most of these “building blocks” got lost, and their authors remain anonymous, unless the blocks were codified in a written form and are still extant, or are cited in later texts. If the blocks did actually circulate in oral traditions, one would expect that some traces of them, and perhaps even the names of their authors, may still be recovered from extant texts. gNubs chen, in the bSam gtan mig sgron, cites from extant dPal dbyangs’s works, and mentions dPal dbyangs as the author of oral teachings. This may give us some idea of whether any oral instruction was codified in a written form, and if any oral teaching can be

“recovered” from extant texts. Thus, in quoting from rDo rje sems dpa’i zhus lan

(twelve times), gNubs chen cites the text by name eight times, and one time cites dPal dbyangs as the author, “the oral instruction of scholar dPal dbyangs” (mkhan po pal dbyangs kyi man ngag). He twice explains teachings which are only identified in the notes as those of dPal dbyangs (these cannot be identified in his works), and twice cites from what he calls “oral instruction” (man ngag) (and these

ORAL AND TEXTUAL TRANSMISSION 52 can be identified in his works) (Takahashi 2009: 261-63). This would suggest that dPal dbyangs’ teachings were circulating in both oral and written form and that while some of his oral instructions were ascribed to him, others were freely circulating as the public domain (anonymous “oral instructions” on specific themes and topics). It further suggests that the “building blocks” could perhaps be located in the passages of the bSam gtan mig sgron where gNubs chen explains the Sems sde, Ch’an, and Mahāyoga teachings.

In his chapter on rDzogs chen (chapter 7), gNubs chen cites and explains nine “principles” of rDzogs chen, that in the interlinear notes are attributed to several rDzogs chen masters (see Esler 2012: 86-103; Karmay 1988: 113-18; and

Ying 2010: 251-80). He also points out that these different approaches and expositions, which are given by different masters, in the end, are all of the same essence (Ying 2010: 252).

The first view takes “non-referentiality” (gza’ gtad dang bral ba) as the main principle and is taught by Vimalamitra and Mahārāja of Oḍḍiyāna (King

Indrabhūti). This view emphasizes the non-conceptual nature of reality and the fact that this nature is free from all conceptions and conceptual fixations. As the self- originated pristine awareness, this nature is spontaneously self-perfected from the very beginning—as such, “Buddhas, sentient beings, saṃsāra, and nirvāṇa are all just designations” (gNubs chen as cited and mentioned by Ying 2010: 253-56).

ORAL AND TEXTUAL TRANSMISSION 53

The second view takes “spontaneous presence” (lhun gyis grub pa) as the main principle and is taught by dGa’ rab rdo rje. This view teaches that the nature of reality is “the state of enlightenment itself.” Metaphorically, it is like “the precious wish-fulfilling jewel” in which “whatever exists appears as whatever one needs and desires”—that being the case, there is no need for effort, urge, or compulsion to realize anything (257-59).

The third view takes “great nature” (bdag nyid chen po) as the main principle and is taught by Vairocana. It teaches that “all phenomena, without exception,” are spontaneously perfected and “naturally clear” to “one’s open awareness, the non-abiding pristine awareness”—as a result, one abandons the fixation on appearances (self and other) and designations (words and letters) (259-

61).

The fourth view takes “self-originated pristine awareness” (rang byung gi ye shes) as the main principle and is taught by dGe slong ma Kun dga’ ma. This view emphasizes the fact that “all phenomena that arise and perish,” from the very beginning, “are in the state of enlightenment,” and that this state is “the essence of the self-originated pristine awareness that is intrinsically free from causes and conditions”—that being the case, what is there to be grasped? Trying to grasp for anything, would be like “the sky trying to grasp for itself” (261-63).

The fifth view takes “effortless non-action” (bya btsal dang bral ba) as the main principle and is taught by Buddhagupta. This view teaches that everything is

ORAL AND TEXTUAL TRANSMISSION 54 already in the state of completeness; there is nothing to search for or to strive for.

As striving and effort will not bring one to awakening, one needs to “abandon the illness of striving” (even that of “trying to abandon” itself), and “stay with the principle without any effort” (263-67).

The sixth view takes “great bliss” (bde ba chen po) as the main principle and is taught by Kukurāja and Śrī Siṃha. This view teaches that having realized that the nature of one’s awakened mind “gives rise to everything,” and that this same mind is “the pure space of great bliss,” one experiences everything as the great bliss—as such, to generate desire and purposefully strive (i.e., for happiness) is to “turn one’s back on this great bliss” (267-70).

The seventh view takes “non-duality” (gnyis su med pa) as the main principle and is taught by Mañjuśrīmitra. This view essentially teaches that as the awakened mind, which “exists from the very beginning,” is blissful and effortless,

“all known extremes are naturally non-dual”—consequently, even striving for awakening is one-sided, for “it prevents the spontaneous presence of one’s open awareness” (270-73).

The eighth view takes “great sphere” (thig le chen po) as the main principle and is taught by Prince Thubo Rājahasti. This view emphasizes the fact that open awareness, that is, the awakened mind, is “naturally free of all conceptual elaborations.” The awareness is the “great sphere,” and the “great sphere” self-

ORAL AND TEXTUAL TRANSMISSION 55 originates—that being the case, it should not be pursued, as it “can never be established or realized through conceptualization and purposeful action” (273-77).

The ninth view takes “the basis of all existence just-as-it-is” (chos thams cad gzhi ji bzhin pa) as the main principle, and is taught by dGa’ rab rdo rje, King

Dhahenatalo, and gNubs chen himself. This view essentially encompasses all rDzogs chen teachings that have been discussed in relation to the “principles.” It points out the fact that there is no such thing as a “concrete,” reified view of a

“spontaneous presence,” “great bliss,” “non-duality,” and so forth—as such:

[one] does not engage in reasoning, does not examine, does not conceptualize, does not actualize, does not investigate, does not measure; he does not [purposefully] do anything whosoever. Rather, in the manner of non-postulation he posits… in non-knowing he knows, in non-acquirement he acquires, in non-achievement he achieves, in non-seeing he sees, in non- realization he realizes, in non-attainment he attains, in non-visualization he visualizes, and in non-experience he experiences. (gNubs chen as cited by Ying 2010: 277-80).

It is important to note that for gNubs chen all these “principles” and designations are synonymous—they refer to the same reality, “suchness” or “just as-it-is.”

Furthermore, even “suchness” is nothing but a label, a name that should not be grasped or reified. Mistaken, one could strive and pursue after empty concepts and designations, without even realizing that there is nothing to grasp; in “pristine

ORAL AND TEXTUAL TRANSMISSION 56 awareness” there is no conceptual reference point to begin with. “Reality” cannot be philosophically understood, it has to be directly experienced.

One earlier text that has already discussed the role of “principles” is Erh-ju ssu-hsing lun (Treatise on the Two Entrances and Four Practices) by Bodhidharma

(d.c. 530), the first patriarch of Chinese Ch’an. The treatise is found in Langqie

Shizi ji (Record of the Masters and Disciples of the Lanka), written by the Jingjue

(c. 688-746), and is preserved in several Dunhuang manuscripts. At the beginning of the text, Bodhidharma points out that even though there is more than one way to attain enlightenment, all of them can be summarized according to the two entrances: “entrance by principle” and “entrance by practice.” Entrance by principle means “relying on the teachings” and “realizing their guiding principle.”

Entrance by practice means “contemplating” (1) karmic consequences and (2) dependent arising, to realize detachment, and (3) emptiness, to realize not seeking; and (4) “acting” in accord with the dharma, that comes about through the realization of the principle (van Schaik 2018: 117-27; see also Broughton 1999;

McRae 1986: 101-17). The main “principle,” for Bodhidharma, is that of “true nature.” It teaches that all sentient beings, whether ordinary or enlightened, are the same in their true nature. Once the principle is realized, one becomes identical with it and no longer follows spoken instructions: “without form, beyond analysis, a stillness without a name” (van Schaik 2018: 19, 24-25).

ORAL AND TEXTUAL TRANSMISSION 57

In the bSam gtan mig sgron, gNub chen is aware of Bodhidharma’s two

“entrances,” and quotes “entrance by principle” two times (57-58, 130). For him, once a practitioner has a direct experience of what the rDzogs chen principles are all about, he (or she) “sees all phenomena with one’s awakened mind, and then acts accordingly.” That is, the practitioner rests in the spontaneous presence of “pristine awareness,” completely free from any grasping and reification, and acts in accord with the rDzogs chen principles (Ying 2010: 2286). For Bodhidharma and gNub chen, acting in accord with a “principle” involves more than engaging in a ritual, a contemplative technique, or a philosophical speculation. They do seem to be in accord when it comes to “principles,” and, yet, gNubs chen specifically points out that he wrote his text precisely because of the similarities (that were causing misunderstanding) between the Sems sde and the Ch’an teachings (Karmay 1988:

105). Was he critical of Ch’an practices because they departed from Bodhidharma’s teachings? Perhaps! Would he be critical of scholars who suggest that Sems sde teachings are similar to Mahāyoga philosophical speculations? We will never know! What is clear though is that even though Sems sde, Tibetan Ch’an, and

Tibetan Mahāyoga have themes and vocabulary in common, they are still three distinct traditions. The similarities could be due to the fact that oral “principles,” as

“building blocks,” had an important role to play in the process of their emergence and development. The traditions were interdependent; they mutually interacted and

ORAL AND TEXTUAL TRANSMISSION 58 shared “building blocks.” With the passage of time, some of these “building blocks” were collated and became written texts.

3.2 Textual transmission

If Davidson’s (2002) assessment of the process of scriptural formation is correct, then oral instructions, taken collectively, represent the potential for the development of the Sems sde texts. Oral teachings, based on “principles” (such as that outlined by gNubs chen in the bSam gtan mig sgron), as well as “experiential instructions”—that derive from a master’s personal experience and are associated with the “principles”—served as the fundamental building blocks from which early

Sems sde texts were formed.

The “original” texts were not created ex nihilo. Instead, their authors (or compilers) worked from a mine of materials, from which they used themes, stock phrases, ideas and terminology. These materials may have been available to them in both written and oral form, as building blocks that could be varied at will according to the needs of the moment, and be modified to suit new experiences. However, the precise relationship between the surviving written texts and the oral instructions that, presumably, lied behind them is not clear, and is a matter for further investigation. It is possible, though, that written texts originally were regarded as nothing more than a permanent form of the teachings already given orally.

ORAL AND TEXTUAL TRANSMISSION 59

Some present-day scholars, such as Wilkinson (2012), still consider the

“fictive” model of scriptural revelation as a possibility. In the case of the Mi nub pa’i rgyal mtshan, one of the five earliest Sems sde texts, he suggests that the original teaching of Vajrasattva, as given to dGa rab rdo rje, “due to turbulent circumstances” may have “scattered into a miscellaneous collection of texts,” only to be “brought together again” at a later time (45). This “conclusion” is based on the fact that ten texts contain verses that are found in the Mi nub pa’i rgyal mtshan, scattered throughout, and the “idea” is that someone like dPal gyi seng ge (b. 8th cent. – d. 9th cent.), or one of his peers, extracted this verses and brought them together to form a whole text again (44-45). The problem with this model—that a scripture was produced by a Buddha (or an individual inspired by a Buddha), got scattered, and was later re-compiled—is that it assumes that most of the texts (if not all) that include verses pre-date the Mi nub pa’i rgyal mtshan, and that probably is not the case. Karmay (1988), suggests that even though most of the “rDzogs chen tantras” are un-datable, in their present form, they “hardly date back beyond the

11th century” (52, n. 45). Furthermore, close textual analysis of specific texts does not seem to support the model of textual revelation.

Liljenberg (2018), in her article “From Treatise to Tantra,” documents how a commentarial text, Byang chub kyi sems sgom pa, attributed to Mañjuśrīmitra (an

Indian scholar), is “literary” transformed into a Tibetan tantric scripture, Byang chub sems sgom pa’i rgyud. This is a “clear-cut case” of the process of

ORAL AND TEXTUAL TRANSMISSION 60 transformation of a treatise into a sacred text equivalent to the Word of the Buddha.

The Byang chub sems sgom pa’i rgyud incorporates the entire Byang chub kyi sems sgom pa text verbatim, with only three lines missing, and greatly expands upon it.

Thus, it is not an actual recension of the Byang chub kyi sems sgom pa, but, rather, a “derivative, but substantially different text” (34). The method of incorporating smaller texts (or parts of texts) into larger ones, and expanding upon them, would seem to suggest that some later texts may be the “reworking” of other, older texts.

Perhaps, the boundaries between texts were not as important to their authors as the task that they were compiled for, that of instruction.

Another interesting case of a textual “reworking” of a short basic text into a more “elaborate” text is that of the rDo la gser zhun and the Thig le drug pa. The

Thig le drug pa is a text that most often than not is listed as one of the “thirteen later” (phyi bcu gsum) texts, however, sometimes it replaces the rDo la gser zhun as one of the “five earlier” (snga lnga)—or “five greater” (che ba lnga)—Sems sde texts. The five texts belong to a group known as Lung chen po bco brgyad

(Eighteen Major Scriptures), Sems rmad du byung ba bco rgyad (Eighteen Marvels of the Mind), and Ma bu bco brgyad (Eighteen Mothers and Children). The thirteen

“later” texts of the group, according to kLong chen pa (1308-1363), are: rTse mo byung rgyal, Nam mkha’i rgyal po, bDe ba ’phrul bkod, rDzogs pa spyi chings,

Byang chub sems tig, bDe ba rab ’byams, Srog gi ’khor lo, Thig le drug pa, rDzogs pa spyi gcod, Yid bzhin nor bu, Kun ’dus rig pa, rJe btsan dam pa, and sGom pa

ORAL AND TEXTUAL TRANSMISSION 61 don grub (see Liljenberg 2012a). gNubs chen, in the bSam gtan mig sgron, may have been the earliest reference to the “Eighteen Texts” as a group when he refers to them as “the eighteen or twenty texts” (Karmay 1988: 97). The numbers of the texts vary, from one list to another, just as much as the titles of the texts; each list suggests a variation, and this leads to considerable confusion as to which text belongs to which group (or to the sub-group).

Of the five basic Sems sde texts (Rig pa’i khu byug, rTsal chen sprug pa,

Khyung chen lding ba, rDo la gser zhun, and Mi nub pa’i rgyal mtshan) the rDo la gser zhun is the most problematic: its position within the group is not stable, and sometimes it is replaced by either the sGom don drug pa or the Thig le drug pa

(both texts are presumed to be lost as independent texts). Thus, Vairo ’dra ’bag chen mo (a biography of Vairocana found in Bairo rgyud ’bum) and dPa’ bo tsug lag’s mKhas pa’i dga’ ston (a fifteenth-century composition) suggest that the Thig le drug pa is one of the five earlier texts, and not the rDo la gser zhun (Liljenberg

2012a: 88-101).

This confusion is understandable if we consider the fact that the text allegedly written by Mañjuśrīmitra is also known as the rDo la gser zhun. The

Byang chub kyi sems sgom pa is a treatise written by a learned Buddhist scholar that employs classical Buddhist philosophy, while its style is logical and based on philosophical arguments. It contrasts the conceptions of the Yogācāra and

Mahāyoga to a rDzogs chen perspective. According to the legendary accounts,

ORAL AND TEXTUAL TRANSMISSION 62

Mañjuśrīmitra (a paṇḍita) wrote the text after he received the rDzogs chen oral teachings from dGa’ rab rdo rje, and attained realization. Apparently, dGa’ rab rdo rje told him that since he was a great scholar, he should compose a book, and thereafter he composed the Byang chub kyi sems sgom pa (Reynolds 1996: 258).

Mañjuśrīmitra’s treatise has only a few emended lines that are in common with the canonical rDo la gser zhun text and is nothing like it; to claim (as some scholars suggest) that the basic text is a loose paraphrase of the treatise is a major understatement (to say the least).

Liljenberg (2012a, 2012b), based on the citations of the Thig le drug pa in the bSam gtan mig sgron (314.3, 347.5, and 452.2), has established that the Thig le drug pa text is still extant as Chapter 3 of the rDzogs pa chen po chos nyid byang chub kyi sems thig le rgya mtsho gnas la ’jug pa’i rgyud (also known as Tb. 124).

This tantra is a ten-chapter text; it includes two introductory chapters, chapter three as its core, and the rest of the chapters as a commentary on chapter three. The commentary seems to be a slightly later elaboration on a text that originally circulated independently of the other chapters (2012a, 139). What is striking is that the newly identified Thig le drug pa text has a lot more in common with the rDo la gser zhun than one would expect, and this, perhaps, would explain why some lists replaced the rDo la gser zhun with the Thig le drug pa. Sog zlog pa bLo gros rgyal mtshan (1552–1624) went as far as to remove the rDo la gser zhun entirely from the group of the Eighteen Texts (49).

ORAL AND TEXTUAL TRANSMISSION 63

Comparison between Derbac (162-165) edition of the rDo la gser zhun and Liljenberg (2012b: 144-146) edition of the Thig le drug pa

The text in the Thig le drug pa that is highlighted as “bold” indicates a variant reading in comparison to the rDo la gser zhun (an insertion, omission, recensional emendation, or a scribal error); and the passages highlighted as “italics” indicate the text that is cited by gNubs chen in the bSam gtan mig sgron.

rDo la gser zhun Thig le drug pa

[1] /bsam ’das brjod du med pa’i byang de nas sangs rgyas kun gyis bstod chub sems/ /ston pa’i sgron mar gyur byas shing/ phyogs rnams dbang pas yon tan rab bsngags pa/ /chos kyi bskur rang gi gsang sngags ston/ snying por gyur pas ’jam dpal gzhon nu zhing khams mtha’ klas mchog gi nyid/ /bya bral lhun rdzogs bde ba’i lhun dkyil ’khor sbyor/ mya ngan rba la gnas/ glong rnam rtog dra ba sel/ kun tu bzang po rigs kyi ’khor lo dag/ phyogs rnams kun tu gzigs shing dgongs mdzad de/ rang gi gsang ba bla na med pa’i tshul/ (fol. 54a) ston pa ’jig rten sgron mar gyur pa kun gyis rab bsngags pa/ chos la chos kyi snying por gyur ba ’jam dpal gzhon nu nyid/ rnam dag byang chub sems kyi rang bzhin gnyis med sgom pa ni/ bya bral bde ba’i klong du ma bcos ji bzhin btang snyom gnas/

[2] /tshul khrims la sogs spyod pa rgya tshul khrims la sogs spyod pa rgya mtsho’i gzhir gyur pa’i/ /ji snyed rnam mtsho gzhir gyur pa/ tha dad ji snyed par grol ba’i lam bstan pa/ /bde gshegs lam du bshad pa yang/ bde gshegs yum yum ’gyur kun gyi mtshungs lam de/ /de du gyur cing kun gyis mtshungs pa’i med mi ’byung de phyir de ni rnam grol lam/ de med mi ’byung de phyir de ni mchog gi lam/ rnal ’byor mchog gi lam/ [314.3]

[3] /phra zhing shes dka’ kun gyi lam ste ’gro zhing shes dka’ kun gyi lam ste mi mi rtog rtog las ’das/ /mi gnas mi dmigs rtog rtog las ’das/ mi gnas mi dmigs spros med bsam pa kun dang bral/ /tshig spros med bsam dang bral/ tshig gis mi gis mi phebs dbyibs dang kha dog spyod theg dbyibs dang kha dog dbang po’i yul med/ /bstan cing brtag par dka’ la spyod yul med/ bstan zhing rtag par brjod du rdul tsam med/ dka’ la brjod du rdul tsam med/

ORAL AND TEXTUAL TRANSMISSION 64

[4] /sngon gyi drang srong lam du gang sngon gyi drang srong rnams kyi lam zhugs pa/ /bsgom chags lam gyi nad du med/ sngon gyi drang srong rnams kyis zin ta re/ /ston pa’i lung bzhin tshig kyi lam du gang ’jug pa/ sgom chags gi mtha’ spyod lam du de mthong na/ rtsol ba lam gyi nad kyis zin ta re/ ston /de ni rtog pa’i rgyun ’brangs smig rgyu pa’i lung bzhin tshig gi mtha’ dpyod lam snyegs pa ’dra/ /yang dag lam ni tshig du de mthong na/ de nyid rtog pa’i gis mtshon du med/ /yang dag nyid du rgyun ’brang ri dvags smig rgyu snyegs bstan pa nga yis ’khrul/ pa ’dra/ brjod med rnam dag lam ni tshig gis mtshon du med/ dag dang ma dag bstan pa tshig nyid tsam ste ’khrul/

[5] /dag dang ma dag gnyis med ’dres dag dang ma dag gnyis med ’dres shing shing mtshungs/ /rnam pa cir yang mi dbyer med mtshungs/ de bas gnyis ’byed ye shes gti mug dang / /thogs med snang spang blang dgag (fol. 54b) gsal ba’i mar me bsam pa kun dang bral/ sgrub med pa’i tshul/ rnam pa cir yang /ngang gis mi g.yo rmugs shing ting mi ’byed ye shes gti mug ngang/ thogs ’dzin rgyal por gnas/ med gsal ba’i mar mer bsam pa kun dang bral/ ngang gi mi g.yo rmugs shing ting ’dzin rgyal por gnas/

[6] /mngon sum mthong bas mthong mngon du mthong med mthong byed de med mngon sum mthong ba’i mig /de nyid thob pa’i dmigs/ de phyir thams phyir thams cad mkhyen pa’i spyan zhes cad mkhyen pa’i spyan zhes de la bya/ de la bya/ /mtha’ dang dbus med yangs de bzhin dbang po drug gi yul gnas de pa’i snying po la/ /mi len mi ’dor nyid ’das/ mtha’ dang dbus med yangs mnyam pa’i rgyal por gnas/ pa’i rang bzhin te/ mi len mi spong mnyam pa’i rgyal por gnas/

[7] /sems dang bag chags gnyis med sems dang bag chags gnyis med ’dres ’dres shing mtshungs/ /’dzin pas brtags shing mtshungs/ ’dzin pas brtags shing shing snang ba’i chos rnams ni/ /rang gi snang ba’i chos rnams ni/ rang gi rgyan rgyan du snang bas mi ’dor spong mi du snang bas mi ’dor spong mi byed/ byed/ /cir yang mi dgongs thabs kyis de rnam par cir yang mi dgongs thabs kyis la rol/ bde la rol/ [452.2] [8] /kun dang mi mthun yongs kyis kun dang mi mthun yongs kyis spong spong ba’i chos/ /nyon mongs lnga dang ba’i chos dag pas/ nyon mongs lnga mtshams med lnga po yi/ /rnam dag lam dang mtshams med lnga yi dngos nyid der zhugs nas mnyam pa’i rgyal po med/ rnam dag lam der zhugs pas thob/ /bud med la sogs kun kyang spong mnyam pa’i rgyal po thob/ ’du byed la mi byed/ sogs kun kyang spang blang dor mi byed/

ORAL AND TEXTUAL TRANSMISSION 65

[9] /lo rgyus don gnyis tshad ma’i blor lo rgyus don nyid tshad ma’i blor bzhags bzhag nas/ /ting ’dzin gsum po brtan nas nas/ ting ’dzin gsum sogs rtags rnams grub pa’i mtha’ spyod pa/ /rtsod bral grub ’dod pas/ rtsol bral lung las gol te lung las gol zhes ’khrul pa yin/ /bya bral ’khrul pa yin/ bya bral lhun rdzogs bde lhun rdzogs bde ba’i lhun la gnas/ ba’i yul la gnas/

[10] /ye shes rang ’byung che ba’i ye shes rang byung che ba’i snying po snying po nyid/ /mi g.yo mi ’gyur nyid/ mi g.yo mi ’gyur bsnyad pa kun bsnyad pa kun dang bral/ /zin pa’i bdud (fol. 55a) dang bral/ zin pa’i bdud rtsis rtsis rtsol ba’i sdug bsngal ’joms/ /bya rtsol ba’i sdug bsngal ’joms/ bya bral bral zin pa’i yul la ji bzhin gnas/ yul la kun tu ji bzhin gnas/ [347.5]

[11] /chos rnams thams cad thig le chen ma bcos mtshan ma’i rtog tshogs po’i rang bzhin yin pas na/ /spros pa mtha’ rnams kun dang bral/ ma ’gags med cing bsdus pa med/ /skye ba med yon tan dus gsum ’grib pa med/ ma cing ’gag pa med/ /’gog pa med par de spangs nyes tshogs rtog pa’i yul bzhin gnas/ /rnam par mi rtog snying po rnams dag/ sgrib med phyi nang kun ’di/ /nam mkha’ bzhin du ye gnas pas/ tu ye shes chen po gnas/ rtsol sgrub /rtog pa’i smra bsam yul las ’das/ mtha’ ’das sangs rgyas kun tu bzang/ ye nas rnam bkod lhun gyis grub pa’i chos/ de dag de rgyud rang gi rnam dag ston/ thig le chen por chos rnams spros pa med/ de dag byang chub sems kyi rang bzhin la/ rang bzhin ngo bo yul kun gsal bar ston/ thams cad byed dang bya ba’i chos rgyud las/ skye ’gag rnam brtags rang bzhin don gyis bral/ yul dang yul can mtshan ma’i tha snyad de/ bsam rtog kun las ’das pa’i ye shes nyid/ ces gsungs so/ rdzogs pa chen po chos nyid byang chub kyi sems thig le rgya mtsho gnas la ’jug pa’i rgyud las/ de kho na nyid kyi don gtan la ’bebs pa’i le’u ste gsum pa’o/

A comparison of the two texts suggests that they are not that different after all. The

Thig le drug pa incorporates most of the rDo la gser zhun text (sections 1 to 10),

ORAL AND TEXTUAL TRANSMISSION 66 and several lines are included verbatim. However, the text also inserts new lines

(sec. 2, 5, and 6), and emends some lines to the point that the reading is altered and cannot be found in the rDo la gser zhun (sec. 2, 4, and 9). All other variant readings can be accounted for in terms of recensional emendation and textual transmission.

The most noticeable difference is at the beginning and the end of the Thig le drug pa, where the text is expended and it reads differently from the rDo la gser zhun.

That being the case, the Thig le drug pa is not a recension of the rDo la gser zhun, but a different text, a more elaborate text that is based on the rDo la gser zhun.

The citations of the Thig le drug pa, from the bSam gtan mig sgron, provide some further clues about the relationship between the two texts. What is interesting about these quotes is that even though they are most definitely from a Thig le drug pa text, they are not citing chapter 3 of the Tb. 124 verbatim. For example, in 10.4 the quote sides with Tib. 124 against rDo la gser zhun, even though the reading is not identical: while rDo la gser zhun reads zin pa’i yul la ji bzhin gnas and Tib. 124 reads yul la kun tu ji bzhin gnas, the quote reads ji bzhin yul la kun tu gnas. This is a common scribal error of order reversal (spoonerism). It also sides with Tib. 124 in 2.3 where it reads yum du gyur cing instead of yum ’gyur, and mtshungs pa’i lam instead of mtshungs lam de, and in 7.4 where it reads rnam pa[r] cir yang instead of cir yang. However, in some instances, the quotes side with rDo la gser zhun instead of siding with Tib. 124. Thus, in 2.2 the quote sides with rDo la gser zhun and reads rnam grol instead of rnam par grol ba’i, but against Tib. 124 that omits

ORAL AND TEXTUAL TRANSMISSION 67 the reading. In the same line, it reads rnam grol mchog, as does rDo la gser zhun, instead of reading rnal ’byor mchog, as does Tib. 124. The quotes also have their own peculiar readings, recensional emendations, such as, rkyen gyis where rDo la gser zhun and Tib. 124 read rgyan du (7.3), and lam du bsnyad pa dang, where rDo la gser zhun reads lam bstan pa and Tib. 124 reads lam du bshad pa yang (2.2). All these variant readings would seem to suggest that gNubs chen had access to an earlier version of the Thig le drug pa, and that the author of the Tib. 124 must have had access not only to this earlier version, which is based on the rDo la gser zhun, but also to a copy of the rDo la gser zhun (to account for the reading that sides with the rDo la gser zhun, and could not be emended without it). The exact relationship between the two texts is not clear, however, there is no doubt that the rDo la gser zhun is a root text, a fundamental text of the Sems sde tradition.

According to the traditional rDzogs chen Sems sde accounts, the Sems sde teachings originated with the supreme Buddha, Samantabhadra (Kun tu zang po), who revealed them to Vajrasattva. Vajrasattva, in turn, transmitted them to a human master, dGa’ rab rdo rje, who, together with his disciple, Mañjuśrīmitra, codified the teachings into a written form. Thus, all Sems sde texts, according to the mythology accounts of textual transmission, are translations from Sanskrit and can trace their “origin” to the mythic Vajrasattva, and the legendary figure, dGa’ rab rdo rje.

ORAL AND TEXTUAL TRANSMISSION 68

Historical and text-critical studies, however, would seem to suggest that these ahistorical and mythological claims are most probably fictive, designed to provide some form of legitimacy to an emerging Tibetan tradition. They are imaginary accounts that have no historical basis as no rDzogs chen Sems sde text has ever been found in Sanskrit, and preliminary research, based on style and syntax, suggests that the texts are not translations from Sanskrit, but, rather, are original Tibetan compositions (see chapter 8; and Liljenberg 2012a: 143-44). The rDzogs chen Sems sde tradition is a Tibetan tradition; a Tibetan “innovation” that was developed by the Tibetans. It co-emerged together with the Tibetan Ch’an and the Tibetan Mahāyoga, based primarily on oral teachings that circulated in Tibet, and a dynamic relationship of mutual interaction and interdependence between the three traditions. The next chapter (chapter 4) will focus on the methods and principles of Textual Criticism, and provide a comprehensive introduction to several methods of stemmatic and phylogenetic analysis.

TEXTUAL CRITICISM 69

CHAPTER 4

Methods and Principles of Textual Criticism

Textual Criticism, sometimes referred to as lower criticism (as opposed to higher criticism), is a branch of textual scholarship that is concerned with the history of the transmission of a text, and its primary goal is to establish the earliest

“attainable” text. Its “business,” as Paul Maas (1958) points out in his classical text,

Textual Criticism, “is to produce a text as close as possible to the original

(constitutio textus)” (1). Methodologically, Textual Criticism is one of several critical research methodologies used in the study of a text as a historical document.

Thus, Historical Criticism seeks to find the particular historical setting (or settings) of texts. Any text may be said to have a history of its own, which includes its time and place of composition, the circumstances in which it was produced or written, its author or authors, how it came to be written, and the audience to which it was addressed (Soulen and Soulen 2001: 79-80). Form Criticism is used to uncover oral traditions that precede the written texts. It investigates diverse forms and structures and seeks to reconstruct the “life setting” of these traditions (Sparks 2007: 111-14).

Source Criticism (sometimes called Literary Criticism) seeks to detect and, in some cases, reconstruct the documentary sources used by various authors in the composition of their works (Kloppenborg 2007: 340-44); whereas Redaction

TEXTUAL CRITICISM 70

Criticism has as its purpose to discern the process of selecting, organizing, interpreting and combining, by a “redactor,” diverse oral and written traditions into a final textual form (Downing 2007: 310-12). Textual Criticism studies the subsequent transmission of the text and has as its aim to transmit and reconstruct the text itself.

Ancient texts in their various forms (as manuscripts, fragments of manuscripts, printed copies, excerpts, paraphrases, quotations, etc.) are either lost or are still extant as single copies, multiple copies, or as manuscripts preserved in multiple editions. Most “autographs” are long lost to us. An autograph is the original edition of a particular work, written or dictated by the author; the earliest copy from which all later copies are descended. The problem with ancient texts (or any “old” text for that matter) is that not only the original documents are lost, but the surviving copies differ from one another. Whenever a manuscript is copied, some mistakes will almost certainly be made even by the scribes who take particular care to omit them. The damage of accidental errors is compounded when ambitious scribes undertake to improve upon the exemplar from which they copy, either to correct perceived errors of spelling and grammar, or to improve style and content. The result is that no two extant texts are exactly the same. A comparison of two or more manuscripts brings to light “variant” readings; textual “corruptions” in the form of lacunae, interpolations, displaced passages, and various scribal errors introduced into a text. In their extreme forms, these corruptions may render

TEXTUAL CRITICISM 71 substantial portions of a text incomprehensible, unreadable. Each form of corruption presents a special challenge for the textual critic, and it must be approached in its own terms. This has led to the development of several methods and principles of Textual Criticism.

4.1 The Classical Method of Textual Criticism

The method of Textual Criticism originated in the Hellenistic Age, at the famed library in Alexandria, where the Greeks had to deal with the “corrupt”

Homeric epics, the Iliad and the Odyssey. The early curators of the library sought to provide “accurate editions” of the Homeric poems. Zenodotus of Ephesus (c.

325–240 BCE.) formulated a comparison of many different manuscripts in order to restore the original text of both the Iliad and the Odyssey. He made several corrections to the Homeric text, eliminating some verses as spurious, marking others as doubtful, transposing the order of some, and introducing new readings previously not present. Another curator of the library, Aristophanes of Byzantium

(c. 257– 180 BCE), produced another edition of the Iliad and the Odyssey. In his edition, he employed a variety of critical symbols to indicate his opinion on the state of the text. His student, Aristarchus of Samothrace (c. 220–144 BCE.), later supplemented these symbols and published two critical editions of the Homeric poems in addition to editing the work of several other Greek authors (Metzger and

Ehrman 2005: 198).

TEXTUAL CRITICISM 72

The Christian scholars who were excommunicated as heretics by the Bishop of Rome, Pope Victor I (served as pope c. 187-98 CE), applied the method of

Textual Criticism developed at Alexandria to the text of the New Testament, so as to ascertain its originality. Theodotus and his followers (late 2nd century CE), it would seem, were the first Christians to have produced a critical recension of the

New Testament (according to Eusebius, c. 263–339 CE, as quoted by Metzger and

Ehrman 2005: 199). Origen of Alexandria and Caesarea (c. 185–254 CE), a contemporary of Theodotus, produced Hexapla—a text-critical study of the entire

Old Testament (in Hebrew as well as several Greek translations). A century later,

St. Jerome (c. 347-420 CE), another Christian textual critic, produced a revised edition of the Latin Gospels based on older Greek manuscripts (Metzger and

Ehrman 2005: 200-201).

The Classical Method of Textual Criticism involves two main processes: recensio and emendatio. Recensio refers to the process of attempting to determine the genealogical relationship of manuscripts and establish an “archetype,” the earliest inferred common ancestor of all surviving manuscripts—insofar as it can be established from the evidence at hand. Emendatio is the attempt to eliminate all errors to be found in the “archetype” (Tanselle 1983). The method of emendation has been known, practiced, and refined since antiquity. The great novelty of the nineteenth-century was the scientific foundation of recensio. It was Karl

Lachmann (1793-1851), a German philologist and critic trained in classical studies,

TEXTUAL CRITICISM 73 who divided textual criticism into recensio and emendatio, and in the process transformed and thus modernized the discipline of Textual Criticism. He enunciated the principle that “the agreement in error” just as well as “the identity of reading” implies “the identity of origin,” and used this principle to demonstrate how by comparing different manuscripts it is possible to draw inferences as to their lost ancestors or archetypes, and consequently create a stemma, that is a genealogical tree. His method was designed to deal with fairly long texts that have come down to us in multiple, complete manuscripts (see Metzger and Ehrman

2005: 207; and Battezzato 2009: 774).

4.1.1 Recensio

The process of recensio begins with collecting information about the text.

In the first stage, the scholar catalogues and dates complete as well as fragmentary manuscripts, partial quotations, and commentaries. In the second, the scholar (1) collates all extant information about the text; (2) notes every omission and addition; and (3) makes a list of all the variant readings found in the extant sources. This step requires a single manuscript to be selected as a collation’s base. This manuscript is either the oldest, the most complete, or one most free of errors. The last step of the recensio process involves assessing the relationship of the manuscript witnesses.

This involves eliminating manuscripts that are direct, late copies of other extant manuscripts, and trying to determine a stemma of the remaining witnesses by

TEXTUAL CRITICISM 74 grouping them into families. In order to accomplish this step, according to

Lachmann, a scholar must look for shared mistakes. If two or more manuscripts share the same (glaring and not easily repeatable) mistakes, they must be related.

Agreement in correct readings must not be used to infer that a group of manuscripts is closely related. This agreement only proves what we already know: that all manuscripts descend from an original text. For Lachmann, “variant” readings are those readings that link extant manuscripts together (Battezzato 2009: 774-75).

Several causes can account for textual variation, and all these causes are classified as being either intentional or unintentional. “Intentional alterations” occur when a scribe makes the reading more plain by interpolating, deleting, or conflating readings, harmonizes related passages, removes readings needing further explanation, emphasizes or safeguards specific teachings, and/or promotes particular customs. In each of these processes, the scribe may intentionally add, omit, conflate, and/or substitute words or phrases. “Unintentional changes” are accidental slips, variant readings that crop into a text involuntarily during the copying process. These variants are found in every manuscript, as no copyist is infallible from making errors. Some of the most common errors are dittographic and haplographic. Repeating a letter, a syllable, a word or even a sentence is known as dittography; whereas omitting a letter, a syllable or a word that should have been written twice is called haplography.

TEXTUAL CRITICISM 75

In the Tibetan context, there are several kinds of technical, “unintentional” variants that a scholar working with Tibetan texts would be interested in, and some of the most important are grouped according to the following criteria: i)

Orthography and Spelling; ii) Homophones; iii) Mistaken Letters; iv) Punctuation; v) Fusion and Fission; vi) Spoonerisms; vii) Homoeoteleuton (Eye-skip); and viii)

Stylistic.

i. Orthography or Spelling - applies to minor spelling mistakes such as gyi (>m) for gyis (>m=), po’i (.}8m) for pos (.}=), ste (%{) for te (){), bzungs (07v$=) for gzungs (#7v$=), and mchos (1&}=) for mchod (1&}+), where the intended word is clear, but the spelling is incorrect. Such variants are not in themselves necessarily compelling in assessing the ancestry of a manuscript, unless they are shared on a large scale.

ii. Homophones - relate to scribal errors arising over words having the same pronunciation as others but differing in spelling: sku (ku !q) for rku (ku Fy); ’gru (dru 8Es ) and ’bru (dru 8Ks) for ’dru (dru 8Hs); bzhin (zhin 06m,) for sbyin (jin ^m,); byas (jé A=) for bcas (ché 0%=); chung (chung &u$) and phyung (chung @p$) for byung (jung Ap$); ser (ser ={:) for gzer (zer #7{:); khro (tro D}) and spros (trö c}=)

TEXTUAL CRITICISM 76

for ’phros (trö 8J}=); phrag (trak J#) for khrag (trak D#); byed de (jé dé A{+-+{) for byas te (jé té A=-){) (Harrison 1992).

iii. Mistaken Letters - are visual errors made by scribes due to confusion of look alike letters. In Tibetan some of the most common mistakes include: ngang ($$) for dang (+$), and tran (F,) for dran (H,); langs (;$=) for yangs (9$=); dba’ (+08) for dpa’ (+.8), and bdud (0`o+) for bdung (0`o$); mtshe (13|) for mche (1&{); lha’i ([8m) for lnga’i (T8m), zung (7v$) for zur (7v:), and sogs (=}#=) for logs (;}#=); stag (%#) for rtags (K#=), and rgyun (Wv,) for rgyud (Wv+); sgoms ("}1) for skoms (!}1=); brtag (0K#) for brnag (0M#), lta (W) for lha ([), and ya la (9-;) for a la (?-;). The identification and/or positioning of the vowel may also be mistaken, as med (1{+) for mdo (1+}), dogs (+}#=) for dgos (+#}=), and dbyings (+Am$=) for dbyibs (+Am0=) (Cantwell and Mayer 2006: 65).

iv. Punctuation - in Tibetan context is a valuable variant reading for determining connections between editions of a text and establishing the stemma of that text. The placement of a single (k) or a double shad (kk, for example, #7v#=- %}$-.8}kk), most often than not, has an impact on the meaning and coherence of the text and tells us much about how a particular text was read (Silk 1994: 70).

TEXTUAL CRITICISM 77

v. Fusion and Fission - refer to scribal mistake of joining two words into one and dividing a word into two. It can occur when a tsheg ( - ), the mark of inter-syllabic punctuation, is omitted or misplaced as in the following examples: nyid ((m+) for nyi da ((m-+); and ba rgya (0-W) for brgya (0W).

vi. Spoonerisms - are common scribal errors of order reversals, such as giving lta stag (W-%#) for sta ltag (%-W#) or byang nub (A$-ao0) for nub byang (ao0- A$). However, with some of this type of errors it is not always easy to tell which one was the original reading when both variants make sense, eg., yang snang (9$- '$) for snang yang ('$-9$). These variants can be either recensional or transmissional.

vii. Homoeoteleuton (Eye-skip) - involves the omission of words or lines between repeated words or phrases when the eye moves from one occurrence to another further down the page, omitting the text in between. In the Tibetan context, it can prove one of the most reliable sources of “indicative” errors. As it is generally not possible to recover from omissions of several lines, eye-skip enables us to see where one brunch of the tradition has corrupted the earlier text, and hence where all versions sharing the error must have had a common ancestor in which the error originated (Cantwell and Mayer 2006: 62).

TEXTUAL CRITICISM 78

viii. Stylistic - are variants that cover the writing style, which may or may not have been inherited. The category primarily includes alternative spellings for the same word, such as writing phrin (Jm,) for ’phrin (8Jm,); ’thun (8*v,) for mthun (1*v,); cig (%m#) for zig (7m#); du (`o) for tu (_p); ci (%m) for ji ('m); or yi (9m) for i (?m).

The consistent use of a particular spelling throughout a text or a collection of texts, even though unconventional and therefore less standard, suggests a deliberate alternative style rather than scribal errors, and may be a stylistic feature that is shared between closely related versions of a text. This category, in and by itself, is not indicative of a straightforward textual affiliation, that two or more versions necessarily used the same exemplar, as the style of writing could have been the style of a specific period or specific geographic area (Cantwell and Mayer 2006:

60).

The distinction between the intentional and the unintentional textual variants is not always clear. In particular, certain spellings could be deliberate or the result of negligence or distraction. Some scholars (Harrison 1992: xxv) also make a distinction between “recensional” and “transmissional” variants.

Recensional variants reveal either extensive or deliberate editorial changes to the text, or adoption of a different text altogether; whereas transmissional variants point to errors resulting from scribal lapses or casual attempts to improve or

TEXTUAL CRITICISM 79 modernize the text. Recording all variants, whether recensional or transmissional, guarantees a relatively full picture of the textual tradition, as each text has its own history with variant readings originating in each of its several phases of transmission. Textual critics assume that a copy will perpetuate the readings of the version from which it is copied, in addition to introducing new variants: this assumption is used when copies are put in chronological sequence. Once all the data has been collated, the textual critic begins his (or her) inquiry into the historical relationship of extant manuscripts, by attempting to answer questions such as: Are there any cases where the variant readings are present in two or more manuscripts? Do the readings collate consistently? Do they collate with a particular group of manuscripts? Can connections be drawn between the manuscripts? Are any of the manuscripts directly descended from other extant copies? Is it possible to recognize groups of closely related copies, or construct a stemma (West 1973: 30)?

According to Martin West (1973: 32-33), it is possible to deduce the relationship between the manuscripts from the pattern of agreements and disagreements among them. The argument could run as follows: Suppose there are six extant manuscripts, ABCDEF (see Fig. 1, [a] [b] [c] denoting lost, hypothetical copies).

TEXTUAL CRITICISM 80

[a] │ ------│ │ A [b] │ ------│ │ │ B C [c] │ │ F ------│ │ D E

Fig. 1

All manuscripts preserve readings that are common to all, but each one also has singular readings that are not reproduced in the rest, therefore they all descend from the same copy, archetype [a], a copy that is now lost. There are further variant readings that are common to BCDEF but cannot be found in A, and variants that are peculiar to each manuscript and not shared by the whole group, therefore they have their own copy from which they descend, hyparchetype [b]. As A has variants of its own, [b] cannot derive from A but derives independently from [a]. An archetype [a] gave rise therefore to both extant manuscripts, A and the lost manuscript [b]. BF share some further variants not found in CDE, but there are no variant readings in B that cannot be found in F, while F has some variants peculiar to itself. Therefore F is directly descended from B. DE share variants not found in

BCF, but each one has also singular readings not found in the other, so neither is

TEXTUAL CRITICISM 81 derived from the other but both are dependent upon another copy, hyparchetype [c] that now is lost. Of the three copies of [b], namely B, C, and [c] as reconstructed from DE, no two have a variant reading where the third has a reading that could not have gotten it by conjecture. No two of them, therefore, are dependent upon one another or on a further copy: all three come independently from [b].

As the lost manuscript [c] can be reconstructed from DE, and [b] from BC and [c], the lost archetype [a] can also be reconstructed from A and [b]. The role of manuscript F, which is a direct copy of manuscript B, is considered as insignificant in the stemmatic process (worthless, according to Maas, 1958: 2) and as such needs to be eliminated, excluded from the stemma and the last text-critical process, emendatio. It would seem that this process of selection and elimination is one of the several shortcomings that plague the classical Lachmannian method. If the manuscript F contains a good selection of singular variant readings, a case could be made that perhaps it is a “mixed” (conflated) manuscript with two or more ancestors, most of them lost, belonging to an extinct textual tradition. By eliminating the manuscript, an important line of transmission, that could, hypothetically speaking, trace its origin directly to the archetype (or even to the autograph itself), may get lost. The possibility of “conflation,” of horizontal transmission, has to be admitted, and this possibility makes the reconstruction problematic and questionable as most classical stemmatic analyses are based on the assumption that conflation has never taken place.

TEXTUAL CRITICISM 82

Furthermore, the greater the number of manuscripts the less likely it is that the analysis will provide a satisfactory account of all interrelationships. Some scholars—such as Robert Mayer (2006) in his article “Textual Criticism of the rNying ma rgyud ’bum Tradition”—have suggested that one way of dealing with conflation is to distinguish between historical and pragmatic stemmatic analyses in the problematic cases, and aim for the pragmatic instead of the historical stemmata.

A historical stemmatic analysis, that is, an analysis that seeks to establish a genealogical tree that represents the proven and exact historical relationships of the texts, may not be possible; however, a pragmatic stemmatic analysis that seeks to demonstrate the relationships of extant witnesses according to a clear pattern of shared evidence, with one or more alternative stemmata, is not only possible but necessary.

4.1.2 Emendatio

After recensio, the text-critical scholar must proceed with the emendatio—a process that is closely related to examinatio as both activities are being performed simultaneously. In this process, the scholar assesses between the “qualities” of the transmitted readings, selects among them, and on a case-by-case basis emends them by “conjecture,” as needed. The stemmatic analysis performed during the recensio process helps to establish the reading of the archetype, that is, the reading not dependent on any of the other variants, but from which all other variant readings

TEXTUAL CRITICISM 83 can be explained. The understanding is that the presumed archetype belongs to the stage of the text that provided the basis for all known textual witnesses.

When the stemmatic analysis is not helpful in the selection of the transmitted readings, the scholar can rely on “canonical principles,” that is textual guides. These are classified according to their criteria as external and internal. The following external criteria are usually accepted as guiding principles: 1) unequal status of textual sources; 2) broad attestation; and 3) age of textual witnesses. The first principle is based on the assumption that some textual sources carry more weight than others. In this context, witnesses that contain the most variant readings are often regarded as inferior in their readings. The second principle assumes that the better attested a reading is, the more trustworthy it is. A variant form represented in a number of different textual traditions is to be preferred to a variant found in only one tradition or one geographical area. The third principle presumes that older witnesses are preferable to the more recent ones because the older ones are likely to have been less exposed to textual corruption than the recent ones. The external criteria may allow a scholar to define some characteristics relevant to specific groups of witnesses and thus form an idea of the quality of their readings; however, all three principles have been recognized as problematic. In his seminal work, Textual Criticism of the Hebrew Bible, Emanuel Tov (1992) points out that almost all arguments based on these principles are questionable and of limited usefulness. For example, reliance on a broad attestation of textual evidence is

TEXTUAL CRITICISM 84 unreliable as it could have been created by historical coincidence; according to

Tov, textual criticism “does not proceed according to democratic rules” (301).

The internal criteria consist of evaluating the intrinsic value and content of the variant readings according to the text and its contexts. The following are some of the best-known principles recognized as internal: lectio difficilior, lectio brevior and usus scribendi. Lectio difficilior, often expressed as difficilior lectio potior, refers to the fact that “the harder (more difficult) reading is to be preferred.” It would seem that this canonical textual guide had been proposed and was used for the first time by Johann Albrecht Bengel, in a 1725 “Prodromus” to his edition of the Greek New Testament, Gnomon Novi Testamenti, published in 1734 (Epp

1993b: 146). The principle is based on the assumption that scribes were generally inclined to make texts simpler rather than harder: a rare word or an unusual construction is more likely to be original than a common word or a usual phrase.

Therefore, when textual variation is encountered by a textual critic, one of the readings is termed more difficult than the other(s) and given a preferable status.

The problem with this rule is that it fails to take into account simple scribal errors that may render a text more difficult than it originally was. In these cases the rule does not apply. Furthermore, in many instances the rule can be applied subjectively, as the same reading may appear difficult to one scholar but not necessarily as difficult to another (Tov 1992: 304).

TEXTUAL CRITICISM 85

Lectio brevior, or brevis potior, refers to the phenomenon that “the shorter reading is to be preferred.” The logic this principle is based on is that the usual tendency of scribes, especially when they were looking for a chance to make some kind of revision, was to add details rather than omit them, that is, to expand the text. Scribes would add explanations to clarify the difficult readings or add details to help a reading be in harmony with parallel passages. In short, the assumption is that their aim was to make the text more understandable. The problem with this principle is that there are too many exceptions, as the rule does not cover scribal omissions, such as haplography and homoioteleuthon. Furthermore, some short readings could be entirely intentional, due to scribal attempts to “smooth out” difficult readings by deleting the difficult words. Finally, some short readings may have their existence due to the inherent features of one particular textual tradition

(Vaganay 1991: 80). Usus scribendi refers to the particular “writing style” of the author. According to this principle, the text-critic must select the variant reading that fits with the general stylistic tendencies of the author; these are elements that make up the writer’s usual practice: his vocabulary, language, style, and the way of composing a text (Battezzato 2009: 775). However, since scribes may well have been perfectly capable of conforming to a peculiar addition, such as a passage, or an author’s style, the rule is impractical and of limited use to a text-critical scholar.

Canonical principles, depending on a case-by-case variant reading, are helpful to some extent in the process of evaluating the readings. Their

TEXTUAL CRITICISM 86 shortcomings, however, such as the logic that underlines some rules (lectio difficilior and lectio brevior) and their subjective nature, render them ineffective and applicable only to a small fraction of the readings that need to be evaluated.

Difficult readings (“nonsense” readings) may require an alternative method to a mechanical application of textual rules, and one such a method is the method of

“conjectural emendation.” This is an attempt to “imagine” what the author actually wrote based on the sense, language, style, and metre of both the immediate context, as well as the multiple contexts of the work as a whole. In this process, from the transmitted readings the “reading which is the most appropriate to its context” is selected as the best reading (Tov 1992: 309).

When the text-critical scholar has completed the entire process (that of recensio and emendatio), he or she has arrived at the final step in the Textual

Criticism, that of preparing either a “diplomatic” or a “critical” edition of the text.

The aim of producing a diplomatic (facsimile) edition is the exact reproduction of a particular manuscript; whereas that of a critical edition is the reconstruction, as far as the extant evidence permits, of the autograph or the archetype, that may or may not be the same as the original text (Tanselle 1983: 27, 48-9). Both editions usually include an apparatus criticus that lists most, if not all, of the variant readings. The apparatus of a critical edition, however, is much more elaborate, as it needs to provide all the information needed to evaluate textual decisions made by the editor.

The edition must present the evidence that allowed the editor to arrive at the printed

TEXTUAL CRITICISM 87 text. It must report the manuscript support for each word of the text and list all cases where a conjectural emendation was made. In short, the apparatus must inform its readers about all “manuscript readings that might point to a different choice of text from the one made be the editor” (see Battezzato 2009: 777; and

Tanselle 1972).

4.2 The Claremont Profile Method

The Claremont Profile Method (CPM) was developed by Paul McReynolds and Fredrik Wisse (students of Colwell at Claremont Graduate School) in the 1960s for the International Greek New Testament Project (IGNTP). The IGNTP was at the time preparing a critical apparatus for the Gospel of Luke, and needed a method to determine which manuscripts, from among nearly 1,400, should be included.

McReynolds and Wisse were assigned the task of classifying the manuscripts, as a full collation, a detailed comparison of all the manuscripts, was impractical. They noticed that witnesses shared readings in certain combinations. When a section of the text (a chapter) was selected as a sample base, and a group of manuscripts were collated against this sample, the variant readings in terms of agreements and disagreements revealed a pattern, a “profile” that could be presented in a graphic form. Manuscripts with similar profiles were classified into group and subgroup affiliations. Once the pattern was established, new manuscripts were collated only in select readings, instead of being collated in full, and were subsequently assigned

TEXTUAL CRITICISM 88 to a group or subgroup based on their profile. The CPM proved valuable, as it allowed a complete classification of the extant, closely related manuscripts.

However, the method has its limitations (it takes into account only partial evidence and is not applicable to “mixed” manuscripts) and cannot serve as a substitute for a full textual analysis (see Epp 1993a: 211-220; and Metzger and Ehrman 2005: 236-

37).

4.3 The Coherence-Based Genealogical Method

The Coherence-Based Genealogical Method (CBGM) is a new method for assessing textual variation; it was developed by Gerd Mink, for the Institute for

New Testament Text Formation Research (INTF). The method is a computer-based program that charts the relation of a given textual variant to other variants in the same variant unit. It is not an automated method of generating a stemma or deriving the textual history of a text, but rather a technique, a way of analyzing the structure of a manuscript tradition. That is, “the method” is not tied to a preconceived method. In its application, one might use either internal or external criteria. The

CBGM was developed to evaluate witnesses in a more rigorous and precise manner, and thereby attempt to account for contamination with the coincidental emergence of variants (Mink 2004: 24, 41).

A stemmatological model in CBGM is applied at each unit of variation; the stemma (known as substemma) presents not the relationship between or among

TEXTUAL CRITICISM 89 manuscripts, but the relationship between the states of the text that manuscripts convey at a single point of variation. Each local stemma indicates the “textual flow” from one manuscript (known as a witness) to another. For example, a reading present in manuscript A (and perhaps others) is the source of reading in manuscript

B (and perhaps others); this is expressed as B←A. Here, the textual flow is determined by the variants that the descendant shares with the ancestor, as well as by the variants of the ancestor from which new variants in the descendant evolved.

They determine the direction of the textual flow, which corresponds to the development of the text (i.e. the variants) throughout its history. Once a number of local stemmata have been made, both the consistency of the textual flow between and/or among individual witnesses and the overall textual flow logically required by the stemmata can be analyzed (see Mink 2002; Parker 2008: 169-170; and Mink

2004: 33, 37).

This new method for assessing textual variation, CBGM, may prove useful to a textual-critic in one’s attempts to evaluate witnesses in a more rigorous and precise manner, by accounting for contamination and the coincidental emergence of variants. However, the method also has its inherent problems. Not all variant readings or witnesses may be taken into consideration. A variant reading, according to CBGM, is either designated as genealogically dependent on another variant, and thus relevant, or its origin is designated as uncertain, and the variant is excluded from further consideration (Mink 2004: 35). The CBGM also excludes from

TEXTUAL CRITICISM 90 consideration any non-extant intermediary witness (hyparchetype), whether it is known or not. Furthermore, the CBGM’s reliance on an “initial text” as the base and potential ancestor of all other witnesses, with all variant readings collated against and classified as later or earlier based on their coherence with readings found in this text, seems to have been structured in such a way as to prove or lend legitimacy to the claim that the text used by the INTF (the Nestle-Aland 27th ed.), an arbitrarily selected text, is the actual “initial text,” the ancestor of all other extant witnesses.

4.4 The Phylogenetic Method

In the past few years, Textual Criticism and biology have collaborated by sharing a tool, the computer-based phylogenetic methods that can be applied to the stemmatic analysis of manuscript relationships. Initially developed to establish relationships between and among living organisms, these methods can also help to reconstruct the history of a text, that is, its textual transmissions, by indicating for each variant reading the source text from which it was copied or its hypothetical ancestor. There are many similarities between Textual Criticism and biology. Thus, unintentional changes and intentional alterations were made during each copying process, accumulating and being propagated in the descendant copies of copies, copies of copies of copies, and so on. This process resembles the biological process in which genetic mutations are introduced and propagated during the evolution of a

TEXTUAL CRITICISM 91 species. Connecting each manuscript version to its ancestor gives a tree-like structure, a stemma, similar to the phylogenetic tree in biology. The node (in biology) represents a manuscript in stemmatology, and an edge represents the relationship of a source manuscript and its direct copy (Zou 2010: 41).

A scribe using more than one exemplar, while copying a text, creates a hybrid, “contaminated” text, which may contain sections from each exemplar used or incorporate into the original text individual readings from two or more copies.

This has parallels with genetic recombination, where related molecules produce a hybrid, a lateral gene transfer between distantly related organisms. The stemmatic analysis aims to recover all relationships of a set of manuscripts; this process is analogous to phylogenetic analysis for reconstructing the relationships between and among species (Windram et al. 2005: 189-190). However, there are significant differences between textual criticism and biology, as well. In a phylogenetic analysis, the family tree is usually assumed to be binary. This is because the separation of more than two species is unlikely to happen at one and the same point in time. As a result, most programs for reconstructing phylogenetic trees are based on the model of reconstructing a binary tree. On the other hand, in stemmatology, more than two copies can be derived from the same source manuscript, thus resulting in a polytomy in a stemmatic analysis. In addition, species evolve all the time, and it is assumed that no ancestral form can be retrieved from the analysis.

Such does not seem to be the case with manuscripts—once produced, they “tend”

TEXTUAL CRITICISM 92 to stay unchanged, and the recovery of a hypothetical ancestral manuscript is plausible (Zou 2010: 40-41).

In spite of the differences, similarities between “textual transmission” and

“biological inheritance” have led scholars to undertake several studies in which phylogenetic methods and techniques developed in evolutionary biology have been successfully applied to textual traditions (Lantin et al. 2004; Windram et al. 2005;

Windram et al. 2008; Yorav et al. 2005). There are several phylogenetic methods and computer programs available to the textual-critic. Thus, Neighbor-Joining,

Maximum Parsimony, and Maximum Likelihood are methods implemented in computer programs such as PAUP (Swofford 2002) and PHYLIP (Felsenstein

1989), whereas NeighborNet is implemented in SplitsTree (Huson 1998). Each phylogenetic method is a computerized algorithm that helps create a chain

(manuscripts linked together) and/or a stemma, a graphical portrayal of relationships between and among variant readings or manuscripts. They all have their strengths and weaknesses and the debate centers on the merits of each of these competing methodologies.

The Neighbor-Joining (NJ) method is based on the distances between sequences. The distance between two sequences measures the differences between them. This method reconstructs the tree by repeating the process of joining the nearest nodes until it reaches the root. The NJ method is a popular and extremely rapid estimator of phylogenetic trees; it performs well when the data set is small. Its

TEXTUAL CRITICISM 93 potentially serious weakness is that the observed differences between and among sequences are not accurate reflections of the evolutionary distances between and among them, and when the goal is to infer older relationships, it may not provide reliable values (Holder and Lewis 2003: 276-77).

The Maximum Parsimony (MP) method operates by selecting the tree or trees that minimize the number of evolutionary steps. It considers all possible bifurcating trees, and identifies the one that requires the smallest number of changes (i.e. is most parsimonious). The various manuscripts are in this way grouped according to their shared derived characters. Not all sites are informative for maximum parsimony analysis. A parsimony informative site is one in which there is more than one variant recorded, and the variants are each recorded at least twice. All other sites are considered parsimony uninformative and are subsequently ignored by the analysis. The MP method is based on the assumption that the most likely tree is the one that requires the fewest number of changes to explain the data; its basic premise is that objects sharing a common characteristic do so because they inherited that characteristic from a common ancestor. MP is one of the most successful and reliable phylogenetic methodologies (Lantin et al. 2004; Yorav et al.

2005).

The Maximum Likelihood (ML) method uses the heuristic search to reconstruct the tree with the highest probability or likelihood. Manuscripts are added successively to the tree topology until all are added. Additionally, before the

TEXTUAL CRITICISM 94 next manuscript is added, local rearrangements of the existing topology are performed and accepted only if the rearrangement increases the probability of the tree. The process is completed when all the manuscripts are added and no improvement can be made by local rearrangements. ML is a reliable method, but it is CPU intensive and very slow (Holder and Lewis 2003: 279; Zou 2010).

The NeighborNet method is a distance-based method for constructing phylogenetic networks based on the Neighbor-Joining algorithm. It represents the relationships between and among manuscripts as a network rather than a simple branching tree, and enables multiple affiliations to be represented in a graphical form. The distance between pairs of manuscripts is calculated as the total number of differences between them. The closest pairs of manuscripts, and pairs of pairs, etc., are then grouped, and when a member of one has been chosen as a member of a second, then both are replaced by a pair of linked hypothetical ancestors, so as to give an overall network showing the relationships between and among the various manuscripts (Windram et al. 2008). The NeighborNet is a consistent method that may turn out a useful tool for a textual-critic.

Overall, the results of the computer-based and the traditional methods have been found to be in agreement with each other. All methods, it would seem, have their own shortcomings. As some scholars have pointed out (Yorav et al. 2005), the main disadvantage of the traditional methods, based almost exclusively on the insights, know-how, and intuition of experienced scholars, is the lack of a

TEXTUAL CRITICISM 95 methodological means for resolving conflicting conclusions reached by different scholars. Additional problems arise when the amount of data to be considered is too large to be dealt with manually, or when the evidence precludes a simple description of the ancestry of various versions. Manual resolution is only possible when the number of text witnesses is small, but becomes increasingly difficult as these proliferate. The computer-based methods can quickly assimilate a large amount of data, and therein generate a tree showing the relationships between and among the texts. The methods are rapid and flexible, allowing a variety of analyses to be performed from one data set; they also allow for specific textual variants or manuscripts to be included or excluded from the analyses. Once the initial coding of texts into a form that can be handled by the computer programs has been completed, phylogenetic methods can be very useful for the rapid identification of possible regions of interest. They can then be further examined by more traditional methods.

One of the major shortcomings of all computer-based methods is that they all provide “un-rooted” trees in the form of chains. They look at data in a purely numerical way arriving at no textual judgment. However, “rooting a tree” means choosing a point on the un-rooted genealogical tree as the earliest time in the transmissional history of the manuscripts. It requires either some information that may be based on textual analysis of individual variants, or external evidence, such as the age of the manuscripts or the chronology of variant readings. Unlike a rooted

TEXTUAL CRITICISM 96 tree that shows a sequence of events, an un-rooted tree only represents groupings between and among the manuscripts. Furthermore, most phylogenetic methods were not designed to deal with mixed, contaminated manuscripts. The conclusion that can be drawn is that these methods, in conjunction with other text-critical methods, can work very effectively to facilitate an understanding of a textual tradition. Perhaps their real value is derived from their power to assist the textual scholar, rather than to replace his or her expertise. The next chapter (chapter 5) will provide a phylogenetic analysis—a practical application of two phylogenetic methods, Maximum Parsimony and NeighborNet—and in the process answer some questions related to the antiquity of the rDzogs chen Sems sde texts.

PHYLOGENETIC ANALYSIS 97

CHAPTER 5

Phylogenetic Analysis of the bSam gtan mig sgron’s Sems sde Quotations

The bSam gtan mig sgron text, by gNubs chen Sangs rgyas ye shes

(9th/10th century), is one of the earliest accounts of the rDzogs chen Sems sde tradition. It contains a series of quotations from Sems sde texts, and, yet, despite its importance, the text has not been included in any major Tibetan collection and it was virtually unknown to the scholarly community until it was published in 1974— there is no known commentary on it, or a text that quotes extensively from it. Based on the fact that Tibetan Buddhism has a long- and well-established tradition of

“rediscovered” (gter ma) texts, could it be that the original, gNubs chen’s text, was lost, and the extant copy is a gter ma text, a text that was rediscovered sometimes in the eighteenth, nineteenth, or early twentieth century? And if that was the case, what readings would such a text preserve? It would seem that a phylogenetic analysis could shed some light on these questions.

In this study, I provide a phylogenetic analysis of the bSam gtan mig sgron’s rDzogs chen Sems sde quotations, and the quoted body of text as found in several editions and canonical collections. By doing so, I intend to establish whether the bSam gtan mig sgron text is indeed an ancient text that preserves ancient readings, or perhaps a more recent composition, a rediscovered text. By

PHYLOGENETIC ANALYSIS 98 implication, I also intend to verify whether the claim made in the Gang ga’i chu rgyun by the Fifth Dalai Lama, Ngag dbang bLo bzang rgya mtsho (1617-1682), is credible or not. In this text, he claims that the tenth-century text, the Kun byed rgyal po, is not composed on the basis of the so-called “canonical” Sems sde texts, but, rather, that these texts are extractions from it (Karmay 1988: 207). My assumption is that an objective phylogenetic analysis, undertaken by the use of several computer programs (such as PAUP and SplitsTree) that rely on methods specifically designed to establish phylogenetic relationships, could shed some light on these questions (on phylogenetic applications see Apple 2014; Marwick 2012; and Maas 2010, 2008).

The second aim of my study is to establish whether the extant Sems sde texts are “the same” texts that the author of the bSam gtan mig sgron, gNubs chen, had access to in the tenth century. Professor Neumaier (1998), in her paper entitled

“The bSam gtan mig sgron and its rDzogs chen Quotations: A study in the

Production of Tibetan Texts,” suggests that that may not be the case, as the textual bodies attached to individual titles “may have undergone significant change in the course of history” (2). My assumption, in this case, is that the process of collating bSam gtan mig sgron’s quotations against the parallel textual material as found in other extant witnesses will clarify this particular question.

PHYLOGENETIC ANALYSIS 99

5.1 The bSam gtan mig sgron text

The bSam gtan mig sgron (SM) is considered by most present-day scholars as one of the most important works that has come to light in the twentieth century.

According to the traditional Tibetan accounts, the text was written by gNubs chen

Sangs rgyas ye shes, who flourished at the end of the ninth and the beginning of the tenth century of the Common Era, and is known under various titles, rDzogs chen gyi man ngag bsam gtan mig sgron, rNal ’byor mig gi bsam gtan, and sGom gyi gnad gsal bar phye ba bsam gtan mig sgron. The only extant copy of the text is that published in 1974. As not much is known about its past or its passage through time, it could be said that the text was “revealed” to the scholarly community after more than a thousand years of mostly undocumented transmission. It has not been included in any major Tibetan collection, such as the bKa’ ’gyur, bsTan ’gyur, rNying ma rgyud ’bum or Bai ro rgyud ’bum, and there is no known commentary on it.

In 1092, Pho brang Zhi ba ’od issued a bka’ shog (open letter) to the

Buddhists of Tibet, more specifically to bKa’ gdams pa practitioners. In this document, he criticizes several works assumed to have been composed by Tibetans that were being passed on as originals, of Indian origin (at this period of time only texts composed in India were considered original). The rDzogs chen doctrine is singled out as one that should not be practiced under any circumstance, as this practice, apparently, “leads people into evil birth” (Karmay 1980: 17). Karmay

PHYLOGENETIC ANALYSIS 100 suggests that one of the texts that is mentioned in the bka’ shog, rNal ’byor mig gi sgron ma, may be our text bSam gtan mig sgron (24, n. 93). Considering the fact that gNubs chen is mentioned as one of the authors of rDzogs chen texts (the other being Drang nga Shag tshul, a peripheral figure in Tibetan historiography), and the composer of Rig pa’i nyi ma, a commentary on rDzogs pa chen po nam mkha’ che, this suggestion does seem credible.

Obscurity of the text has led some Western scholars, such as Capriles

(2014), to claim that perhaps the bSam gtan mig sgron was one of the texts that was

“entombed at Dun-huang from tenth or eleventh century CE until 1908” when it was re-discovered (10-11). A great collection of ancient manuscripts was indeed discovered at the beginning of the twentieth century in China; but the bSam gtan mig sgron was not one of these manuscripts. The “Preface” to the text states that this edition, 1974, is a copy of a block-print from eastern Tibet, commissioned by

’Jam dbyangs bLo gros rgya mtsho. The blocks were prepared on the basis of a manuscript that belonged to Lo chen Dharma shri (1654-1718), and this manuscript was based on a copy from the library of the famous Tibetan Buddhist scholar

Tāranātha (1575-1634).

What is interesting is that ’Jam dbyangs bLo gros rgya mtsho (1893-1959), also known as ’Jam dbyangs Chos kyi blo gros, Chos kyi bLo gros, and bLo gros rGya mtsho, was recognized as an incarnation of ’Jam dbyangs mKhyen brtse’i dbang po (1820-1892) by ’Jam mgon Kong sprul (1813-1899/1900). Both mKhyen

PHYLOGENETIC ANALYSIS 101 brtse and Kong sprul, collaborated closely with one of the most famous nineteenth- century Treasure revealers (gter ston), mChog gyur gLing pa (1829-1870). mKhyen brtse (who himself was a great gter ston) instructed mChog gyur gLing pa on rDzogs chen, and collaborated with him on the revelation of mChog gyur’s most famous Treasure, a rDzogs chen text known as rDzogs chen sde gsum (Three

Sections of Great Perfection) (Doctor 2005: 88-89).

mChog gyur also revealed teachings that had originally been taught by gNubs chen Sangs rgyas ye shes to gNubs Khu lung pa Yon tan rgya mtsho, his student. He did this by “remembering his past existence” as Khu lung pa (Doctor

2005: 96). This form of Treasure “revelation” was known as “recollected

Treasures” (rjes dran gter), and was one of the four major forms of revelation, the other three being: “earth Treasures” (sa gter ngos), “rediscovered Treasures” (yang gter), and “mind Treasures” (dgongs gter ngos). The purpose of “recollected

Treasures” was to “revive past teachings” that have been lost and allow a “treasure finder” (gter ston) to claim access to works associated with some of the earliest

Tibetan Buddhist scholars and masters (27).

Both mChog gyur and mKhyen brtse practiced this form of revelation extensively, and the question is then: could it be that the incarnation of mKhyen brtse, ’Jam dbyangs bLo gros rgya mtsho, was in possession of the bSam gtan mig sgron’s manuscript because the text was a “recollected Treasure,” a treasure revealed by either mChog gyur or mKhyen brtse, or by both of them as a

PHYLOGENETIC ANALYSIS 102 collaborative project? But, then, if that is the case, bSam gtan mig sgron has to have a very close stemmatic (phylogenetic) relationship with at least one extant edition that includes Sems sde texts, and this edition would be the one that the quotations are based on. Either way, it is striking that the text may have been extant and known (at least by some scholars), and yet no scholar seems to have ever relied on it, or describe its content before it was published in 1974.

5.2 Methodology

There are several phylogenetic methods and computer programs available to a textual-critic, such as Neighbor-Net method used in SplitsTree (Huson 1998); and

Neighbor-Joining, Maximum Parsimony, and Maximum Likelihood methods implemented in PAUP (Swofford 2002), and PHILIP (Felsenstein 1989). Each phylogenetic method is a computerized algorithm that helps create a graphical portrayal of relationships between texts, editions, and collections of texts. The aim of the methods is, therefore, not to identify a single source, but, rather, to establish

“patterns of relationships” that would aid a textual critic in his (or her) textual analysis. They may be a poor substitute for rigorous textual scholarship when it comes to interpretation, but they do produce accurate results that a textual scholar can depend on. As all methods have their own strengths and weaknesses, perhaps, the best way to proceed is to apply multiple methodologies to the same dataset in order to test the results.

PHYLOGENETIC ANALYSIS 103

In this study, I intend to apply two phylogenetic methods: Neighbor-Net and

Maximum Parsimony. The Neighbor-Net method is a distance-based method for constructing phylogenetic networks that are based on the Neighbor-Joining algorithm (distance matrix method). The program is designed to calculate the best network based on the number of differences between witnesses, with the length of the branches in the network being proportional to the number of differences between manuscripts. Substantive agreements, whether in a small or a large number, can all be depicted in the network, and witnesses can be shown to reside in a network of relationships instead of being portrayed as descending from a single source, an ancestor (Howe, Connolly, and Windram 2012: 54-57).

The Maximum Parsimony method operates by selecting the tree, or trees, that minimize the number of evolutionary steps. It considers all possible bifurcating trees and identifies the one that requires the minimum overall number of changes

(i.e. is most parsimonious). The various manuscripts are grouped according to their shared derived characters. Thus, the presence of a given variant reading in two or more witnesses is taken as evidence that those manuscripts were copied from the same ancestor who also carried that reading. However, not all sites are informative for maximum parsimony analysis. A parsimony informative site is one where there is more than one variant reading, and where at least two variants are each recorded at least twice. All other sites are treated as parsimoniously uninformative and are

PHYLOGENETIC ANALYSIS 104 ignored by the analysis. A “parsimonious” tree is un-rooted—it gives no indication of the origin of a textual tradition (Howe, Connolly, and Windram 2012: 54-65).

5.3 Data set

The earliest form of rDzogs chen is represented by the pre-tenth-century texts that fall under the rubric of Sems sde, such as the texts included in the Kun byed rgyal po and a group of texts traditionally known as Sems sde bco brgyad, the

“eighteen Sems-sde texts.” The first five texts of this group are called sNga ’gyur lnga, the “five earlier translations” of Vairocana, and the last thirteen Phyi ’gyur bcu gsum, the “thirteen later translations” of Vimalamitra. The texts are considered to have been translated from Sanskrit into Tibetan, even though no Sanskrit originals have ever been found, and the texts are most likely to be Tibetan compositions. The bSam gtan mig sgron text, in its chapter dedicated to rDzogs chen (Ch. 7, 290-494), quotes extensively from several Sems sde bco brgyad texts, and it seems that sNga lnga texts must have been considered important as most of them are included as chapters in the tenth-century text, Kun byed rgyal po.

5.3.1 Collections and Editions Used

The data of the present study is based on the quotes from the bSam gtan mig sgron and the quoted textual body as found in four Sems sde texts, Rig pa’i khu byug, rTsal chen sprugs pa, Khyung chen lding ba and Mi nub pa’i rgyal mtshan.

PHYLOGENETIC ANALYSIS 105

The texts are extant as independent texts in Bai ro’i rgyud ’bum and the gTing skyes edition of the rNying ma rgyud ’bum, as well as chapters in the Kun byed rgyal po—a text preserved in several editions of the two major canonical collections the bKa’ ’gyur and the rNying ma rgyud ’bum. The bSam gtan mig sgron quotes 162 lines that can be identified in the four texts, out of a total of 354 lines that make up the texts. In the bKa’ ’gyur editions, the “core” of the texts, that is the text as found in the independent versions, is preceded by introductory remarks and is followed by a conclusion. These introductory remarks and conclusions have not been taken into consideration in the present study. The Kun byed rgyal po is a rDzogs chen Sems sde text that is found in Bai ro rgyud ’bum

(2nd and 3rd sections of the text only, chs. 58-84), in all editions of the rNying ma rgyud ’bum, and several editions of the bKa’ ’gyur (for more on Kun byed rgyal po, see Neumaier-Dargyay 1992). Three editions of the rNying ma rgyud ’bum were consulted, sDe dge, gTing skyes and mTshams brag, and five editions of the bKa’

’gyur, sDe dge, Peking Qianlong, Co ne, sNar thang and lHa sa.

The Bai ro’i rgyud ’bum is a collection in the rNying ma tradition that was published in 1971, by Tashi Y. Tashigangpa. This is the only known extant edition of this collection. The introductory remarks accompanying the work note that it was reproduced from the rare manuscript belonging to rTogs ldan Rimpoche, and that it contains ancient tantras and esoteric instructions compiled and translated by the eighth-century Tibetan master Vairocana. According to Kapstein, the Bai ro’i

PHYLOGENETIC ANALYSIS 106 rgyud ’bum most probably derived from earlier, smaller collections. During the twelfth century, there was an “ongoing process of compilation” of texts and traditions within the Zur lineage, all connected with the rDzogs chen Sems sde teachings. The Bai ro’i rgyud ’bum was developed as a result of this process and most likely includes “a truncated version of the rDzogs chen teachings of Zur”

(Kapstein 2008: 283-84).

The rNying ma rgyud ’bum is the most important canonical collection of the rNying ma tradition. It is based on texts stored at ’Ug pa lung in gTsang (11th century) and is extant in several editions that vary in structure, sequencing, and extent. The collection contains texts that for the most part were excluded from the broader canons of Tibetan Buddhism, the bKa’ ’gyur and the bsTan ’gyur. The sDe dge rNying ma rgyud ’bum edition was produced between 1794 and 1798 in sDe dge, eastern Tibet. It was commissioned by the Queen of sDe dge, sGa rje Tshe dbang lha mo, and edited by ’Gyur med Tshe dbang mchog grub (1761-1829). The gTing skyes edition was produced in southern Tibet, in the early eighteenth century, by gSang sngags bstan dzin. The mTshams brag was produced in Bhutan, calligraphed between 1728 and 1748, on the order of mTshams brag Sprul sku

Ngag dbang ’brug pa (1682-1748) (see Derbac 2007: 8-35).

The bKa’ ’gyur collection is the primary canon in Tibetan Buddhism and all traditions accept it as canonical. It is divided into several branches, with Them spangs ma (1431) and Tshal pa (1347-1349) as the most well known. It is not clear,

PHYLOGENETIC ANALYSIS 107 however, when rNying ma texts were first included in the bKa’ ’gyur. Bu ston Rin chen grub (1290-1364), a prominent Tibetan scholar of the fourteenth century, excluded the so-called rNying ma tantras from his version of the Tibetan Tripiṭaka and considered that these texts should not be in the bKa’ ’gyur canon (Dargyay

1985: 285-86). It would seem that the first editor who inserted the rNying ma texts in this canon may well have been a contemporary of Bu ston, Kun dga rdo rje (alias

Si tu dGe ba’i blo gros). In one of his works, sNga ’gyur rgyud ’bum, ’Jigs med gling pa (1729-1798) specifically mentions Kun dga rdo rje as the editor of a Tshal pa edition of the bKa’ ’gyur, and points out that he did include a section specifically reserved for rNying ma texts (rnying rgyud) (Dargyay 1985: 286-87).

The rNying rgyud section, where the Kun byed rgyal po is located, is not included in most of Them spangs ma editions (such as the sTog Palace, London, and Tokyo) with the exception of Ulan Bator (not accessible). All Tshal pa editions include this section. The sDe dge (1729-1733) printed bKa’ ’gyur is based primarily on the ’Jang sa tham (1608-1621) edition and was checked against lHo rdzong manuscript edition. It was commissioned by bsTan pa tshe ring (1678-1738) and redacted by Si tu Chos kyi ’byung gnas (1699/1700-1774). The Peking

Qianlong edition of the bKa’ ’gyur was made in Peking, in 1737, under the reign of

Emperor Qianlong (1735-1796). It is a reprint based on the plates of the Kangxi edition (1684-1692)—itself an “emended” adaptation of the Yongle edition (1410).

The Co ne edition was printed from a ’Jang sa tham exemplar, between 1721 and

PHYLOGENETIC ANALYSIS 108

1731, and is considered a direct copy of this edition. The sNar thang is a 1730-1732 edition, and the lHa sa, a 1934 edition (Harrison 1992: xvi-xxiv; Zimmermann

2002: 164-172).

5.3.2 The Five Early Seme sde Texts

There is some confusion about which texts actually belong to the sNga lnga group as rNying ma sources are far from being consistent in their lists, and this fact requires some clarification. Nyang ral Nyi ma ’od zer (1136-1204) is the first scholar to have suggested such a list of texts (as far as our sources tell us). In his two major works, Me tog snying po (a Buddhist religious history) and Zhangs ling ma (a hagiography of Padmasambhava), he points out that the following works are the five early translations:

1) Rig pa’i khu byug (Ch. 31 in KG)

2) rTsal chen sprugs pa (Ch. 27 in KG)

3) Khyung chen lding ba (Ch. 22 in KG)

4) rDo la gser gzhun (Ch. 26 in KG)

5) Mi nub pa’i rgyal mtshan nam mkha’ che = Nam mkha’ che (Ch. 30 in KG) (Chos ’byung me tog snying po sbrang rtsi’i bcud, 420; on Zhangs ling ma see Liljenberg 2009: 51, n. 2).

PHYLOGENETIC ANALYSIS 109

Two centuries later, kLong chen pa (1308-1364) also mentions these same five texts (in the same order), in two of his major works, Lung gi gter mdzod (p. 749) and Grub mtha’ mdzod (p. 1189). However, Vairo ’dra ’bag chen mo, a thirteenth to fourteenth-century hagiographical account of Vairocana, gives a slightly different version, as it replaces the rDo la gser gzhun with the Thig le drug pa

(Karmay 1988: 23-24). The rNying ma scholar, Sog bzlog pa bLo gros rgyal mtshan (1552-1624), also gives a different list of texts in his work entitled Nges pa’i don gyi ’brug sgra. He suggests sKu la ’jug pa and rDzogs pa spyi gcod instead of Khyung chen lding ba and rDo la gser gzhun, but acknowledges that some people do list these texts, “instead of” (tshab tu) the ones that he has given

(Nges pa’i don gyi ’brug sgra, 470.1).

The question is, then, “Why do we have different lists of texts?” And the answer would seem to be, “It is because of the text Nr. 4 on Nyang ral’s list, rDo la gser gzhun.” Citing sKu la ’jug pa (an obscure text) instead of Khyung chen lding ba (a well known text) seems to be more of an anomaly on Sog bzlog pa’s part than a real problem, as no other scholar ever suggested that Khyung chen should not be on the list. With rDo la gser gzhun the problem is slightly more complex as the reason why this text should be excluded from the list appears to have a lot more to do with the body of the text that circulated under the name of rDo la gser gzhun, than with the title of the text itself. The Bai ro rgyud ’bum text, Pan sgrub rnams

PHYLOGENETIC ANALYSIS 110 kyi thugs bcud snying gi nyi ma, gives the same list as Nyang ral and kLong chen pa. The text is expository in nature and has chapters dedicated to each of the five sems sde texts. Yet, in its chapter on rDo la gser gzhun (49.1-64.4) it quotes from

Byang chub sems bsgom pa’i rdo la gser zhun, a text specifically attributed to

Mañjuśrīmitra (’Jam dpal bshes gnyen) (On Pan sgrub rnams kyi thugs bcud snying gi nyi ma see Kapstein 2008). The identification of the rDo la gser gzhun with the text authored by Mañjuśrīmitra may have rendered it as unworthy to be included on the list. The sNga lnga texts were considered to be canonical in nature, therefore, not philosophical treatises. Apparently, scholars such as Nyang ral and kLong chen pa identified the rDo la gser gzhun title with a different text; a text that was a scripture—no longer extant as an independent text but still preserved as

Chapter 26 in the Kun byed rgyal po (see chapter 3).

Rig pa’i khu byug = rDo rje tshig drug — TkK vol. 1, fol. 209a1-209a4; Bg vol. 5, p. 306; Tk vol. 1, ff. 48b7-49a3; Tb vol. 1, fol. 57a2-57a5; Dg vol. 5, fol. 39b3-

39b5; D vol. 97, fol. 38b3-38b5; Q vol. 20, fol. 36b5-36b8; C vol. 19, fol. 43a5-

43a7; N vol. 98, ff. 53a6-53b2; H vol. 97, fol. 54b1-54b4. The text consists of a total of six lines, and bSam gtan mig sgron gives two quotes with a total of two lines. The second quote repeats one of the lines of the first quote (347).

PHYLOGENETIC ANALYSIS 111 rTsal chen sprugs pa — TkK vol. 1, ff. 211a3-211b1; Bg vol. 5, pp. 306-308; Tk vol. 1, ff. 42b1-43a2; Tb vol. 1, ff. 49b7-50b1; Dg vol. 5, fol. 34b1-34b7; D vol.

97, fol. 33b1-33b7; Q vol. 20, ff. 32a4-32b2; C vol. 19, ff. 37b7-38a6; N vol. 98, ff. 46a6-46b7; H vol. 97, ff. 47a6-47b7. Two quotes of this text are found in bSam gtan mig sgron, and four lines out of a total of twenty-one are cited.

Khyung chen lding ba — TkK vol. 1, ff. 209a4-211a3; Bg vol. 5, pp. 308-314; Tk vol. 1, ff. 37a6-39a7; Tb vol. 1, ff. 44a2-46a7; Dg vol. 5, ff. 30a6-32a1; D vol. 97, ff. 29a6-31a1; Q vol. 20, ff. 28a7-29b8; C vol. 19, ff. 33b1-35a6; N vol. 98, ff.

40b4-42b7; H vol. 97, ff. 41b3-43b7. This text is extensively quoted in SM.

Eighteen quotes as given by bSam gtan mig sgron, with a total of forty-eight lines, are identifiable in Khyung chen lding ba. The text has a total of 108 lines. There are some minor problems with some quotes as one quote cannot be identified: rang byung ye shes mi rtog kun tu ji bzhin gnas/ (341); and one is not to be found in the

Khyung chen but it is traceble to rDo la gser zhun, as preserved in Kun byed rgyal po (ch. 26): ston pa’i lung bzhin tshig gi mtha’ spyod las du de mthong na/ de ngag

’khrul te don med pas mi ched po rnams kyis de ltar mi lta’o// (304). What is interesting about this misattributed quote is that it cannot be found in the Byang chub sems bsgom pa’i rdo la gser zhun, the text that bSam gtan mig sgron quotes.

PHYLOGENETIC ANALYSIS 112 rDo la gser zhun — Tk vol. 1, ff. 41b1-42b1; Tb vol. 1, ff. 48b4-49b7; Dg vol. 5, ff. 33b3-34b1; D vol. 97, ff. 32b3-33b1; Q vol. 20, ff. 31a7-32a4; C vol. 19, ff.

37a1-37b7; N vol. 98, ff. 45a3-46a6; H vol. 97, fol. 46a3-47a6. This text is not quoted at all in bSam gtan mig sgron. The text that is quoted (ten times) is Byang chub sems bsgom pa’i rdo la gser zhun, and this text has only a few lines in common with the rDo la gser zhun, some in an expended form: bde gshegs ma lus yum du gyur pas rgyal ba kun gyi lam gcig go/ vs. bde gshegs yum gyur kun gyi mtshungs lam ste/.

Mi nub pa’i rgyal mtshan nam mkha’ che = Nam mkha’ che — TkK vol. 1, ff.

211b1-214b2; Bg vol. 5, pp. 384-395; Tk vol. 1, ff. 45b4-48b7; Tb vol. 1, ff. 53b1-

57a2; Dg vol. 5, ff. 37a2-39b3; D vol. 97, ff. 36a2-38b3; Q vol. 20, ff. 34b1-36b5;

C vol. 19, ff. 40b2-43a5; N vol. 98, ff. 49b6-53a6; H vol. 97, ff. 50b7-54b1. This text is one of the most extensively quoted texts in bSam gtan mig sgron with a total of thirty-four quotes. Out of the 219 total lines, almost half, 108 lines are quoted in bSam gtan mig sgron and all are identifiable.

PHYLOGENETIC ANALYSIS 113

NGB

Fig. 2 - Maximum Parsimony (PAUP 4.0.b10)

Fig. 3 - NeighborNet (SplitsTree 4.11.3)

PHYLOGENETIC ANALYSIS 114

5.4 Analysis of data

In the present study, the Maximum Parsimony method, as implemented in the phylogenetic software PAUP 4.0 beta 10, and the Neighbor-Net method, as implemented in SplitsTree 4, were applied to a collation of the bSam gtan mig sgron’s rDzogs chen Sems sde quotations and the identical textual body as found in four extant ancient texts, Rig pa’i khu byug, rTsal chen sprugs pa, Khyung chen lding ba and Mi nub pa’i rgyal mtshan. The collation was done manually from eleven block-prints and manuscripts, and the collated data was inputted into the program developed by Stefan Hagel, the Classical Text Editor. The positive critical apparatus was exported as a genealogical data in a matrix format, and this file was later converted to a NEXUS format—a format that can be read by most phylogenetic programs. Data of the matrix included 11 taxa (extant editions), and

287 characters (sites or points of variation). Variant readings ranged from 1 (single reading) to 5 different readings per site. 131 points of variation were considered as parsimony-informative and 156 as un-informative. The data file was analyzed and a graphical representation of the results was generated as an out-put file (see Fig. 2 and Fig. 3 above).

In SplitsTree, the Neighbor-Net method was used to construct a phylogenetic network. The inclusion or exclusion of uninformative sites had no

PHYLOGENETIC ANALYSIS 115 significant impact on the position of the 11 taxa or the number of splits (29). The out-put graphic (Fig. 3) displays both informative and un-informative sites. In

PAUP, Exhaustive Search with optimality criterion set at Maximum Parsimony was selected to search through all possible trees for the ones having the best parsimony score. The scores of the trees ranged from 486 for the best to 651 for the worst. Out of a total of 34 million trees evaluated, 2 trees were retained as most parsimonious.

The program then computed a strict consensus of the two trees in order to depict only those relationships that are unambiguously supported by all the data. The overall Consistency Index (CI) turned out to be 0.82 out of 1.00. The Consistency

Index represents the degree of consistency between the available data and the structure of the most parsimonious tree.

By default, the trees re-constructed by the phylogenetic programs are not rooted, and based on the external evidence it would seem that three points could be considered as most plausible points of origin: point a, a1, and a2 (Fig. 2). The tree could be rooted manually at point a, as an automatic rooting requires that an out- group be selected. This would lead to a polytomy and suggest that the independent gTing skyes (TkK) and the bSam gtan mig sgron (SM) traditions go back directly to the archetype, while the independent texts as found in the Bai ro rgyud ’bum

(Bg) collection belong to a textual tradition that also includes texts as chapters in the Kun byed rgyal po, in both the rNying ma rgyud ’bum (NGB) and the bKa’

’gyur (K) collections. The same grouping would result if point a1 is selected as the

PHYLOGENETIC ANALYSIS 116 point of origin, and the gTing skyes (TkK) and bSam gtan mig sgron (SM) as the out-groups. The only difference would be that this time we would have a dichotomy instead of a polytomy. There is some historical support for these groupings. Zur po che, who flourished in the eleventh century, established the ’Ug pa lung monastery and initiated a process of collecting texts that by the fourteenth century transformed the monastery into a major library and a great center of learning. In the fifteenth century, Rat na gLing pa used the library to prepare his own edition (Lhun drup) of the rNying ma rgyud ’bum (NGB).

There are, however, some problems with placing the Bai ro rgyud ’bum

(Bg) in the same textual tradition as the rNying ma rgyud ’bum (NGB) and the bKa’ ’gyur (K), both textual and historical. In his article, “The Sun of the Heart and the Bai ro rgyud ’bum,” Kapstein (2008) points out that Bai ro rgyud ’bum (Bg) most probably derived during the twelfth century from earlier, smaller collections, all connected with the rDzogs chen Sems sde teachings. This would explain why the collection does not include tantric texts, but only rDzogs chen (a total of 196 texts). It would seem that initially, as a rDzogs chen practitioner, Zur po che collected mostly rDzogs chen texts, and the Bai ro rgyud ’bum (Bg) collection was established based on these texts. Furthermore, as the collection includes the early rDzogs chen texts only as independent texts the logical step would be to place it in the same group with gTing skyes (TkK) and the bSam gtan mig sgron (SM). This grouping is also supported by the Neighbor-Net method (Fig. 3) and suggests that

PHYLOGENETIC ANALYSIS 117 the most probable point of origin, the oldest reconstructable archetype, is located at point a2.

In our stemma, two places of contention are most striking. In one case (Fig.

2), sDe dge (Dg) rNying ma rgyud ’bum (NGB) edition and sDe dge (D) bKa’

’gyur (K) edition are group together and depicted as a collapsed branch by the

PAUP program. This results in a trichotomy at node x2 between the two sDe dge editions and the hypothetical ancestor of the group represented by the Peking and the Cone editions, at node x3. As all resolved nodes in a phylogenetic tree are binary, a trichotomy suggests that there is some ambiguity in the data. It would seem that the conflict is due to contamination. The two sDe dge editions (Dg and

D) belong to two different canons, the rNying ma rgyud ’bum (NGB) and the bKa’

’gyur (K); yet, the same establishment, sDe dge par khang, produced both editions and both are highly edited. This fact is supported by the study—even though the total number of single readings is 158, the sDe dge (Dg) rNying ma rgyud ’bum

(NGB) does not contain even one reading that is not shared with sDe dge (D) bKa’

’gyur (K), and sDe dge (D) bKa’ ’gyur (K) includes only one single reading, a spelling error.

In another place of contention, Co ne (C), known to be a direct copy of the

Lithang (1621) edition, is grouped together with Peking Qianlong (Q). Peking

Qianlong is a corrected copy in the Chinese tradition of the Tibetan bKa’ ’gyur (K) that dates back to 1410, the Yongle edition. The correction was based on the

PHYLOGENETIC ANALYSIS 118

Lithang edition and this would explain why Co ne (C) and Peking Qianlong (Q) are depicted as being closely related even though the two editions only share a distant common ancestor (the Tshal pa edition of the bKa’ ’gyur) and not a direct recent ancestor (the x3).

Despite these caveats, the consensus trees in both Fig. 2 and Fig. 3 suggest that at one time the transmission of the Sems sde texts did split into two main branches, independent texts and texts preserved as chapters in the Kun byed rgyal po. Furthermore, both methods successfully identify the three main groups: the independent group formed by gTing skyes (TkK) rNying ma rgyud ’bum (NGB), bSam gtan mig sgron (SM), and Bai ro rgyud ’bum (Bg); the Kun byed rgyal po

(KG) rNying ma rgyud ’bum (NGB) group, formed by gTing skyes (Tk) and mTshams brag (Tb); and the Kun byed rgyal po (KG) bKa’ ’gyur (K) group, formed by sDe dge (D), Qianlong (Q), Co ne (C), sNar thang (N) and lHa sa (H).

5.5 Conclusion

As a conclusion, it can be said that the phylogenetic analysis clearly points out that the bSam gtan mig sgron is indeed an ancient text that preserves an independent ancient textual tradition. Furthermore, there is no doubt that the present-day rDzogs chen Sems sde texts are “the same” texts that the author of the bSam gtan mig sgron, gNubs chen Sangs rgyas ye shes, had access to in the late ninth to early tenth century. The correspondence between the quotes as given in the

PHYLOGENETIC ANALYSIS 119 bSam gtan mig sgron and the wording found in the extant Sems sde texts suggests that the textual bodies did undergo some changes in the course of time. However, these alterations are not extensive enough or significant enough for the texts to be treated as different texts altogether. They are variant readings that emerged during the transmissional process. Textual transmission is a long and complex process and no copyist is infallible from making errors. The texts obviously have seen some alteration, and the recorded 158 single readings testify to that. Overall, however, there is no doubt that the texts that we have today are the same texts that were extant in the past.

The phylogenetic analysis supports the historical evidence. The bSam gtan mig sgron text may not have been a well-known text, but its antiquity and the fact that it preserves an ancient textual tradition is unquestionable. By implication, the claim that the early rDzogs chen Sems sde texts are “extractions from the Kun byed rgyal po,” turns out, after all, as nothing but a claim that has no phylogenetic or historical support whatsoever. The bSam gtan mig sgron does not mention Kun byed rgyal po, and its author, gNubs chen, seems to be unaware of the existence of this particular text. Yet he quotes extensively from several Sems sde texts, that were later included as chapters in the Kun byed rgyal po. From a methodological point of view, the four Sems sde texts are grouped as independent texts, or as chapters of the Kun byed rgyal po, for a good reason—they belong together, they belong to the same textual tradition. If the early Sems sde texts were extracted from

PHYLOGENETIC ANALYSIS 120 the Kun byed rgyal po (or the bSam gtan mig sgron was a gter ma text), the variant readings would betray their origin. The “independent” witnesses, that is, the texts as preserved in the gTing skyes (TkK) rNying ma rgyud ’bum (NGB) edition, the

Bai ro rgyud ’bum (Bg) collection, and the bSam gtan mig sgron’s (SM) quotations, would be grouped together with the texts preserved as chapters in the

Kun byed rgyal po, or be depicted as direct descendants of these texts. Obviously, that is not the case! A phylogenetic analysis, based on methods specifically designed to establish phylogenetic (stemmatic) relationships, may have its shortcomings, however, it proves again (and again) that a computer-aided stemmatic is a useful and indispensable tool to a textual critic.

PHYLOGENETIC ANALYSIS 121

5.6 Collated bSam gtan mig sgron’s Sems sde quotations

Rig pa’i khu byug

<1> zin pas rtsol1 ba'i nad spangs te/A

/lhun gyis gnas pas2 bzhag3 pa yin/[323]

rTsal chen sprugs pa

<1> rtog4 dpyod5 rnam dag gnyis med chos dbyings la/

/log6 rtog byis pa'i spyod7 pas ’ga’8 zhig bslu9/[294]

<2> slub byed10 rtsal sprugs smon pa med pa yi11/

/byar12 med lhun rdzogs snga ma de nyid don13/[347]

Khyung chen lding ba

<1> mi gyo14 rang bzhin15 gnas pa’i16 sman17 chen ma byas na/[314]

Variant Readings:

1 rtsol SMTkKTkTbDgDCNH; brtsol BgQ | 2 gnas pas SMBgTkDgDQCNH; gnas pa TkK; grup pas Tb | 3 bzhag SMBgQCNH; gzhag TkKTkTbDgD 4 rtog SMTkKTkTbDgDQCNH; rtogs Bg | 5 dpyod SMTkKTkTb; spyod BgDgDQCNH | 6 log SMTkKTkTbDgDQCNH; blos Bg | 7 spyod SMTkKBgTkTbDgDCNH; spyos Q | 8 ’ga’ SM; ci TkKBgTkTbDgDQCNH | 9 bslu SMBgTb; slu TkKTk; bslus DgDQCNH | 10 slub byed SM; sgrub med TkKTk; bsgrub med BgTbDgDQCNH | 11 pa yi SMTkK; pa’i Bg; pas na TkTbDgDQCNH | 12 byar SM; bya TkKBgTkTbDgDQCNH | 13 don SMTkTbDgDQCNH; do TkKBg 14 mi gyo SMTkK; mi g.yo BgTkTbDgDQCNH | 15 rang bzhin SMTkKBg; mnyam par TkTbDgDQCNH | 16 pa’i SMTkK; pas BgTkTbDgDQCNH | 17 sman SMTkKBgTkTbDgDCH; smad QN

a This line (zin pas rtsol ba’i nad spangs ste/) is also quoted at page 347. PHYLOGENETIC ANALYSIS 122

/mtho ris bgrod pa’i rgyu des18 nyon mongs zin par ’gyur/[315]

<2> bdud rtsi chen po’i rang bzhin la19/

/bza’ ba’i thabs la rag pa20 med zhes pa dang/[317]

<3> lam21 med lam du ’jug pa22 lam gyi23 nad chen yin24/

phyin par ’dod pa25 ri dgas smig rgyu snyed pa26 ’dra/

/rnyed pa’i27 yul med ’jig rten gsum las yongs mi28 ’byung29/

/sa bcur30 ltos pa’i gnas kyang byang chub sgrib pa yin/[319]

<4> ye nas sangs rgyas31 bdag nyid mngon sum32 gnas pa la/

/grib33 ma gzugs kyi dpe bzhin de la rtog pa ’byung/[339]

<5> rang byung34 ye shes mi rtog kun tu ji bzhin35 gnas/[341]

<6> rang byung36 ye shes nyid kyang tshig gi mtha’ dang bral/[341]

18 des SMTkKTkTbDgDQCNH; de Bg | 19 la SMTkKBgDgDQCNH; pas TkTb | 20 rag pa SMTkKTkTbNH; rag ba DgDQC; regs pa Bg | 21 lam SMBgTkTbDgDQCNH; la TkK | 22 ’jug pa SMTkKBg; zhugs pa TkTbDgDQCNH | 23 lam gyi SM; rnams kyi TkKBgTkTbDgDQCNH | 24 chen yin SM; che ba TkKBgTkTbDgDQCNH | 25 pa SM; pas TkKTkTbDgDQCNH; par Bg | 26 snyed pa SM; snyeg pa TkKBgTkTbC; snyegs pa DgDQNH | 27 rnyed pa’i SMTkKBgTkTbDgDC; snyed pa’i QNH | 28 las yongs mi SM; las yong mi TkKBg; las mi TkTb; du mi DgDQCNH | 29 ’byung SMTkKBg; ’byung zhing TkTbDgDQCNH | 30 sa bcur SMTkKBgTkTbDgD; sa bcud QCNH | 31 sangs rgyas SMTkKBg; rgyas pa’i TkTbDgDQCNH | 32 sum SMTkKBgTkTbDgDQCH; gsum N | 33 grib SMBgTkTbDgDQCNH; sgrib TkKBg | 34 rang byung SMTkK; rang ’byung BgTkTbDgDQCNH | 35 ji bzhin SMTkKTkTbDgDQCNH; rang bzhin Bg | 36 rang byung SMTkK; rang ’byung BgTkTbDgDQCNH PHYLOGENETIC ANALYSIS 123

<7> zin pas37 chags dang bral yang38 zhi/

/dngos po med dang rnam par39 spangs40/

/bdud rtsi chen po’i rang bzhin la41/

/bza’ ba’i thabs la rig pa42 med/[347]

<8> shan pa43 smad44 ’tshong mtshams45 med lnga46/

/yongs su rdzogs pa47 chos kyi rtsi48/

/bde ba che la49 gzhan med do50/[353]

<9> bde ba ’dod pa51 bde bar52 rgyab kyis phyogs/

/bde ba53 zin te bde bas bde ba ’tshol54/

/byang chub ’khrul pa55 ye rdam56 chos la za57/

/de lta58 yul59 can sangs rgyas mthong ba med/[356]

37 zin pas SMTkK; zin dang BgTkTbDgDQCNH | 38 yang SM; dang TkKBgTkTbDgDQCNH | 39 par SMTkKBgTkTbDNH; pa DgQC | 40 spangs SM; spang TkK; spong BgTkTbDgDQCNH | 41 la SMTkKBgDgDQCNH; pas TkTb | 42 rig pa SM; rag pa TkKTkTbNH; rag ba DgDQC; regs pa Bg 43 shan pa SMTkKBgTkTbH; shen ba DgDQ; shan ba C; shen pa N | 44 smad SMBgTkTbDgDQCNH; smang TkK | 45 ’tshong mtshams SMTkKTbDgDQCNH; mtshong mtshams Bg; ’tshong ’tshams Tk | 46 lnga SMTkKTkTbDgDQCNH; byed Bg | 47 pa SMTkKDgDQCNH; pas BgTkTb | 48 rtsi SMTkKBg; rtsis TkTbDgDQCNH | 49 la SM; las TkKBgTkTbDgDQCNH | 50 med do SMTkKTkTbDgDQCNH; min no Bg | 51 ’dod pa SM; ’dod pas TkKTkTb; ’dod na DgDQCNH; ’dos pas Bg | 52 bde bar SM; bde la TkKBgTkTbDgDQCNH | 53 bde ba SM; bde des TkK; bde bas BgTkTbDgDQCNH | 54 ’tshol SMTkKBg; tshol TkTbDgDQCNH | 55 pa SMTkTbDgDQCNH; bas TkK; pas Bg | 56 ye rdam SMTkKBgTkTbDgDC; ye rbam QNH | 57 za SMTkKBgTkTbDgDCNH; zab Q | 58 de lta SM; de lta’i TkKBg; ’di lta’i TkTbDgDQCNH | 59 yul SMTkKBgTkTbDgDCNH; sul Q PHYLOGENETIC ANALYSIS 124

<10> chos dang chos nyid60 rtag tu ’dres shing61 dbyer med la62/

don dam chos shes63 lhag par64 bshad pa65 med/

’khrul66 dang67 byang chub lam de68 rtog pas thob pa69 med/[368]

<11> brtags na med la bzhag70 na [382] sna tshogs71 ’byung ba’i legs pa che72/

dngos73 su mi snang chags pa’i cha lugs kun la ston74/[383]

<12> btang ngo75 bzhag go rang bzhin de76/

/’dod pa med cing len pa med/

bred77 pa rdul tsam mi skye zhing78/

/khyung chen nam mkhar79 ldings pa ltar80/

/spros pa81 med cing bsdu82 ba med/

/stong par83 mi dgos bza’ ba84 med/

60 chos nyid SMTkK; chos min BgTkTbDgDQCNH | 61 shing SMTkKBgTkTbDgDQCH; shid N | 62 dbyer med la SMTkKBg; mtshungs TkTbDgDQCNH | 63 chos shes SMBgTkDgDQCNH; chos zhes TkKTb | 64 lhag par SMTkKBg; lhag pa’i gnas Tk; lhag pa’i gnas shes DgDQCNH; lhag pa’i gnas zhes Tb | 65 bshad pa SMTkK; bshad pa’ang TkTbDgDQCNH; bshad Bg | 66 ’khrul SMTkKBgTkTbDgDCNH; la brul Q | 67 dang SMTkKBgTkDgDQCNH; dang ma ’khrul Tb | 68 de SMBgTkTbDgDQCNH; ste TkK | 69 pa SMTkKBg; pa’ang TkTbDgDQCNH | 70 bzhag SMBgTkDgDQCNH; gzhan TkK; gzhag Tb | 71 sna tshogs SMTkKBg; legs par TkTbN; legs pa DgDQCH | 72 ’byung ba’i legs pa che SMTkKBg; ’byung ba che TkTbDgDQCNH | 73 dngos SMTkKTkTbDgDQCNH; sngos Bg | 74 ston SMTkKTkTbDgDQCNH; bston Bg | 75 btang ngo SMTkKBgTkTbDgDQNH; btad do C | 76 de SM; te TkKBgTkTbDgDQCNH | 77 bred SM; brod TkKBgTkTbDgDQCNH | 78 skye zhing SMTkK; skyed cing Bg; bskyed de TkTb; bskyed do DgDQCNH | 79 nam mkhar SMTk; nam mkha’ TkKBgTbDgDQNCH | 80 ldings pa ltar SMTkK; lding pa bzhin BgTk; lding ba bzhin DgDQCNH; ldings pa bzin Tb | 81 spros pa SM; spro ba TkKBgTkTbDgDQCNH | 82 bsdu SMBgTkTbDgDQCNH; bsru TkK | 83 stong par SMTkK; stor bar TkTbDgDQCNH; om. Bg | 84 ba SMTkKDgDQCNH; ba’ang TkTb; om. Bg PHYLOGENETIC ANALYSIS 125

rgya85 mtsho bzhin du ye gnas la/

/chos rnams sna tshogs ’byung bar byed/

/yon tan nam mkha’i mtha’ dang86 mnyam/

/bsdu ba’i gnas su nges pa med/[383]

<13> shin tu myur ba’i87 bsam pa88 kun dang bral/[412]

<14> snang ba mtsho89 chen po bzhin/

mi rtog90 nam mkha’i mtha’91 ltar yangs/[413]

<15> ji bzhin pa zhes rnam rtog de phyir92 rtag sgom pa93/

bde ba chen bo94 ’dod pas95 chags pa’i96 nad97 yin te/[421]

<16> mngon sum thob par ’dod pas98 rgyu la rtag tu bsgom99/

/bsgom100 chags bde ba de las nyams101 pa’i don mi ’byung/[431]

85 rgya SMTkKC; rgyu BgTkTbDgDQNH | 86 dang SMTkKBgTkTbDNH; dag DgQ; dad C | 87 shin tu myur ba’i SMTkTbDgDQCNH; shin tu myur pa’i TkK; shin myur ba’i Bg | 88 bsam pa SM; ye shes bsam pa TkKBgTkTbDgDQCNH | 89 mtsho SM; rgya mtsho TkKBgTkTbDgDQCNH | 90 mi rtog SMTkKBgTkTbDgDQCH; mi rtogs N | 91 mtha’ SMBgTkTbDgDQCNH; mkha’ TkK | 92 de phyir SMBgTkDgDQCNH; ji bzhin TkK; de’i phyir Tb | 93 rtag sgom pa SMTkK; rtag bsgom pa Tk; rtog bsgom pa TbDgDQCNH; rtog bsgos pa Bg | 94 bo SM; po TkKBgTkTbDgDQCNH | 95 pas SMTkKBgTkTbDgDQCH; pa’i N | 96 pa’i SMBgTkTbDgDQCNH; pas TkK | 97 nad SMTkKTkTbDgDQCNH; gnas Bg | 98 pas SMTkKBg; nas TbDgDQCNH; na Tk | 99 bsgom SMBgTkTb; bsgoms TkK; sgom DgDQCNH | 100 bsgom SMBgTkTbQC; sgom TkKDgDNH | 101 las nyams SM; la mnyam TkKBgDgDQCNH; las mnyam TkTb PHYLOGENETIC ANALYSIS 126

<17> bya bral yul la mi gnas gnyen pos bcos su med/[434]

<18> ’dod spyod snying rjes brlan pa la102/

/gang du dga’103 bar byang chub104 spyod/[451]

Mi nub rgyal mtshan nam mkha' che

<1> ma skyes pas na ’jig pa105 med/b

gdod106 nas snying po byang chub la/

/rtsal107 ba’i bsam pas dbyings mi108 bskyod/[294]

<2> don rnams ji bzhin mi bskyod do109/

/bya ba med pas110 grol bas grol/

rang111 byung112 ye shes btsal113 med pas/

/grol nas114 grol ba’i lam yang ston/[295]

<3> gong nas gong du tshangs pa’i lam/

102 brlan pa la SM; brnyal pas na TkK; brlan pas na BgTkTbDgDQCNH | 103 dga’ SMTkKBgTkTbDgDCNH; gda’ Q | 104 byang chub SMTkKBgTbDgDQCNH; snying rje Tk 105 ’jig pa SMBgTkTbDgDQCH; ’jigs pa TkK; ’jig ba N | 106 gdod SMTkKBgTkTbDgDQCH; gdon N | 107 rtsal SM; brtsol TkK; btsal TkTbDgDQCNH; brtsal Bg | 108 mi SMTkKBgTkTbDgQCNH; mir D 109 bskyod do SM; skyod de TkKBgQ; bskyod te Tk; bskyod de TbDgDCNH | 110 pas SMTkKBgTkDgDQCNH; par Tb | 111 rang SMTkKBgTkTbDgDQC; nges NH | 112 byung SMTkKTkTb; ’byung BgDgDQCNH | 113 btsal SMBgTbDgDQCNH; brtsal TkK; brtsol Tk | 114 nas SMTkKBgTkTb; na DgDQHCN b This line (ma skyes pas na jig pa med/) is also quoted at page 342. PHYLOGENETIC ANALYSIS 127

/bya bral chos dang mthun115 pa min116/

/ci ste lam gyi117 bgrod gyur118 na/

nam mkha’i119 mtha’ bzhin thug120 pa med/[319]

<4> yid kyi121 bsam pas122 yon bdag ste123/

/bltas pas124 stobs ni125 bkram pa’o126/

/mthong bar gyur pa127 dngos grub la/

/de nyid128 mnyam bzhag129 rdzogs pa’o/[323]

<5> gang dang gang gi130 gang du131 yang/

/kun ’khol132 kun spyod bdag la133 ’byung134/[331]

<6> nam mkha’ bzhin de135 mtha’ mnyam zhing/

/gzhan la bstos136 ’gyur chos ma yin/

/mtshungs pa med pa’i ye shes137 kyis138/

115 mthun SMTkKBgTbDgDQNH; ’thun TkC | 116 pa min SMDgDQCNH; pa bzhin TkK; pa’i phyir Bg; pa men TkTb | 117 gyi SM; la TkKBgTkTbDgDQCNH | 118 gyur SMTkTb; ’gyur TkKBgDgDQCNH | 119 nam mkha’i SMBgTkTbDgDQCNH; nam mkha’ TkK | 120 thug SMTkKBg; thob TkTbDgDQCNH | 121 kyi SMBg; kyis TkKTkTbDgDQCNH | 122 bsam pas SMDgDQCNH; bsams pas TkKTkTb; bsam pa’i Bg | 123 bdag ste SMTkKBgTkTbDgDQ; gdag ste CN; gdags te H 124 bltas pas SMBgTkTbDgDQCNH; ltas pa’i TkK | 125 stobs ni SMTkKBg; tshogs ni TkTbDgDQCNH | 126 pa’o SMTkKTkTbDgDQCNH; pa po Bg | 127 pa SMTkK; pas BgTkTbDgDQCNH | 128 de nyid SMTkTbDgDQCNH; de ni TkKBg | 129 bzhag SMTkKBgDgDQCNH; gzhag TkTb | 130 gi SMTbDgDQCNH; gis TkKBgTk | 131 du SMTkKBgTkTbDgDCNH; tu Q | 132 ’khol SMTkDgDQCNH; bkol TkK; khol Bg; ’khor Tb | 133 la SM; ba TkK; las BgTkTbDgDQCNH | 134 ’byung SMBg; byung TkKTkTbDgDQCNH | 135 de SM; te TkKDgDQCNH; du BgTkTb | 136 bstos SM; stobs TkK; ltos BgTkTbDgDQCNH | 137 ye shes SMBgTkTbDgDQCNH; chos nyid TkK | 138 kyis SMTkKBgDgDQH; kyi TkTbCN PHYLOGENETIC ANALYSIS 128

/rang gi mthu yis139 rig pa las/

/chos ni gzhan nas140 ’byung141 mi ’gyur142/[331]

<7> rgyu143 nyid rdo rje rkyen dang ’dra/

ma skyes pas na ’jigs pa144 med/[342]

<8> byis pas145 bslad pas ’khrul pa’i146 phyin/c

sman nyid sman pa147 tshol148 ba bzhin/[344]

<9> de dang der ni149 rnam150 brtsal na151/

/de las152 de bzhin de mi ’byung/[352]

<10> go153 ba’i yul na bde ba che/

/de nyid154 rnam dag ’jig rten yin/[353]

139 yis SMTkTb; yi TkKDgDQCNH; ’i Bg | 140 gzhan nas SMBgTkTb; om. TkK; gzhan las DgDQCNH | 141 ’byung SMBgTkTbDgDQCNH; ’byung bar TkK | 142 ’gyur SMBgTkTbDgDQCNH; ’gyur ro TkK | 143 rgyu SMBgTkTbDgDQCNH; rgyud TkK | 144 ’jigs pa SMTkK; ’jig pa BgTkTbDgDQCH; ’jig ba N | 145 byis pas SM; bus pa Bg; bus pas TkTbDgNH; bus bas DQC; om. TkK | 146 bslad pas ’khrul pa’i SMTkTbNH; ’khrul pas bslad pa’i Bg; bslad pas ’khrul ba’i DgDQC; om. TkK | 147 pa SMBgTkKTbDgDQCNH; pas Tk | 148 tshol SMTkKTkTbDgDQCNH; ’tshol Bg | 149 der ni SMTkKTkTbDgDQCNH; de ni Bg | 150 rnam SMTkTbDgDQCNH; rnams TkKBg | 151 brtsal na SMTkKBg; btsal na TkTbDgDCNH; btsal de Q 152 las SMTkTbDgDQCNH; la TkKBg | 153 go SMTkKTkTbDgDQCNH; ’gro BgTk | 154 de nyid SM; ’di ni TkKBgTkTbDgDQCNH c Instead of this line (byis pas bslad pas ’khrul pa’i phyin/) TkK has the following line: /thabs mkhas pa'i spyod yul yin/. PHYLOGENETIC ANALYSIS 129

<11> gcig155 ste rnam pa156 yongs kyis157 med/

/rnal ’byor nam mkha’i158 bya lam159 gnas/

/ma byung ma skyes160 snying po la/

sgros btags161 chos kun162 ga la yod/

/phyi nang gnyis ga163 phyi nyid nang164/

/zab mo’i165 cha shas rtogs166 yul med/[364]

<12> /de la tha tshigs167 phyi dang [364] nang/

rang168 bzhin phung po khams bzhin gnas/

/dus gsum ’di dang mi ’bral bas/

/tha tshig169 ming du btags170 pa med/[365]

<13> kun gyi mnyam nyid yin pa la171/

phyogs su lta bas172 grub173 pa med/[368]

<14> chos skus174 khyab pa’i175 gcig pa’i phyir/

155 gcig SMTkKTkTbDgDCNH; cig BgQ | 156 pa SMBgTkTbDgDQCNH; par TkK | 157 kyis SMTkKTkTbDgDQCH; kyi BgN | 158 nam mkha’i SMBgTkTbDgDQCNH; nam mkha’TkK | 159 lam SMTkKTkTbDgDQCNH; la Bg | 160 ma skyes SMTkTbDgDQHCN; mi skye TkKBg | 161 btags SMTkKTkTbDgDQCNH; btab Bg | 162 kun SMBgTkTbDgDQHCN; su TkK | 163 gnyis ga SM; gnyis ka TkKBgTkTbDgDQCNH | 164 nang SMBgTkTbDgDQCNH; na TkK | 165 mo’i SMBgTkTb; mo TkKDgDQCNH | 166 shas rtogs SMTkKTkTbDgDCNH; shas rtog Bg; shes rtogs Q | 167 tshigs SMQCN; tshig TkKBgTkTbDgDH | 168 rang SMBgTkTbDgDQCNH; de TkK | 169 tshig SMTkKBgTkTbDgDH; tshigs QCN | 170 btags SMTkKBgTbDgDCNH; brtags Tk; rtags Q 171 nyid yin pa la SMBgTkTbDgDQCNH; pa de nyid yin TkK | 172 lta bas SM; ltas pas TkK; bltas pas BgTkTbDgDQCNH | 173 grub SMTkKDgDQCNH; ’grub BgTkTb | 174 skus SMBgTkTbDgDQCNH; sku TkK | 175 pa’i SM; pas TkKBgTkTbDgDQCNH PHYLOGENETIC ANALYSIS 130

/che bas176 chen po’i177 rang bzhin gnas/[370]

<15> de bas de lung de nyid las178/d

de la179 de byar180 sgrib181 par ’gyur/

/de ltar de la de rtog na/

/de la de nyid grub pa med/[374]

<16> mchog tu gsang ba’i chos nyid ni/

/rna dbang gzhan las182 thong183 mi ’gyur/

/de bzhin lce’i184 dbang po185 kyang/

/de la brjod du rdul tsam186 med/[383]

<17> rang bzhin gcig187 tu ma nges pas/

/ji ltar188 brtags pa189 de ltar snang190/[385]

<18> bsam gtan nyid pas [412] bsam du med/

176 bas SMTkK; ba’i BgTkTbDgDQCNH | 177 po’i SMC; por TkKBgTkTb; po DgDQNH | 178 nyid las SM; nyid la BgTbDgDQCNH; snyed la Tk; om. TkK | 179 de la SMDgDQCNH; de bas TkK; de las BgTkTb | 180 de byar SM; de bya TkK; de sbyar BgTkTbDgDQCNH | 181 sgrib SMTkKBgTkTbDgDCNH; bsgrib Q | 182 gzhan las SMTkKBgTkDgDCNH; gzhan la Tb; bzhan las Q | 183 thong SM; thob TkK; ’byung Bg; thos TkTbDgDQCNH | 184 lce’i SMTkKBgQN; lce yi TkTbDgDQCH | 185 po SM; pos TkKBgTkTbDgDQCNH | 186 tsam SMTkKBgTkTbDgDQCH; rtsam N | 187 gcig SMTkKBgTkTbDgDCNH; cig Q | 188 ji ltar SMBgTkTbDgDQCNH; ci ltar TkK | 189 brtags pa SM; ltas pa TkKTk; gnas pa Bg; bltas pa TbDgDQCNH | 190 de ltar snang SMBgTkTbDgDQCNH; der snang ste TkK d This line (de bas de lung de nyid las/) is omitted in TkK. PHYLOGENETIC ANALYSIS 131

/ma bsams191 ma sbyangs chos bzhin192 du/

/rnam rtog193 nyid las skye194/[413]

<19> snang ba med pa’i chos nyid ni195/

/ma btsal196 bzhag pas197 bsgom198 pa yin/[414]

<20> /khyad par cir199 yang mtshungs pa la/

/las so200 zhes ni201 rnam par brtags/

/ci202 ste las kyi203 dbang gyur204 na/

/rang byung205 ye shes yod ma yin/[424]

<21> mngon sum206 pa dang rgyab nas byung207/

/de yang snang ba’i208 skyon yin te209/[433]

<22> ’phra210 ba’i sgo mor ming btags211 pas212/

191 bsam«s» SMTkKTkDgD; bsam BgTbQCNH | 192 bzhin SMBgTkTbDgDQCNH; nyid TkK | 193 rtog SMTkKTkTbDgDQCNH; rtogs Bg | 194 la skye SM; la ye shes skye TkKBgTkTbDgDQC; la ye shes skyes N; pas ye shes skyes H | 195 ni SMTkKBgDgDQCNH; ’di TkTb | 196 ma btsal SMTkTbDgDQCNH; mi rtsal TkK; ma brtsal Bg | 197 bzhag pas SMTkKBgTkDgDQCNH; gzhag pas Tb | 198 bsgom SMBgTkTbC; sgom TkKQ; bsgoms DgDNH | 199 cir SMBgTkTbDgDCNH; ji TkK; bcir Q | 200 so SMTkKTkTbDgDQCNH; so so Bg | 201 zhes ni SMBgTkTbDgDQCNH; zhes pa TkK | 202 ci SMBgTkTbDgDQCNH; ji TkK | 203 las kyi SMBgTkTbDgDQCNH; de’i TkK | 204 gyur SMTkK; ’gyur BgTkTbDgDQCNH | 205 byung SMTkKTkTb; ’byung BgDgDQCNH | 206 sum SMTkKTkTbDgDCNH; gsum BgQ | 207 byung SMTkKBgTkTb; ’byung DgDQCNH | 208 snang ba’i SM; rnam pa’i TkKBgTkTbDgDQCNH | 209 yin te SM; yin pa TkKBgTkTbDgDCNH; yin bas Q | 210 ’phra SMBg; phra TkKTkTbDgDQCNH | 211 btags SMTkKTkTbDgDC; ’dogs Bg; brtags QNH | 212 pas SM; nas TkK; te BgTkTbDgDQCNH PHYLOGENETIC ANALYSIS 132

/sems kyis213 dben pas214 lam tshol215 zhing/

/dgon pa’i rgyud du dben ’dzin pa216/

/brtags na rnam par rtog217 ’gyur bsgom218/[434]

<23> ’di219 la brtul zhugs220 drag shul kyis221/

/gnas su bya ba’i ming med de222/

a dang par ni rnam ldan na/

/rgyu ba’i223 bde ba bskyed par224 ’dod/[435]

<24> srid pa [435] gsum na sbyor ba225 yang/

/ming tsam sgyu226 mar snang ba ste227/

/’khor lo228 bsgyur229 ba’i gnas chen yang/

/sgyu mas230 sbyor231 ba’i bsti232 gnas yin/[436]

<25> byams pas233 don nyid rnam234 sbyangs235 phyir/

213 kyis SMTkTb; kyi TkKBgDgDQCNH | 214 pas SMTkKBg; pa’i TkTb; pa DgDQCNH | 215 tshol SMTkKTkTbDgDQCNH; ’tshol Bg | 216 ’dzin pa SMBg; ’dzin te TkK; ’dzin zhing TkTb; ’dzin cing DgDQCNH | 217 rtog SMBgTkTbDgDQCNH; rtogs TkK | 218 bsgom SMTkKBgTkTbDgDCNH; sgom Q | 219 ’di SMBgTkTbDgDQCNH; de TkK | 220 brtul zhugs SMTkKTkTbDgDQCH; brtul shugs Bg; btul zhugs N | 221 kyi«s» SM; gyis TkKBgTbDgDQCNH; gyi Tk | 222 de SMTkKBgTbDgDQCNH; te Tk | 223 rgyu ba’i SM; sgyu ma’i TkKBgTkTbDgDCNH; rgyu ma’i Q 224 bskyed par SM; skye bar TkK; ’byung bar BgTkTbDgDQCNH | 225 sbyor ba SMBgDQCNH; sbyor TkK; sbyong ba TkTbDg | 226 sgyu SMTkKTkTbDgDCNH; sgyur Bg; rgyu Q | 227 ste SMTkKBgTkTbDgDCNH; dang Q | 228 lo SMTkKTkTbDg; om. Bg; los DQCNH | 229 bsgyur SMBgTkDgH; sgyur TkKTbDQCN | 230 sgyu mas SM; sgyu ma TkKBgTkTbDgDCNH; rgyu ma Q 231 sbyor SMTkK; sbyong TkTbDgDQCNH; sbyang Bg | 232 bsti SMBgTkTbDgDQCNH; sti TkK 233 pas SMTkKBgTbDgDQCNH; pa Tk | 234 rnam SMTkTbDgDQCNH; rnams TkKBg | 235 sbyangs SMTkTbDgDQCNH; sbyong TkK; sbyang Bg PHYLOGENETIC ANALYSIS 133

/snying rje236 chen po237 cir mi mdzad238/[445]

<26> mnyam pas rim par bkod pa med/

/gcig las239 phyogs su240 bsngo241 dang bral/

/tshogs242 kyi brgyan dang243 bkram pa244 yang/

/rang bzhin gnas nas245 ’grim pa246 med/

/lhun gyis grub247 pas mi bsngo248 zhing/

/gdod nas dag pas bdud rtsi yin/[448]

<27> mi bzung249 gtong pas250 gtor251 ma yin/

/bya ba252 med pa’i253 las rnams te254/[448]

<28> kun spyod yon tan rnam lnga yang/

/chos nyid dbyings kyi255 rgyan zhes gsungs256/[452]

<29> rnam grangs brgya stong brag257 yangs pas258/

/gang ltar spyad259 kyang me tog260 skye/[452]

<30> bla ma mchod dang gtong261 ba dang/

/de bzhin bsod nams thams cad kyang/

/ma chags mi g.yo’i262 stobs med na/

/byas kyang263 ’ching ba chen por ’gyur/[456]

<31> sla264 zhing dka’ la265 sla266 phyir dka’/

mngon sum267 mi gnas kun tu268 khyab/

ming tsam269 ’di zhes bstan par ni/

/rdo rje sems dpa’270 ’tshon271 du med/[463]

<32> de ni snying po de bas na/

/de las de ’byung272 ngo mtshar che/[484]

<33> ngo mtshar rmad byung273 rol pa274 ’di/

/bya bral nam mkha’ ji bzhin te275/

/cir276 yang mi dmigs gti mug las/

/de ma dag277 tu rang278 las ’byung279/[484]

<34> yon tan kun dang280 stobs kyi rnams281/

/ji bzhin rtogs pa282 phra283 ba las284/

/de ma thag tu rang las ’byung285/[489]

275 ji bzhin te SMTkKBg; bzhin du gnas TkTbDgDQCNH | 276 cir SMTkTbDgDQCNH; ci TkKBg 277 dag SM; thag TkKBgTkTbDgDQCNH | 278 rang SMBgTkTbDgDQCNH; nga TkK | 279 ’byung SMBgTkTb; byung TkKDgDQCNH | 280 dang SMTkKBgTkTbDgDCNH; ldan Q | 281 kyi rnams SMTkKBgDgDQCNH; rnams kyi TkTb | 282 rtogs pa SMBgTkTbDgDQC; rtogs pa’i TkK; rtog pa NH | 283 phra SMTkKTkTbDgDQCNH; ’phra Bg | 284 las SMBg; yis TkKTkTbDgDQCNH | 285 rang las ’byung SMTkTbDgDQCNH; ’byung bar ’gyur TkK; rang la ’byung Bg

INTRODUCTION TO THE CRITICAL EDITIONS 136

CHAPTER 6

Introduction to the Critical Editions

The theoretical problems associated with the process of establishing a critical edition are many and complex, but all, in one way or another, are related to the purpose of the edition, and the extant witnesses that one has access to and can rely upon in order to establish that edition. Sometimes, the aim is too high, as when a textual critic attempts to reconstruct an “archetype,” based on witnesses that are removed from the presumed “original” text by several centuries. At other times, a critic may aim to reconstruct an earlier version of a text, and end up “establishing” a text that is critically sound but literally unreadable, leading to a translation that is beyond comprehension.

The purpose of the present study is to establish a reliable critical edition of the five “earlier” Sems sde texts. Its aim is two-fold: (1) contribute to the study of the Sems sde textual transmission, by providing data for a phylogenetic analysis

(see chapter 5); and (2) contribute to the transmission and dissemination of the rDzogs chen teachings, by establishing readable editions that can lead to accessible translations (see chapter 8). The editions are aimed to be as comprehensive as possible; they take into consideration all extant witnesses (that are accessible) and carefully record all variant readings (whether recensional or transmissional). Some

INTRODUCTION TO THE CRITICAL EDITIONS 137 editions (such as Co ne, sDe dge and lHa sa), may provide little or no information about the earlier strata of the texts, however, they may shed some light on the

Tibetan orthography and how spelling conventions have changed over time.

In editing the texts, I have not made any conjectural emendations. However,

I did “restore” what I considered to be the correct reading, the “best” reading—if I had a compelling reason for doing so, and considered that some of the readings defy reasonable expectations—based on careful examination and comparison of all textual variants. In this sense, it can be said that the editions are eclectic editions, sometimes preferring one textual tradition, and sometimes another. Given the large number of variants, it is not always easy to make editorial decisions, and it is quite possible to select other readings, still, in most cases the selection is self-evident. In terms of spelling and punctuation, I have adopted the modernized form that can be found in relatively later editions, such as the sDe dge.

The critical editions of the five Sems sde texts (Rig pa’i khu byug, rTsal chen sprug pa, Khyung chen lding ba, rDo la gser zhun, and Mi nub pa’i rgyal mtshan) are based on fourteen extant witnesses. With the exception of the rDo la gser zhun, the texts are found as independent texts in the Bai ro’i rgyud ’bum (Bg,

14th c.) collection and the gTing skyes (TkK, 18th c.) edition of the rNying ma rgyud ’bum canon, and as citations in the bSam gtan mig sgron. All texts are also found as chapters of the Kun byed rgyal po’i mdo in the sDe dge (Dg, 1794-98), mTshams brag (Tb, 1728-48), and gTing skyes (Tk, 18th c.) editions of the rNying

INTRODUCTION TO THE CRITICAL EDITIONS 138 ma rgyud ’bum canon; and the ’Jang sa tham/Lithang (J, 1608-1621), Kangxi

Taiwan (K, 1669), Basgo (Ba, 17th c.), Co ne (C, 1721-31), Qianlong Peking (Q,

1737), sNar thang (N, 18th c.), sDe dge (D, 18th c.), and lHa sa (H, 1934) editions of the bKa’ ’gyur canon.

6.1 “Recensional” and “transmissional” readings

All editions contain readings, to a lesser or a greater extent, that are either singular or are variant—shared by two or more witnesses—and thus betray a family resemblance. Furthermore, a distinction can also be made between “recensional” and “transmissional” readings, which can be either single or variant. According to

Harrison (1992), recensional readings, on the one hand, reveal “extensive and deliberate editorial changes to the text,” and involve “alterations to the wording” or

“the substitution of terminology.” Transmissional readings, on the other hand, are minor and most often trivial in scope “resulting from scribal lapses or casual attempts to improve or modernize the text” (xxv, xxxi). For the Sems sde texts, most variant readings are transmissional in character, and even though recensional variants can be found, most recensional readings are single readings.

As most bKa’ ’gyur editions “derive” from the so-called “Old sNar thang” manuscript edition, the Tshal pa bKa’ ’gyur (1351), it is inevitable that many editions would share a great number of readings. The Tshal pa continued to be revised after it was compiled, and was used as the base for further editions, such as

INTRODUCTION TO THE CRITICAL EDITIONS 139 the Yongle (1410) and the ’Jang sa tham/Lithang. Furthermore, as Qianlong and

Kangxi are impressions of the Yongle edition, Co ne is a copy of the ’Jang sa tham edition, and sDe dge is based on the ’Jang sa tham, the variant readings clearly indicate group relations between the Tshal pa based witnesses. The same can be said about the rNying ma rgyud ’bum editions that carry the Kun byed rgyal po text.

What is striking is that the Basgo bKa’ ’gyur edition contains a great number of single readings, such as, sbyin khyab instead of ltar mnyam (9.4), rnyeg pa ’dra instead of bsnyabs pa bzhin (20.4), chos sku’i instead of sku yi (32.1), ye shes de nyid ni instead of pas ni ye shes te (32.2), rol ba instead of nam mkha’ (44.3), and

’chi ba myed par instead of ’ching ba chen por (54.4), to cite just a few examples

(Mi nub pa’i rgyal mtshan). It also seems to have a total disregard for “metrics” when it comes to the number of syllables in a line. For example, in 33.2 it omits kun and ends up with six syllables instead of seven (as the standard form), in 33.3 omits med ces and in 34.1 drag shul and ends with five syllables, and in 34.2 it uses bya’i (usually counted as one syllable) for bya ba’i and ends with six syllables (Mi nub pa’i rgyal mtshan). Furthermore, Basgo even sides with Bg, TkK and SM, and shares variant readings with these witnesses of the freestanding Sems sde texts, such as reading rang bzhin for mnyam par (10.3, Khyung chen lding ba), and yangs pas for yas pa (39.1, Mi nub pa’i rgyal mtshan).

The extant Basgo edition is fragmentary, and we do not know much about the edition or its line of transmission. Bruno (2009) points out that the Basgo

INTRODUCTION TO THE CRITICAL EDITIONS 140 manuscript belongs to a group of newly discovered collections, originating from

Western Tibet, that are “not bKa’ ’gyurs per se, but are rather proto-bKa’ ’gyurs.”

These proto-bKa’ ’gyur collections include: two bKa’ ’gyurs in Basgo (Ba sgo,

Ladakh), one incomplete bKa’ ’gyur in Chemde (lCe bde; Ladakh), one collection in Hemis (He mis, Ladakh), one collection in Charang (Kinnaur), one collection in

Rangdum (Zanskar), and one collection in Stongde (sTong bde, Zanskar) (5-7).

Apple (2014), who in his critical edition of the Dunhuang Tibetan version of the

Mañjuśrīvihārasūtra includes for the first time both Basgo and Hemis editions in a critical study (stemmatic and phylogenetic analysis), suggests that the two editions are closely affiliated, and belong to a line of transmission that may be based on the collection that at one time was kept at Tholing (mTho lding). They are independent of the Tshal pa textual tradition and contain readings that are older than that of the

Tshal pa based witnesses (296-300; on Tholing see also De Rossi Filibeck 2007). If the Basgo and the Hemis editions do indeed preserve older readings, this would explain why the Basgo edition has so many single and variant readings—the extant

Tshal pa based editions are known to be highly revised and edited.

It is important to note that, sometimes, transmissional readings may be as important as the recensional readings. Reversed words (spoonerism) and lines omitted (eye-skip), if they do make sense or are accepted as variants, can be either transmissional single readings or recensional variant readings. Thus, in the Khyung chen lding ba (13.1) we have some words reversed: while Bg, SM, and TkK read

INTRODUCTION TO THE CRITICAL EDITIONS 141

’byung ba legs pa che, all other editions read legs pa ’byung ba che. In this case, a onetime transmissional single reading has become a variant reading. In the rTsal chen sprugs pa we also have a case of spoonerism when Bg and TkK read slob pon tshad mar ma gyur instead of reading tshad mar ma gyur slop dpon, as do all the other editions.

The Bg and TkK editions preserve the five Sems sde texts as independent texts. Do they also preserve the older readings? It is hard to say as sometime it would help if we knew what was the “standard,” that is, the ideal form that the anonymous authors (or redactors) were aiming for. For instance, in rTsal chen sprugs pa, they omit ye nas in 1.1 and 1.2, and dang ni in 2.3, and read gyis instead of dang gnyen po in 2.4. It is unclear as to which edition emended the verse, the Bg and TkK emended the lines from nine-syllables to seven-syllable, or the other editions emended the lines from seven-syllables to nine-syllables. This highlights not only the close relationship between the two editions (Bg and TkK), but also the idiosyncratic structure of the rTsal chen sprugs pa: some of its verses have lines of nine-syllables (1, 2 and 5), while others are highly irregular, ranging from nine to eleven and thirteen-syllables (3 and 4). There are also many “peculiar” transmissional readings that can be found in the extant editions, which might be of interest to those who specialize in Tibetan linguistics, and they include: i)

Orthographic Peculiarities; ii) Graphic Variants; iii) and Abbreviations.

INTRODUCTION TO THE CRITICAL EDITIONS 142

i. Orthographic Peculiarities - refer to orthographic features, that may or may not be archaic (see Zimmermann 2002: 207), such ya btags and the reversed gi gu. Ya tags appear in words that begin with m- followed by i, e, or o, such as, dmyigs (དགས), myig (ག), myin (ན), smyig (སག), myed myi (ད་ ), and myongs (ངས) for dmigs (དགས), mig (ག), min (ན), smig (སག), med mi (ད་), and mongs (ངས) (Ba). A reversed gi gu “originally” may have been used when there was a lack of space (e.g. at the end of a line) for a standard gi gu, and in later editions it appears as a peculiarity, such as, gyīs (ྀས), mtsho’ī (མ), nyīd (ད), and lī () (Q). Other “irregularities” may include the use of dra dag, gyurd (རད) for gyur (ར), and zind (ཟིནད) for zin (ཟིན), as well as unusual letter combinations, ci () for ji (), and ces (ས) for zhes (ས) (Ba).

ii. Graphic Variants - usually include the subjoined s and the anusvāra-like abbreviation for m, and refer to the placement of these letters as subscripts and superscripts. Subjoined s can be found in most editions, and in several forms, such as, bas () and rtags () (Ba); gis () and rtogs () (Bg); chags (ཆ), las (ལྶ), ’joms (འ), nyams (ཉ), and stobs () (C); chos (᮶ོ) and khams (ཁ) (N); gshes (གཤྶེ), shes (ཤྶེ), bsgoms (བ), lugs (), rnams (), and rtogs () (Q); and dmigs (ད) (CK) and mtshungs (མངྶ) (BgTbQ). The abbreviated letter m is not as common as the subjoined letter s, but it does appear in

INTRODUCTION TO THE CRITICAL EDITIONS 143

some editions as gsuṃ (གཾ) for gsum (གམ) (C); laṃ () for lam (ལམ) and rnaṃs (སཾ) for rnams (མས) (Bg); and suṃ (ཾ) for sum (མ) (SM).

iii. Abbreviations - may include two different kinds of contractions: bsdys yig and bskungs yig (Orofino 1994: 31-32). The first type, bsdys yig, refers to the practice of eliminating a consonant where the final consonant is the same as the first consonant of the following syllable. Some abbreviations of this type are: rtogyur (ར) for rtog ’gyur (ག་འར) (Bg); gnasu (གན) for gnas su (གནས་), phyogsu (ག) for phyogs su (གས་), rjesu () for rjes su (ས་), and yongsu (ང) for yongs su (ངས་) (N); sumthong (མང) for sum mthong (མ་མང) (Tb); namkha’i (ནམཁ) for nam mkha’i (ནམ་མཁ) (BgN) and namkha’ (ནམཁའ) for nam mkha’ (ནམ་མཁའ) (BaBgN). The second type, bskungs yig, is a more complex form of abbreviation where letters are dropped, as in the following examples: bya’i () for bya ba’i (་བ) and bder (བར) for bde bar (བ་བར) (Ba); rtogyur (ར) for rtog ’gyur (ག་འར) and mes (ས) for me yis (་ས) (Bg); and thaṃd (ཐམད) for thams cad (ཐམས་ཅད) (N).

6.2 Explanation of the layout

In choosing the “layout” for the critical editions, my aim has been to present the texts in a legible and reader-friendly form, and the critical apparatus in a

INTRODUCTION TO THE CRITICAL EDITIONS 144

“visible” position, for easy access and future reference, so that the apparatus can be read together with the main text. It is true that some readers may be more interested in a text than in the apparatus, however, for a scholar interested in critical editions, the appended apparatus may be more important than the text itself. The editions depart from the original Tibetan format in several ways: the use of Latin script, the

Western-style paragraphing and layout of verses, and the critical apparatus.

Personally, I tend to agree with Harrison’s (1992: l-li) assessment that in normal circumstances it would be preferable to print a Tibetan text in a Tibetan script.

Unfortunately, technical problems associated with the creation of a data set that can be used in phylogenetic programs have made the use of the Tibetan script impractical. The transliteration is based on the Wylie system, as described by

Wylie (1959) in his ground-breaking article, “A Standard of Tibetan

Transcription.” Also for practical reasons (reference and editing purposes), the texts are divided into paragraphs (rDo la gser zhun) and verses (Rig pa’i khu byug, rTsal chen sprug pa, Khyung chen lding ba, and Mi nub pa’i rgyal mtshan). Each number of a paragraph or verse is highlighted in bold and appears within square brackets

(e.g. [4]), to indicate that it is not part of the text.

The critical apparatus and the notes are placed at the bottom of the page, beneath the main text. The apparatus is divided into four layers: 1) variant readings;

2) single readings; 3) orthography and punctuation; and 4) pagination. In the “main apparatus,” section one, are included all variant readings, recensional as well as

INTRODUCTION TO THE CRITICAL EDITIONS 145 transmissional, that are shared by more than one manuscript or xylograph edition.

In order to keep the apparatus as concise as possible, I follow the lead of Harrison

(1992, liii) and take the Tibetan syllable as the unit of reference, not the whole word. To illustrate how the variants are cited and listed in the main apparatus, a good example can be found in the Mi nub pa’i rgyal mtshan (5.1):

10 dka’] BaBgCDDgHJNTbTkTkK; bka’ KQ

The number (10) refers to the line number that is suggested by the numbers located at the right margin of the critical text—to keep the page “clean,” only numbers for line 4, 8, and 12 are printed. The number is followed by the lemmata, in bold type, that cites the main text (dka’), a square bracket (]), and all textual witnesses that support the main text (BaBgCDDgHJNTbTkTkK). A semicolon separates the list of witnesses from the first variant (bka’), which in turn is followed by the witnesses that share this variant reading (KQ). If there is more than one variant for a reading, the process continues until all variants (whether single or not) and all witnesses

(thirteen, or fourteen when we have a citation from SM) are accounted for. All witnesses are listed in alphabetical order by letter code, and the ones that do not transmit a variant are listed in italics type as omitting it (om.).

Sections two to four form the “supplementary apparatus.” Section two includes all singular readings (lectiones singulares), that is, readings that are not shared as variants with other witnesses, and consists primarily in simple errors,

INTRODUCTION TO THE CRITICAL EDITIONS 146 omissions, insertions or orthographical variations (scribal attempts to improve the text by adopting more modern spellings). Section three includes orthographic peculiarities, graphic variants, abbreviations and punctuation (the place of a shad).

Section four refers to pagination, and references are made in the usual manner by folio, verso (a) or recto (b), and line. For example, Ba215a1, refers to the Basgo manuscript of the bKa’ ’gyur edition, folio two hundred fifteen, verso (a), and line

1. Blank spaces within and between the lines have been ignored, and the position of the corrections (above the line, below the line, or on the margins) have not been noted. The “notes,” located at the bottom of the page, include readings that are not considered to be part of the texts but can be found in most editions. All variants as found in these readings are noted in full. Chapter 7, therefore, will establish a critical edition of the five earlier rDzogs chen Sems sde texts, and chapter 8 will provide an English translation based on the critically edited texts.

RIG PA’I KHU BYUG 147

CHAPTER 7

Critical Editions

7.1 The Critical Edition of the Rig pa’i khu byug

[1] /sna tshogs rang bzhin mi gnyis kyang /

/cha shas nyid du spros dang bral/

[2] /ji bzhin pa zhes mi rtog kyang /

/rnam par snang mdzad kun tu bzang / 4

[3] /zin pas rtsol ba’i nad spangs te/

/lhun gyis gnas pas gzhag pa yin/

4 tu] BaBgHNTbTkK; du CDDgJKQTk 5 rtsol] BaCDDgHJNSMTbTkTkK; brtsol BgQ; bcol K 6 pas] BgCDDgHJKNQSMTbTk; pa BaTkK || gzhag] DDgJTbTkTkK; bzhag BaBgCHKNQSM

3 /ji] /ci Ba | pa] ba Bg 4 kun] ku Ba 5 ba’i] pa’i Bg | nad] nang K | te/] ste/ Ba 6 gnas] grup Tb

3 mi] myi Ba 5 /zin] zind Ba

1 /sna] [Ba216b2]; [Bg306-2]; [C43a6]; [D38b4]; [Dg39b4]; [H54b2]; [J39a2]; [K53b1]; [N53a7]; [Q36b6]; [Tb57a3]; [Tk49a1]; [TkK209a2] 3 rtog] [N53b1] 6 yin/] [Ba216b3]; [Bg306-3]; [C43a7]; [D38b5]; [Dg39b5]; [H54b3]; [J39a3]; [K53b3]; [N53b1]; [Q36b7]; [Tb57a5]; [Tk49a2]; [TkK209a3]

1 /sna Bg and TkK read: //rgya gar skad du/ san ti dar pa/ bod skad du/ rdzogs pa chen po sa gcig pa/ sva sti dpal gyi dpal/ bcom ldan ’das dpal kun tu bzang po sku gsung thugs rdo rje bde ba chen po lhun gyis rdzogs pa la phyag ’tshal lo/ (Bg reads san ti ma ha dar pa/ instead of san ti dar pa/ ). All other editions (BaBgCDDgHJKNSMTbTkQ) read: //de nas byang chub kyi sems kun byed rgyal po des kun byed nyid kyi rang bzhin bya med rdzogs pa lhun gyis grub pa ’di gsungs so/ /kye sems dpa’ chen po nyon cig (CDDgHJKNQ om. nyid and Ba om. lhun gyis grub pa). 6 yin/ Bg and TkK read: /byang chub kyi sems rig pa khu byug rdzogs so// (Bg om. kyi and reads rig as rigs). All other editions (BaCDDgHJKNSMTbTkQ) read: /zhes gsungs so/ /byang chub kyi sems kun byed rgyal po las rdo rje tshig drug gi le’u ste sum cu rtsa gcig pa’o// (Ba and Tk read /zhes as /ces; Tb and Tk read drug pa'i instead of drug gi; BaJK read cu as bcu; and BaCHJKNQTk om. rtsa). RTSAL CHEN SPRUGS PA 148

7.2 The Critical Edition of the rTsal chen sprugs pa

[1] /phyi dang nang kun ye nas chos kyi dbyings/

/ye nas rnam par dag pa’i spyod yul la/

/sangs rgyas dang ni sems can gnyis su med/

/lam dang gnyen pos bcos su ga la yod/ 4

[2] /bsgrub med rtsal sprugs smon pa med pas na/

/bya med lhun rdzogs snga ma de nyid don/

/rtog spyod rnam dag gnyis med chos dbyings la/

/log rtog byis pa’i spyod pas ci zhig bslus/ 8

1 ye nas] BaCDDgHJKNQTbTk; om. BgTkK 2 /ye nas] BaCDDgHJKNQTbTk; om. BgTkK 3 dang ni] CDDgHJKNQTbTk; om. BaBgTkK || su med/] BaCDDgHJKNQTbTk; med de/ BgTkK 4 dang… p os] BaCDDgHJKNQTk; la gnyen pos Tb; gyis BgTkK 5 /bsgrub med] BaBgCDDgHJKNQ; /sgrub med TkTkK; /slub byed SM || pas na/] BaCDgHJKNQTbTk; pa’i/ Bg; bas na/ D; pa yi/ SMTkK 6 snga] BgCDDgHJNQSMTbTkTkK; lnga BaK || don/] CDDgHJKNQSMTbTk; de/ Ba; do/ BgTkK 7 spyod] BaBgCDDgHJKNQ; dpyod SMTbTkTkK 8 bslus/] BaCDDgHJKNQ; bslu/ BgSMTb; slu/ TkTkK

1 kun] na K; du Q 4 /lam] /las Ba 5 rtsal] brtsal Ba | pa] pas Ba; ba D 6 /bya] /byar SM 7 /rtog] /rtogs Bg 8 /log rtog] /rtog spyod Ba; /blos rtog Bg | spyod] spyos Q | ci] ’ga’ SM

3 med/] myed/ Ba 5 med1] myed Ba | smon pa] {smon pa} Bg | med2] myed Ba 6 med] myed Ba 7 med] myed Ba

1 /phyi] [Ba211a1]; [Bg306-4]; [C37b8]; [D33b2]; [Dg34b2]; [H47b1]; [J33b6]; [K46b4]; [N46b1]; [Q32a5]; [Tb50a2]; [Tk42b3]; [TkK211a3] | chos] [C38a1] 3 /sangs] [Bg307-1]

1 /phyi Bg and TkK read: //padma’i spyan la phyag ’tshal lo/. All other editions (BaCDDgHJKNSMTbTkQ) read: //de nas byang chub kyi sems kun byed rgyal po des nyid kyi gdams ngag phog pa dang mthun pa’i lung chen po rim gyis bstan te/ /kye sems dpa’ chen po nyon cig /chos rnams thams cad nga las byung ba na/ /ji ltar snang ba’i chos rnams thams cad kun/ /ye nas rnam dag dbyings kyi lung du bstan/ (BaJK read gdam and N reads bdag instead of gdams; J and N read ’thun instead of mthun; Tk reads rims kyis and K ram kyis instead of rim gyis; C and Tb read byung bas while Ba and Tk byung pas instead of byung ba. Ba has several single readings—it reads chen po’i instead of chen po, ba ni instead of ba na, ci instead of ji, and dbyings kyis instead of dbyings kyi. It also om. te in bstan te and rnams in chos rnams). RTSAL CHEN SPRUGS PA 149

[3] /gnyis med bde chen sems can kun spyod la/

/’khrul pas nor ba’i lam du brtags pa yang /

/gong du bstan pa’i lam chen de la mi gnyis pas/

/mnyam par shes pas sangs rgyas kun gyi rje/ 4

[4] /nga dang bdag tu rtog pa mu stegs ’khrul pa’i lam/

/byis pas bslus pas chos spyod rtog pa’i lam la btsud pa ni/

/phyin pa’i dus med rtogs pa’i dus med pas/

/chos nyid chos tshol rnyed par ga la ’gyur/ 8

[5] /tshad mar ma gyur slob dpon spre ’dra’i lung /

/bstan dang lam nor rtog pas zin ta re/

/de bas gser la nag tshur gtong ba yi/

2 ba’i] BaDDgQTb; pa’i BgCHJKNTkTkK || brtags] CDDgJTb; bstan BgHKNQ; btags BaTkTkK 3 de la] CDDgHJKNQ; de las BgTbTkTkK; om. Ba 5 stegs] BaBgCDDgHJTbTkTkK; rtegs N; steg KQ 6 bslus] BgCDDgHJKNQTb; slus TkTkK; om. Ba || pas2] CDDgHJNTbTk; pa TkK; pa’i BgKQ; om. Ba 7 pa’i1] BaBgCHJKNQTbTkTkK; ba’i DDg || dus1] BaCDDgHJKNQ; lam BgTbTkTkK || pas/] BaCDDgHJKNQ; te/ Tk; de/ BgTbTkK 9 /tshad… d pon] CDDgHJKNQTbTk; /tshad mar gyurd te slob dpon Ba; /slob dpon tshad mar ma gyur BgTkK 10 /bstan] BaCDDgHJKNQ; /rten BgTbTk; /brten TkK || rtog] CDDgHJKNQTbTk; rtogs BaBgTkK 11 ba yi/] CDDgHJKNQTb; pa’i Bg; pa yi/ Tk; ba’i/ BaTkK

1 med] myed Ba 3 mi] myi Ba 4 shes] shes Q 7 med1] myed Ba | med2] myed Ba 11 /de] de Bg

6 /byis] [J34a1] 8 tshol] [K47a1] 9 slob] [Q32b1] 11 /de] [Tk43a1] RTSAL CHEN SPRUGS PA 150

/slob dpon yang dag ston ba’i rin chen mdzod/

/rin thang dpag tu med pas blu bar ’os/

1 yang… b a’i] BaCDDgHJKNQTbTk; om. BgTkK || ba’i] DQ; pa’i BaCDgHJKNTbTk; om. BgTkK || mdzod/] CDDgHJKNQTbTk; mdzad/ Ba; de/ Bg; te/ TkK 2 /rin] BaBgCDDgHTbTkTkK; /ring JKNQ || pas] BaCDDgHJKNQTbTk; pa’i rin gyis BgTkK || blu] CDDgHJKQ; bslu BaBgNTb; slu TkTkK

1 /slob] /blo K 2 thang] ’thang Ba | bar ’os/] ba’os/ Ba

1 /slob] slob BgTkK 2 med] myed Ba | ’os/] ’os Q

1 mdzod/] [TkK211b1] 2 thang] [Tb50b1] | ’os/] [Ba211a6]; [Bg307-5]; [C38a5]; [D33b7]; [Dg34b7]; [H47b7]; [J34a2]; [K47a2]; [N46b6]; [Q32b1]; [Tb50b1]; [Tk43a1]; [TkK211b1]

2 ’os/ TkK reads: /rtsal chen sprugs pa rdzogs so//, while Bg reads: /byang chub kyi sems brtsal chen sprugs pa/ /rdzogs sho//. Tb and Tk read: /shes gsungs so/ /bya med rdzogs pa’i lung bcu las / bya med rnam dag dbyings kyi lung bstan pa’i le’u ste rtsa bdun pa’o//. The other editions (BaCDDgHJKNQ) read: /zhes gsungs so/ /byang chub kyi sems kun byed rgyal po las bya med rdzogs pa rnam dag dbyings kyi lung bstan pa’i le’u ste rtsa bdun pa’o// (CJKNQ read shes and Ba reads ces instead of zhes; Ba reads rdzogs pa’i lung and Q rdzogs ba instead of rdzogs pa; Ba om. bstan; and K and Q read bstan as stan). KHYUNG CHEN LDING BA 151

7.3 The Critical Edition of the Khyung chen lding ba

[1] /mi gnas dmigs pa’i yul med mi rtog chos kyi lam/

/bsngo ba’i cha shas phra mo’i yul la byung ba ste/

/rtog sgom chos kyi sku ni khyad par don med pas/

/rang ’byung ye shes mi rtog kun tu ji bzhin gnas/ 4

[2] /bya bral yul la mi gnas gnyen pos bcos su med/

/yan lag chos kyi dbang gis snying po’i yul tshol zhing /

/rnam pa cir yang mi rtog thabs kyis de la rol/

/snying po rang bzhin ’byung bas chos sku gzhan na med/ 8

1 rtog] BaCDDgHJNTbTkTkK; rtogs BgKQ 2 /bsngo] BaBgCDDgHJNTbTkTkK; /sngo KQ || mo’i] BaCDDgFHJKNQTk; ba’i BgTkK; ma’i Tb || la] CDDgHJKNQ; las BaBgTbTkTkK || byung ba] BaCDDgHJKNQTb; ’byung ba Bg; byung pa TkTkK 3 sgom] DDgHKNQTbTkTkK; bsgom BaBgCJ 4 ’byung] BaBgCDDgHJKNQTk; byung TbSMTkK 7 cir] BgCDDgHJNTbTkTkK; ci Ba; bcir KQ

2 ba’i] pa’i N | phra] ’phra Bg 4 mi rtog kun tu ji bzhin] kun tu ci zhig Ba; mi rtog kun tu rang bzhin Bg | tu] du Tk 5 pos] por Ba 6 /yan lag] /lag yan Bg | kyi] kyis Bg | snying po’i] gnyen po Tk; snying po TkK 7 kyis] kyi Bg 8 po] por Ba; bo Q

1 /mi] [Ba206a5]; [Bg308-1]; [C33b2]; [D29a7]; [Dg30a7]; [H41b5]; [J29a8]; [K41a5]; [N40b6]; [Q28b1]; [Tb44a4]; [Tk37a7]; [TkK209a4] 3 sku] [D29b1]; [J29b1]; [Dg30b1] 4 /rang] [Tk37b1] 7 rtog] [N41a1]

1 /mi Bg and TkK read: /bcom ldan ’das ’jam pa'i dbyangs la phyag ’tshal lo/ (Bg reads dpal instead of pa'i). All other editions (BaCDDgHJKNQTbTk) read: //de nas byang chub kyi sems kun byed rgyal po des sems rnal du gzhag pa’i phyir bya med rdzogs pa’i lung khungs chen po ’di gsungs so/ /kye sems dpa’ chen po rdo rje khyod nyon cig /kye bya med yul ni mi gnas nam mkha’ ’dra/ (Tb om. sems in byang chub kyi sems; Ba reads rgyal po las instead of rgyal po des, and bzhag pa'i instead of gzhag pa'i; HKNQ om. chen po in sems dpa' chen po, and Ba om. rdo rje khyod). KHYUNG CHEN LDING BA 152

[3] /rdul phran gcig la gtses shing phyogs bcu kun dang bral/

/mi gnas don nyid legs pa’i ye shes rang ’byung ste/

/mngon sum kun la mi rtog yangs pa’i snying po la/

/rnam dag la ma der zhugs nas mnyam pa’i rgyal po thob/ 4

[4] /mi ’gyur ’gyur ba med pas chags pa’i gnas med de/

/de bzhin bzung ba’i yul med sems kyi gnas kyang med/

/mngon sum thob par ’dod nas rgyu la rtag tu sgom/

/sgom chags bde ba de la mnyam pa’i don mi ’byung / 8

[5] /sku gcig kun la khyab pas bsnan pa’i chos med de/

/mtha’ la thug pa med pas dbyings la bri ba med/

/chos kyi rol pas nyams phag lhag pa’i gnas med de/

1 la] BgCDDgHJKNQ; las TbTkTkK; om. Ba || gtses] CDDgHJN; ces Ba; gses BgTbTk; brtsegs TkK; btses KQ || bcu] BaCDDgHJKNQTkK; cha BgTbTk 2 ’byung] BaBgCDDgHJKNQ; byung TbTkTkK 3 sum] BaBgCDDgHTbTkTkK; gsum JKNQ || la] BaCDDgHJKNQ; tu BgTbTkK; du Tk 4 la ma] DDg; lam chen Ba; lam BgCHJKNQTbTkTkK || der] BaCDDgHJKNQTbTk; ’dir BgTkK || nas] BaCDDgHJKNQTbTk; pas ni Bg; pas TkK 6 bzung] BaBgCDDgHJN; gzung KQTkTbTkK || sems] BaCDDgHJKNQTbTk; chos BgTkK 7 sum] BgCDDgHJQSMTbTkTkK; gsum BaKN || nas] BaCDDgHJKNQTb; pas BgSMTkK; na Tk || sgom/] CDDgHJKNQ; bsgom/ BaBgSMTbTkTkK 8 /sgom] DDgHNTkK; /bsgom BaBgCJKQSMTbTk || la] BgCDDgHJKNQTkK; las BaSMTbTk 9 la] BaCDDgHJKNQ; tu BgTbTkK; du Tk 10 pas] BaCDDgHJKNQTk; cing BgTbTkK 11 nyams] CDDgHJKNQ; mnyams Ba; nyam BgTbTkTkK || phag] CDDgHJKNTbTk; la Ba; bag BgTkK; pag Q

1 gcig] cig Ba | shing] om. Ba 2 pa’i] ba’i N 3 yangs] spongs Ba; yongs N 4 dag] ’dag Bg 5 ’gyur2] gyur K pa’i] pa Ba 6 ba’i] pa’i Ba 7 par] bar K; dang N 8 mnyam] nyams SM 9 gcig] cig Ba | pas] bas N | bsnan] snod Ba; mnan Bg; bsman N; bsnun TkK 10 bri] phri C | ba] pa D 11 kyi] nyid TkK | pas] pa Bg; bas N | lhag] dpag Ba

1 shing] [H42a1] 3 /mngon] [K41b1] 4 thob/] [TkK209b1] 5 med1] [Tb44b1] | pa’i] [Ba206b1] KHYUNG CHEN LDING BA 153

/rang ’byung che ba’i yul la kun tu ji bzhin gnas/

[6] /ngo mtshar mthong ba’i yul med de la mthong ba’i mig

/gang la thos par mi rigs snyad pa kun dang bral/

/chos dang chos min rtag tu ’dres shing mtshungs/ 4

/don dam chos shes lhag pa’i gnas shes bshad pa’ang med/

[7] /’khrul dang byang chub lam de rtog pas thob pa’ang med/

/rang ’byung ye shes nyid kyang tshig gi mtha’ dang bral/

/ye nas rgyas pa’i bdag nyid mngon sum gnas pa la/ 8

/grib ma gzugs kyi dpe bzhin de la rtog pa ’byung /

[8] /med de med pa ma yin med pa’i snying po ’byung /

1 ’byung] BgCDDgHJKNQ; bzhin Ba; byung TbTkTkK || tu] BgHKNQTbTkK; du BaCDDgJTk 2 mig] BaCDDgHJKNQTbTk; dmig BgTkK 3 snyad] CDDgHJNQTk; snyam Ba; bsnyad BgTb; snyed K; rnyed TkK 4 min] BaBgCDDgHJKNQTbTk; nyid SMTkK || mtshungs/] BaCDDgHJKNQTbTk; dbyer med la/ BgSMTkK 5 shes1] CDDgHJKNQSM; zhes BaBgTbTkTkK || lhag … shes2] CDDgHJKNQ; lhag par BgSMTkK; lhag pa’i gnas zhes BaTb; lhag pa’i gnas Tk || bshad pa’ang] BaCDDgHJKNQTbTk; bshad pa SMBgTkK 6 de] BaBgDDgHJKNQSMTbTk; med C; ste TkK || pa’ang] CDDgHJKNQTbTk; pa yang Ba; pa BgSMTkK 7 ’byung] BaBgCDDgHJKNQTk; byung SMTbTkK 8 rgyas pa’i] CDDgJKNQTbTk; sangs rgyas pa’i Ba; sangs rgyas BgSMTkK; rgyal ba’i H || sum] BgCDDgHJQSMTbTkTkK; gsum BaKN

1 la] [Bg309] 4 /chos] [C34a1] 9 kyi] [Tk38a1] 10 po] [N41b1] | ’byung] [H42b1] KHYUNG CHEN LDING BA 154

/stong ste stong pa ma yin stong pa’i yul la gnas/

/nam mkha’i rang bzhin de la ’byung ba’i dran pa skye/

/bya bral zin pa’i bde ba ’dod pa med pas len/

[9] /bza’ ba’i yul med de la ye shes ’byung ba rtsom/ 4

/sngon gyi drang srong rjes su chags pa’i yid kyang ’jog

/rnam rtsol ’bad pa’i gdung bas kun tu zin ta re/

/rang bzhin lam der zhugs pas thams cad mkhyen pa yi/

[10] /ji bzhin pa zhes rnam rtog de phyir rtog bsgom pa/ 8

/bde ba chen po ’dod pas chags pa’i nad yin te/

/mi g.yo mnyam par gnas pas sman chen ma byas na/

/mtho ris bgrod pa’i rgyu des nyon mongs zin par ’gyur/

1 ste] BaBgCDDgJKQTbTkTkK; de HN || la] CDDgHJKNQTbTkTkK; las BaBg 2 la] BgCDDgHJKNQTkK; las BaTbTk 4 /bza’] BgCDDgHJKNQTkK; bzang Ba; /gza’ TbTk || la] BaBgCDDgHJKNQ; las TbTkTkK || ba] BaCDDgHJKNQ; bar BgTbTkTkK 6 /rnam… r e/] BaCDDgHJKNQTbTk; om. BgTkK || tu] BaCDDgHJNTb; du KQTk; om.BgTkK 7 pa yi/] CDDgHJKNQTbTk; pa’i/ Ba; pa med/ BgTkK 8 rtog2] BaBgCDDgHJKNQTb; rtag SMTkTkK || bsgom] CDDgHJKNQTbTk; sgom BaSMTkK; bsgoṃs Bg 10 mnyam par] CDDgHJKNQTbTk; rang bzhin BaBgSMTkK || pas] BaBgCDHJKNQTbTk; bas Dg; pa’i SMTkK

2 /nam mkha’i] /namkha’i BgN 3 med] myed Ba 4 med] myed Ba 5 rjes su] rjesu N | ’jog] ’jog/ J 10 /mi] /myi Ba 11 mongs] myongs Ba

4 /bza’] [J30a1] | shes] [Q29a1] | ’byung] [K42a1] 6 tu] [Tb45a1] 7 bzhin] [D30a1]; [Dg31a1] 9 /bde] [Bg310] 11 bgrod] [TkK210a1] | ’gyur/] [Ba miss. 207a-208b] KHYUNG CHEN LDING BA 155

[11] /lam med lam du zhugs pa rnams kyi nad che ba/

/phyin par ’dod pas ri dags smig rgyu snyegs pa ’dra/

/rnyed pa’i yul med ’jig rten gsum du mi ’byung zhing /

/sa bcur ltos pa’i gnas kyang byang chub sgrib pa yin/ 4

[12] /shin tu myur ba’i ye shes bsam pa kun dang bral/

/bshes gnyen kun las ’byung ba’i nor bu rin chen bzhin/

/dmigs pa med cing ’gyur ba’i gnas la mi ltos par/

/rang gi rang bzhin legs pas re ba thams cad skong / 8

[13] /brtags na med la bzhag na legs pa ’byung ba che/

/dngos su mi snang chags pa’i cha lugs kun la ston/

/bdag dang gzhan du bral ba’i ston mkhan rin chen mdzod/

/thams cad grub pa’i yul zhes bdag med snying rjes bstan/ 12

1 zhugs] CDDgHJKNQTbTk; ’jug BgSMTkK 2 snyegs] DDgHKNQ; snyeg BgCJTbTkTkK; snyed SM || pa] BgCHJKSMTbTkTkKQ; ba DDgN 3 /rnyed] BgCDDgKQSMTbTkTkK; /snyed HJN || du mi] CDDgHJKNQ; las yong mi BgTkK; las yongs mi SM; las mi TbTk || zhing] CDDgHJKNQTbTk; om. BgSMTkK 4 bcur] BgCDDgJKQSMTbTkTkK; bcud HN 5 tu] BgCDDgHNSMTbTkK; du JKQTk 6 /bshes] BgCDDgHJTbTkTkK; /gshes KNQ 7 ba’i] BgCDDgJKNTbTk; pa’i HQ; ba med pa’i TkK || par/] BgCDDgHJKNQTkK; pas/ TbTk 8 pas] BgCDDgJKQTbTkTkK; par HN 9 bzhag] BgCDDgHJNSMTk; gzhag KQTb; gzhan TkK || legs … che/] CDDgHJKNQTbTk; sna tshogs ’byung ba’i legs pa che/ BgSMTkK || pa] BgCDDgHJKNQSMTkK; par NTbTk

5 tu] {tu} Bg; /tu SM 6 /bshes] /gshes Q 8 thams cad] thaṃd N 10 lugs] lugs K

11 /bdag] [C34b1] KHYUNG CHEN LDING BA 156

[14] /nang nas ma g.yos nang du btsal ba’i gnas med de/

/yul phyir chags la yul la bsngo ba’i dmigs pa med/

/mi ’byung mi ’jug bdag med snying rje de yis ni/

/gzhan gyi nor min mi ’ong ye nas gnas pa yin/ 4

[15] /bde ba ’dod na bde la rgyab kyis phyogs/

/bde bas zin te bde bas bde ba tshol/

/byang chub ’khrul pa ye rngam chos la za/

/’di lta’i yul can sangs rgyas mthong ba med/ 8

[16] /sangs rgyas med pas sangs rgyas ming yang med/

/sangs rgyas ston pas ming du btags pa nor/

/sangs rgyas gzhan nas thob byed nor ba’i lam/

/gzugs med chos kun bshad pa rdul tsam med/ 12

3 yis ni/] CDDgHJKNQTbTk; yin na/ BgTkK 5 ’dod na] CDDgHJKNQ; ’dos pas Bg; ’dod pas SMTbTkTkK 6 tshol/] CDDgHJKNQTbTk; ’tshol/ BgSMTkK 7 rngam] BgJKNQTbTkSM; rdam CDDgHTkK 8 /’di] CDDgHJKNQTbTk; /de BgSMTkK 10 btags] BgCDDgJTbTk; brtags HKNQTkK || pa] CDDgHJKN; pas BgTbTkTkK; ba Q 11 ba’i] CDDgKQTb; pa’i BgHJNTkTkK

1 gnas] yul TkK 2 la yul la] sha’i yul la Bg; shing yul la TkK; la yul Q | bsngo ba’i dmigs pa med/] dmigs pa’i bsngo ba med/ Bg | bsngo] sngo K 3 rje] rjes K | de] nga C 4 min] men Bg 5 la] bar SM 6 bas1] ba SM; des TkK 7 pa] pas Bg; bas TkK 8 lta’i] lta SM 9 pas] pa N 10 ston] bston Bg 11 nas] na TkK | byed] med TkK 12 pa] ba J

2 chags] chags C 10 rgyas] rgyas C

1 /nang] [Tk38b1] | gnas] [N42a1] | med] [H43a1] 2 phyir] [Bg311] 4 pa] [K42b1] 5 rgyab] [Tb45b1] 10 btags] [J30b1] KHYUNG CHEN LDING BA 157

[17] /zin dang chags dang bral dang zhi/

/dngos po med dang rnam par spong /

/bdud rtsi chen po’i rang bzhin la/

/bza’ ba’i thabs la rag pa med/ 4

[18] /yangs so che’o chos chen po/

/chud du rnams kyi gnyen po yin/

/che la dmigs pa’i cha bsnyams nas/

/chung che de rnams dmigs dang bral/ 8

[19] /gsung dang rgyas dang dmigs dang snang /

/sgyu mkhan ri mo’i rnam pa bzhin/

/’byung ’jug ye shes rmugs pa yis/

/de yi dbang las skye bar ’gyur/ 12

1 dang1] BgCDDgHJKNQTbTk; pas SMTkK 3 la/] BgCDDgHJKNQSMTkK; pas/ TbTk 5 che’o] BgCDDgHJNTbTkTkK; che bo KQ 6 /chud du] CDDgJTbTk; /chung ngu BgHKNQ; /chung du TkK 7 bsnyams] CDDgHJNTb; mnyams Bg; snyams K; mnyam QTkK; snyam Tk 9 /gsung] CDDgHJKNQTk; /gsungs BgTbTkK || dang2] CDDgHJKNQTbTk; la BgTkK 10 /sgyu] BgCDDgHJNTbTkTkK; /rgyu KQ

1 dang3] yang SM | zhi/] ni/ N 2 spong] spangs SM; spang TkK 4 rag pa] regs pa Bg; rigs pa K; rig ba Q; rig pa SM 6 kyi] kyis Bg | gnyen] snyen K 7 /che] cha K | nas/] na/ Tb 8 che] ngu TkK 9 rgyas] rgyal Bg 10 pa] ba Dg 11 ’jug] ’jugs Bg | rmugs pa yis/] smugs pa yis/ Bg; dmugs pa yis/ K; drmugs pa yis/ Q; rmug pa yis/ Tb; bying rmugs kyis/ TkK 12 /de yi] /de’i Bg

1 dang1] dang / Bg | dang2] dang / Bg 1– 2 zhi/ /dngos] zhi / dngos Bg 3 po’i] po’ī Q 5 che’o] che ’o BgJ | chos] chos N

1 /zin] [TkK210b1] 6 yin/] [Q29b1] 7 /che] [D30b1]; [Dg31b1] 10 /sgyu] [Bg312] KHYUNG CHEN LDING BA 158

[20] /’di ni theg pa kun gyi rgyal/

/btang ngo bzhag go rang bzhin te/

/’dod pa med cing len pa med/

/brod pa rdul tsam mi bskyed de/ 4

[21] /khyung chen nam mkha’ lding ba bzhin/

/spro ba med cing bsdu ba med/

/stor bar mi dogs bza’ ba med/

[22] /rgya mtsho bzhin du ye gnas la/ 8

/chos rnams sna tshogs ’byung bar byed/

/yon tan nam mkha’i mtha’ dang mnyam/

/bsdu ba’i gnas su nges pa med/

[23] /byang chub snying por de ma thag 12

1 /’di … rgyal/] CDDgHJKNQTbTk; om. BgTkK 2 /btang ngo] DDgHKNSMTbTkTkK; gtang ngo Bg; /btad do CJ; /btang nga Q 4 bskyed de/] Tk; skyed cing/ Bg; bskyed do/ CDDgHJN; skyed do/ KQ; skye zhing/ SMTkK; skyed de/ Tb 5 nam mkha’] CDDgHJKNQTbTkK; nam mkhar BgSMTk || lding] BgCDDgHJKNQ; ldings SMTbTkTkK || ba] CDDgHJKNQ; pa BgSMTbTkTkK || bzhin/] BgCDDgHJKNQTbTk; ltar/ SMTkK 6 bsdu] BgCDDgHJNSMTbTkTkK; sdu KQ 7 /stor bar] BgCDDgHJKNQTbTk; /stong par SMTkK || ba] BgCDDgHJKNQSMTkK; ba’ang TbTk || med/] CDDgHJKNQTbTk; med/ /bza’ ba’i yul dang bral de re dogs med/ BgTkK 11 /bsdu] BgCDDgHJNSMTbTkTkK; /sdu KQ

2 te/] de/ SM 4 /brod] /bred SM 6 /spro ba] /spros pa SM 7 dogs] dgos K 8 /rgya] /rgyu SM 10 tan] gtan Q mnyam/] snyam/ K 11 nges pa] des ba N 12 por] po Tk

5 nam mkha’] namkha’ N 10 nam mkha’i] namkha’i KN 11 gnas su] gnasu N 12 thag] thag/ J

9 rnams] [N42b1]; [H43b1] 10 mtha’] [Tk39a1] KHYUNG CHEN LDING BA 159

/ting ’dzin rgyal po chen po ’byung /

/snang ba rgya mtsho chen po bzhin/

/mi rtog nam mkha’i mtha’ ltar yangs/

[24] /kun tu bzang po’i spyod yul la/ 4

/skye zhing ’pho ba’i chos med de/

/rgyu rkyen yan lag bcu gnyis po/

/sgur bas brtags pas bshad pa yin/

/’khrul pa rnams kyi sgo tsam du/ 8

/mkhas pas de bzhin shes par gyis/

[25] /’gro ba rgyud drug snang ba yang /

/dang po’i lam du shes par bya/

/’dod spyod snying rjes brlan pas na/ 12

/gang du dga’ bar byang chub spyod/

2 rgya] BgCDDgHJNTbTkTkK; brgya KQ; om. SM 3 rtog] BgCDDgHJSMTbTkTkK; rtogs KNQ 4 tu] BgDDgHKNTbTkK; du CJQTk 5 chos] CDDgHJKNQTbTk; gnas BgTkK 7 /sgur] DDg; /skur BgCHJKNQTbTkK; /bskur Tk || bas] CTkK; pa Bg; bas DDgHJKNQTbTk || brtags] DDgHN; btab Bg; btags CJKTbTkTkK; gtags Q || pas] BgCDDgHJKNQTkK; te TbTk 9 pas] BgDgHJKNQTbTkTkK; bas CD

1 po2] por Bg 3 mtha’] mkha’ TkK 5 /skye] /skya Bg | de/] te/ Tk 10 ba1] ba’i TkK 11 bya/] gyis/ TkK 12 brlan] brnyal TkK | pas na/] pa la/ SM 13 byang chub] snying rje Tk

3 rtog] rtogs Q | nam mkha’i] namkha’i BgN 4 po’i] po’ī Q 7 brtags] brtags H 9 gyis/] gyīs/ Q

1 ’dzin] [C35a1] | chen] [Tb46a1] 2 mtsho] [K43a1] 9 shes] [Bg313] KHYUNG CHEN LDING BA 160

[26] /shen pa smad ’tshong mtshams med lnga/

/kha na ma tho ’jig rten spong /

/yongs su rdzogs pa chos kyi rtsis/

/bde ba che las gzhan med do/ 4

/yang chos thams cad de bas na/

/chos kyi rang bzhin de yin pas/

[27] /chos kyi chos nyid btsal ba dang /

/nam mkhas nam mkha’ btsal ba dang / 8

/gzhan la ltos pa’i chos nyid ni/

/me yis me gsod pa ’dra ste/

/shin tu dka’ ba’i las byed pa’o/

1 /shen] DDg; /shan BgCHJKNQSMTbTkTkK || pa] BgCDgHJKNSMTbTkTkK; ba DQ 2 ma tho] BgCDDgJKNQTbTk; mtho HTkK 3 pa] CDDgHJKNQSMTkK; pas BgTbTk || rtsis/] CDDgHJNQTbTk; rtsi/ BgSMTkK; tsis K 5 /yang] DDgH; /yong BgJNTbTk; /yod CKQ; om. TkK 6 yin] BgCDDgJKQTbTk; bzhin HN; om. TkK 7 kyi] BgTk; kyis CDDgHJKNQTb; om. TkK || nyid] CDDgHJKNQTbTk; om. BgTkK 8 /nam… m kha’] BgCDDgJKQTbTk; nam mkha’ nam mkhas HN; om. TkK || btsal] BgCDDgHJNTbTk; stsal KQ; om. TkK 9 /gzhan la] CDDgHJKNQTbTk; om. BgTkK 10 pa] CDDgHJKNQ; pa dang Bg; dang TbTk; om. TkK 11 tu] BgCDDgHJNQTb; du KTk; om. TkK

1 smad] smang K | ’tshong mtshams] mtshong mtshams Bg; ’tshong ’tshams Tk | lnga/] byed/ Bg 4 gzhan] bzhan Bg | med do/] min no/ Bg 5– 1 1 /yang chos thams cad de bas na/ /chos kyi rang bzhin de yin pas/ 27 /chos kyi chos nyid btsal ba dang / /nam mkhas nam mkha’ btsal ba dang / /gzhan la ltos pa’i chos nyid ni/ /me yis me gsod pa ’dra ste/ /shin tu dka’ ba’i las byed pa’o/] om. TkK 5 de] da Tb | bas na/] lta bas na/ Bg; lta bas/ Tb; lta na/ Tk; om. TkK 7 btsal] brtsal Q 8 ba] pa N; om. TkK 9 ltos] ltas N; om. TkK | nyid ni/] rnams rten pa ni/ Bg; om. TkK 10 gsod] bsod N; om. TkK 11 pa’o/] do/ Bg; om. TkK

3 /yongs su] /yongsu N 4 las] las C 4– 5 do/ /yang] do// //yong Bg 8 /nam mkhas nam mkha’] /namkhas namkha’ BgH; namkha’ namkhas N 10 /me yis] /mes Bg

3 /yongs] [TkK211a1] 6 /chos] [J31a1] KHYUNG CHEN LDING BA 161

[28] /rnam par mi rtog snying po ’di/

/yongs kyi rgyud la sbas pa med/

/mi len byang chub spyod rnams la/

/rnam pa kun tu rang bzhin gnas/ 4

4 tu] BgDDgHJNTbTkK; du CKQTk || bzhin] BgCDDgHJTbTkTkK; om. KNQ

2 rgyud] brgyud Bg; rgyus TkK 3 rnams] lam TkK

4 gnas/] [Ba miss.]; [Bg313-4]; [C35a5]; [D30b7]; [Dg31b7]; [H43b7]; [J31a2]; [K43a6]; [N42b7]; [Q29b7]; [Tb46a6]; [Tk39a6]; [TkK211a2]

4 gnas/ Bg and TkK read: /yul thams cad la ’jug pa’i bsam gtan du bya ba/ /rnam par mi rtog pa’i snying po/ /man ngag ’di rnams kyang sems chos nyid du gyur pa rnams kyi rna bar rgyud pa/ khyung chen ldings pa rdzogs so// (Bg reads rna ba nas rna bar brgyud pa las gzhun du gleng bar mi bya’o/ /byang chub sems khyung chen rdzogs so // instead of rna bar rgyud pa/ khyung chen ldings pa rdzogs so//). The other editions (CDDgHJKNQTk) read: /zhes gsungs so/ /bya med rdzogs pa’i lung bcu las mi gnas yul med kyi le’u ste nyi shu rtsa gnyis pa’o// (CJKNQ read shes and Tk reads ces instead of zhes; Tb om. med in yul med and reads kyi as gyi; and Tk om. nyi shu). RDO LA GSER ZHUN 162

7.4 The Critical Edition of the rDo la gser zhun

[1] /bsam ’das brjod du med pa’i byang chub sems/ /ston pa’i sgron mar gyur pas yon tan rab bsngags pa/ /chos kyi snying por gyur pas ’jam dpal gzhon nu nyid/ /bya bral lhun rdzogs bde ba’i lhun la gnas/

[2] /tshul khrims la sogs spyod pa rgya mtsho’i gzhir gyur pa’i/ /ji snyed 4 rnam par grol ba’i lam bstan pa/ /bde gshegs yum ’gyur kun gyi mtshungs lam de/ /de med mi ’byung de phyir de ni rnam grol mchog gi lam/

1 brjod] BaCDDgHJNTbTk; rjod KQ 2 yon tan] CDDgHJKN; kun gyis Ba; yon gtan Q; gzhan gyis TbTk || bsngags] BaCDDgHJQTb; sngags KNTk 3 lhun2] BaCDgJKNQTbTk; lhul DH 4 gzhir] BaCDDgHJNTbTk; bzhir KQ 5 bstan] BaCDDgHJNQ; stan K; ston TbTk || ’gyur] CDDgHJKNQ; gyur BaTbTk || mtshungs] BaCDDgHJQTbTk; mtshung KN 5–6 lam de/] CDDgHJNQ; lam te/ Ba; la med/ K; lam ste/ TbTk 6 de2] CDDgHJKNQ; da BaTk; nga Tb

1 /bsam] [Ba210a1]; [C37a2]; [D32b4]; [Dg33b4]; [H46a5]; [J32b6]; [K45b2]; [N45a5]; [Q31a8]; [Tb48b6]; [Tk41b2] 2 yon] [Q31b1] 5 par] [J33a1] | /bde] [Tb49a1]

1 /bsam All extant editions (BaCDDgHJKNQTbTk) read: //de nas byang chub kyi sems kun byed rgyal pos sems las byung ba’i chos chen po lnga ston par byung ba’i rtags su ma skyes pa’i byang chub kyi sems ston par zhal gyis bzhes pa ’di gsungs so/ /kye sems dpa’ chen po nyon cig (Ba reads rgyal po des/ and K rgyal po las/ instead of rgyal pos; Q reads chos chen bo instead of chos chen po; and N reads la instead of lnga. Tb and Tk read ston par phyung ba'i instead of ston par byung ba'i, while Ba reads pa'i as ba'i and N reads it as pi. BaCDDgHJKNQ om. kyi in byang chub kyi sems ston par. Q reads gzhes instead of bzhes and chen po kyod instead of chen po). RDO LA GSER ZHUN 163

[3] /phra zhing shes dka’ kun gyi lam ste mi rtog rtog las ’das/ /mi gnas mi dmigs spros med bsam pa kun dang bral/ /tshig gis mi phebs dbyibs dang kha dog spyod yul med/ /bstan cing brtag par dka’ la brjod du rdul tsam med/

[4] /sngon gyi drang srong lam du gang zhugs pa/ /bsgom chags lam gyi 4 nad kyis zin ta re/ /ston pa’i lung bzhin tshig gi mtha’ spyod lam du de mthong na/ /de ni rtog pa’i rgyun ’brangs smig rgyu snyegs pa ’dra/ /yang dag lam ni tshig gis mtshon du med/ /yang dag nyid du bstan pa nga yis ’khrul/

[5] /dag dang ma dag gnyis med ’dres shing mtshungs/ /rnam pa cir yang 8 mi ’byed ye shes gti mug dang / /thogs med gsal ba’i mar me bsam pa kun dang bral/ /ngang gis mi g.yo rmugs shing ting ’dzin rgyal por gnas/

3 med/1] CDDgHJKNQTk; min/ BaTb || /bstan] BaCDDgHJKNQ; /brtan TbTk || cing] CDDgHJNQ; ci Ba; zhing TbTk || brtag] CDDgHJTb; btags BaN; rtag KQTk 4 /bsgom] BaCDDgHJKNQ; /sgom TbTk 5 /ston pa’i] BaCHJKTbTk;/ston ba’i DDgQ; /bston pa’i N || gi] CDDgHJNQTb; gis BaKTk 6 rgyun ’brangs] DDgHN; lam zhugs Ba; rgyun ’brang C; rgyun ’brengs JKQTbTk || snyegs] CDDgHJN; snyeg BaTb; bsnyegs KQ; snyog Tk 7 bstan pa] CDDgHJNTbTk; bstan pa’i Ba; stan pa KQ 8 cir] CDDgHJNTbTk; cing Ba; bcir KQ 9 ba’i] DDgHJKNQTbTk; pa’i BaC

1 ste] [H46b1] | mi2] [N45b1] 6 /yang] [D33a1]; [Dg34a1] 10 bral/] [K46a1] | rmugs] [Tk42a1] RDO LA GSER ZHUN 164

[6] /mngon sum mthong bas mthong med mngon sum mthong ba’i mig /de phyir thams cad mkhyen pa’i spyan zhes de la bya/ /mtha’ dang dbus med yangs pa’i snying po la/ /mi len mi ’dor mnyam pa’i rgyal por gnas/

[7] /sems dang bag chags gnyis med ’dres shing mtshungs/ /’dzin pas 4 brtags shing snang ba’i chos rnams ni/ /rang gi rgyan du snang bas mi ’dor spong mi byed/ /cir yang mi dgongs thabs kyis de la rol/

[8] /kun dang mi mthun yongs kyis spong ba’i chos/ /nyon mongs lnga dang mtshams med lnga po yi/ /rnam dag lam der zhugs nas mnyam pa’i rgyal 8 po thob/ /bud med la sogs kun kyang spong mi byed/

[9] /lo rgyus don gnyis tshad ma’i blor bzhag nas/ /ting ’dzin gsum po brtan nas grub pa’i mtha’ spyod pa/ /rtsod bral lung las gol zhes ’khrul pa yin/

/bya bral lhun rdzogs bde ba’i lhun la gnas/ 12

2 spyan] BaCDDgHJNTbTk; can KQ 3 snying… l a/] CDDgHJKNQ; rang bzhin shes/ BaTbTk 5 brtags] CDDgHJTbTk; rtags BaKNQ 11 /rtsod] CDDgHJKNQ; /rtsol BaTbTk || pa] BaHKNQTbTk; ba CDDgJ

2 /mtha’] [C37b1] 5 shing] [Ba210b1] | /rang] [Tb49b1] 8 med] [H47a1] | pa’i] [N46a1] 11 grub] [J33b1] | pa/] [Q32a1] RDO LA GSER ZHUN 165

[10] /ye shes rang ’byung che ba’i snying po nyid/ /mi g.yo mi ’gyur bsnyad pa kun dang bral/ /zin pa’i bdud rtsis rtsol ba’i sdug bsngal ’joms/ /bya bral zin pa’i yul la ji bzhin gnas/

[11] /chos rnams thams cad thig le chen po’i rang bzhin yin pas na/ 4

/spros pa med cing bsdus pa med/

/skye ba med cing ’gag pa med/

/’gog pa med par de bzhin gnas/

/rnam par mi rtog snying po ’di/ 8

/nam mkha’ bzhin du ye gnas pas/

/rtog pa’i smra bsam yul las ’das/

2 bsnyad] BaCDDgHJNTb; snyed KQ; snyad Tk || ba’i] DDgHJNTbTk; pa’i CKQ; om. Ba 10 smra] BaCDDgHJNTbTk; smras KQ

1 /mi] /myi Ba | mi] myi Ba; mī Q 2 ’joms/] ’joms/ C 4 le] cig/ J 4– 5 na/ /spros] na / spros Tb 5 med] myed Ba med/] myed/ Ba 6 med] myed Ba | med/] myed/ Ba 7 med] myed Ba 8 mi] myi Ba 9 /nam mkha’] /namkha’ N

8 snying] [K46b1] 10 bsam] [Tk42b1] | ’das/] [Ba210b6]; [C37b6]; [D33a7]; [Dg34a7]; [H47a5] [J33b4]; [K46b1]; [N46a5]; [Q32a3]; [Tb49b6]; [Tk42b1]

10 ’das/ All extant editions (BaCDDgHJKNQTbTk) read: zhes gsungs so/ /byang chub kyi sems kun byed rgyal po las bya med rdzogs pa bsgom du med pa’i le’u ste rtsa drug pa’o// (Tb and Tk read /bya med rdzogs pa’i lung bcu las/ instead of /byang chub kyi sems kun byed rgyal po las. CJKQ read shes and Ba and Tk read ces instead of zhes. Ba reads des/ instead of las and bya bral lhun rdzogs instead of bya med rdzogs pa. C reads rdzogs pa as rdzogs pas and Tk as rdzogs pa'i, while H and N read bsgom du as sgom du). MI NUB PA’I RGYAL MTSHAN 166

7.5 The Critical Edition of the Mi nub pa’i rgyal mtshan

[1] /rdo rje sems dpa’ nam mkha’ che/

/kun bzang yangs pa chos kyi dbyings/

/rnam dag lam chen kun sgrol phyir/

/mi skye mi ’gag cir mi dgongs/ 4

[2] /byams pas don nyid rnam sbyangs phyir/

/snying rje chen po cir mi mdzad/

/che bas che ba’i zab mo yi/

2 pa] BgCDDgHJTbTkTkK; ba N; pas KQ 5 rnam] CDDgHJKNQSMTbTk; rnal Ba; rnams BgTkK 7 bas] BaCDDgHJKNQTbTk; ba’i BgTkK || ba’i] DDgHJKNQTbTk; ba BaTkK; pa’i BgC || yi/] BaBgCDDgJKQTbTkTkK; yis/ HN

3 sgrol] grol TkK 4 ’gag] ’gags Tb 5 pas] pa Tk | sbyangs] ’byor Ba; sbyang Bg; sbyong TkK 6 /snying rje] /rdo rje TkK | po] pos TkK

1 /rdo] rdo TkTkK | nam mkha’] namkha’ BgN 4 /mi] /myi Ba | mi1] myi Ba | mi2] myi Ba 6 mi mdzad/] midzad/ Bg

1 /rdo] [Ba miss 213b]; [Bg384-1]; [C40b4]; [D36a4]; [Dg37a4]; [H51a3]; [J36b1]; [K50a5]; [N50a2]; [Q34b3]; [Tb53b3]; [Tk45b6]; [TkK211b2] 2 chos] [Ba214a1] 7 /che] [Tk46a1]

1 /rdo Bg reads bcom ldan ’das dpal rdo rje sems dpa’ la phyag ’tshal lo/ /chos nyid kyi skabs/, while TkK reads //rgya gar skad du/ Arya pad badzra sad tra/ bod skad du/. Tb and Tk read: //de nas byang chub kyi sems kun byed rgyal po des nyid kyi snying po rdo rje sems dpa' la/ /sems dpa’ rdo rje nyid kyi rang bzhin bya med rdzogs pa’i lung mi nub pa’i rgyal mtshan ’di gsungs so/ /kye sems dpa’ rdo rje khyod nyon cig /sems dpa’ rdo rje khyod nyid la/ /nyid kyi rang bzhin bstan par bya/ /nyid ni kun byed nga yin te/ /nga ni ye nas byang chub sems/ /byang chub sems ni ’di lta ste/ (Tk om. pa'i in mi nub pa'i, and reads /kyi sems dpa' chen po instead of /kye sems dpa'). All other editions (BaCDDgHJKNQk) read: //de nas byang chub kyi sems kun byed rgyal po des nyid kyi snying po sems dpa’ rdo rje la sems dpa’ rdo rje nyid kyi rang bzhin bya med rdzogs pa’i lung mi nub pa’i rgyal mtshan ’di gsungs so/ /kye sems dpa’ rdo rje khyod nyon cig /sems dpa’ rdo rje khyod nyid la/ /nyid kyi rang bzhin bstan par bya/ /nyid ni kun byed nga yin te/ /nga ni ye nas byang chub sems/ /byang chub sems ni ’di lta’o/ /des na byang chub kyi sems kun byed rgyal po des mi nub pa’i rgyal mtshan nam mkha’ che ’di gsungs so/ (J and Q read /da ni ye nas instead of /nga ni ye nas). MI NUB PA’I RGYAL MTSHAN 167

/yon tan cir yang bsngags pa med/

[3] /don rnams ji bzhin mi bskyod de/

/bya ba med pas grol bas grol/

/rang ’byung ye shes btsal med pas/ 4

/grol na grol ba’i lam yang ston/

[4] /’byung ba chen po bcom ldan ’das/

/’gro ba kun la rang bzhin gnas/

/phyin ci log tu rnam brtags kyang / 8

/grol ba rang ’byung gzhan las min/

[5] /che ba’i ye shes rnyed dka’ ba/

/shes rab thabs la brten pas ’grub/

/ming tsam gzhan la brten ’dra yang / 12

2 bskyod] CDDgHJNSMTbTk; bskyed Ba; skyod BgKQTkK 4 /rang] BaBgCDDgJKQSMTbTkTkK; /nges HN || ’byung] BaBgCDDgHJKNQ; byung SMTbTkTkK 5 na] CDDgHJKNQ; bas Ba; nas BgSMTbTkTkK 7 /’gro] BaBgCDDgJKQTbTk; /grol HNTkK 9 ’byung] BaBgCDDgHJKNQ; byung TbTkTkK 10 dka’] BaBgCDDgHJNTbTkTkK; bka’ KQ 11 brten] BaCDDgHTbTkTkK; rten BgJKNQ 12 brten] DDgHTkK; rten BaCJKNQTbTk; brten pa Bg

1 med/] myed/ Ba 2 rnams] rnams Q | mi] myi Ba 3 med] myed Ba 4 med] myed Ba | pas/] {pas}/ Bg 5 ston/] ston// Bg 6 /’byung] ’byung Ba 9 min/] myin/ Ba 12 /ming] /mying Ba

9 gzhan] [K50b1] 10 /che] [Bg385] MI NUB PA’I RGYAL MTSHAN 168

/mngon sum bde ba rang las ’byung /

[6] /cho ’phrul chen po dka’ ba min/

/yon tan kun dang stobs kyi rnams/

/ji bzhin rtogs pa phra ba yis/ 4

/de ma thag tu rang las ’byung /

[7] /snang ba med pa’i chos nyid ni/

/ma btsal bzhag pas bsgoms pa yin/

/de dang der ni rnam btsal na/ 8

/de las de bzhin de mi ’byung /

[8] /mchog tu gsang ba’i chos nyid ni/

/rna dbang gzhan las thos mi ’gyur/

1 sum] CDDgHJKNQTbTkTkK; gsum BaBg || ’byung] BaBgCDDgHJNQTbTk; byung KTkK 3 kyi rnams/] BgCDDgHJKNQSMTkK; kyis rnams/ Ba; rnams kyi/ TbTk 4 rtogs] BaBgCDDgJKQSMTbTkTkK; rtog HN || phra] CDDgHJKNQSMTbTkTkK; ’phra BaBg || yis/] CDDgHJKNQTbTkTkK; yin/ Ba; las/ BgSM 6 ni/] BaBgCDDgHJKNQSMTkK; ’di/ TbTk 7 btsal] CDDgHJKNQSMTbTk; brtsal BaBg; rtsal TkK || bsgoms] DDgHN; bsgom BaBgCDJSMTbTk; sgom KQTkK 8 rnam] BaCDDgHJKNQSMTbTk; rnams BgTkK || btsal] CDDgHJKNQTbTk; brtsal BaBgSMTkK || na/] BgCDDgHJNSMTbTkTkK; ni/ Ba; de/ KQ 11 gzhan] BaBgCDDgHJNSMTbTkTkK; bzhan KQ

1 ba] bar K 2 dka’] dpa’ Ba | min/] men/ Bg 3 dang] la K; ldan Q 4 /ji] /ci Ba | pa] pa’i TkK 5 rang las ’byung /] nga las ’byung / Ba; rang la ’byung / Bg; ’byung bar ’gyur/ TkK 6 pa’i] pa yi Ba | nyid] om. Ba 7 /ma] /mi TkK | bzhag] gzhag Tb; gzhan TkK | pas] pa Ba 8 der] de Bg 9 las] la Bg 10 nyid] ’di Ba 11 las] la Tb | thos] ’byung Bg; thong SM; thob TkK

1 ’byung /] ’byung // Bg 2 min/] myin/ Ba 6 med] myed Ba 8 ni] {ni} Bg 9 mi] myi Ba 11 mi] myi Ba

1 /mngon] [Tb54a1] 2 chen] [H51b1] 5 rang] [D36b1]; [Dg37b1] 6 med] [C41a1] 8 btsal] [N50b1] 11 gzhan] [TkK212a1] MI NUB PA’I RGYAL MTSHAN 169

/de bzhin lce yi dbang pos kyang /

/de la brjod du rdul tsam med/

[9] /’gro ba’i sdug bsngal byang chub sems/

/kun tu chub pas rnam par rol/ 4

/de la bskyod pa med bzhin du/

/nam mkha’i mtha’ ltar mnyam par gnas/

[10] /khyad par cir yang mtshungs pa la/

/las so zhes ni rnam par brtags/ 8

/ci ste las kyi dbang ’gyur na/

/rang ’byung ye shes yod ma yin/

[11] /rgyu nyid rdo rje rkyen dang ’dra/

/ma skyes pas na ’jig pa med/ 12

1 lce yi] CDDgHJTbTk; lce’i BaBgKNQSMTkK 4 tu] BaBgHNTbTkK; du CDDgJKQTk || rol/] BaBgCDDgJKQTbTkTkK; grol/ HN 5 bskyod pa] BaBgCDDgHJNTbTkK; skyod ba KQ; skyo ba Tk 7 cir] BaBgCDDgHJNSMTbTk; ji TkK; bcir KQ 9 ’gyur] BaBgCDDgHJKNQTbTk; gyur SMTkK 10 ’byung] BaBgCDDgHJKNQ; byung SMTbTkTkK

2 med/] myed/ Ba 3 ba’i] ba ’i J 3– 4 sems/ /kun] sems kun J 6 /nam mkha’i] /namkha’i BgN 7 mtshungs] mtshungs Bg 12 med/] myed/ Ba

5 bskyod] [Bg386] 6 mtha’] [Q35a1] 9 /ci] [J37a1] MI NUB PA’I RGYAL MTSHAN 170

/gdod nas snying po byang chub la/

/btsal ba’i bsam pas dbyings mi bskyod/

[12] /yon tan chen po’i bsam gtan ni/

/bsam gtan nyid pas bsam du med/ 4

/ma bsams ma sbyangs chos bzhin du/

/rnam rtog nyid las ye shes skye/

[13] /phra ba’i sgo mor ming btags te/

/sems kyi dben pa lam tshol zhing / 8

/dgon pa’i rgyud du dben ’dzin cing /

/brtags na rnam par rtog ’gyur bsgom/

[14] /rgyu dang ’bras bur ming btags te/

2 /btsal] CDDgHJKNQTbTk; /brtsal BaBg; /rtsal SM; /brtsol TkK 5 bsams] DDgSMTkTkK; bsam BaBgCHJKNQTb || bzhin] BgCDDgHJKNQTbTk; nyid BaTkK; bnyid SM 6 las … skye/] BaBgCDDgJKQTbTkTkK; las ye shes skyes/ HN; la skye/ SM 7 /phra] CDDgHJKNQTbTkTkK; /’phra BaBgSM || btags] BaCDDgJSMTbTkTkK;’dogs Bg; brtags HKNQ 8 kyi] BaBgCDDgHJKNQTkK; kyis SMTbTk || pa] CDDgHJKNQ; ba’i Ba; pas BgSMTkK; pa’i TbTk 9 cing /] CDDgHJKNQ; te/ BaTkK; pa/ BgSM; zhing / TbTk 10 bsgom/] BaBgCDDgHJNSMTbTkTkK; sgom/ KQ 11 ming] BaBgCDDgHJNTbTkTkK; ’di KQ || btags] BaCDDgHJKQTb; brtags NTk; ’dogs BgTkK || te/] BaCDDgHJKNQTbTk; shing/ BgTkK

1 /gdod] /gdon N 2 pas] pa’i K; bas Q | mi] mir D | bskyod/] bskyed/ Ba 3 chen po’i] che ba’i TkK 4 pas] las Bg; la Ba 5 sbyangs] spyangs Ba 6 rtog] rtogs Bg 7 te/] pas/ SM; nas/ TkK 8 dben] bde Ba | tshol] ’tshol Bg 9 /dgon] /sgom K 10 na] nas Ba | rtog] rtogs TkK

2 mi] myi Ba | bskyod/] bskyod// Bg 3 chen] che{n} Bg 4 /bsam] bsam Bg; /bsaṃ SM | bsam du med/] {bsam du med/} Ba 5 /ma] {/ma} Ba | bsams] bsam{s} SM | bzhin] bzhīn Q 7 ming] mying Ba 10 rtog ’gyur] rtogyur Bg bsgom/] {b}sgom/ Bg 11 /rgyu] /rgyud Ba | ming] mying Ba

2 bskyod/] [Ba214b1] 4 /bsam] [Tk46b1] 10 /brtags] [K51a1] | rtog] [Tb54b1] MI NUB PA’I RGYAL MTSHAN 171

/dge sdig gnyis ka rnam par sel/

/’jig rten ’di la ’byung ngo zhes/

/blang dor brod pa mchog tu bskyed/

[15] /chags dang ma chags tshig gi lam/ 4

/dbu ma bzhin te brag cha ’dra/

/bde dang sdug bsngal rgyu mthun zhes/

/’gro ba’i mgon po sems dpas gsungs/

[16] /’dod chags zhe sdang gti mug kyang / 8

/byang chub chen po’i lam las byung /

/kun spyod yon tan rnam lnga yang /

/chos nyid dbyings kyi rgyan zhes gsungs/

[17] /nam mkha’ rtog pa skye med cing / 12

/rtog pa de nyid nam mkha’ ’dra/

1 gnyis] BaBgCDDgHJKNQTb; gnyi TkTkK || ka] BgCDDgHJKNQTb; ga BaTkTkK 2 la] CDDgJKNQ; las BaBgHTbTkTkK 3 bskyed/] BgCDDgHJNTbTkTkK; skye/ Ba; skyed/ KQ 5 cha] BaBgCHJKNQTbTkTkK; ca DDg 6 mthun] BaBgDDgHKQTbTkK; ’thun CJNTk 7 dpas] BaCDDgHJQTbTkTkK; pas BgKN 11 gsungs/] BaCDDgHJKNQSMTbTk; gsangs/ BgTkK 12 /nam mkha’] BaBgCDDgHJKNQ; /nam mkha’i TbTkTkK

3 brod] phrod H 4 tshig] chig Q | lam/] yam/ Dg 5 /dbu] /sgyu TkK | te] du K 7 mgon] ’gon Ba 9 byung] ’byung Bg 10 /kun] ’dod Ba 11 kyi] kyis TkK | rgyan] rgyun Ba; brgyan Bg 12 cing /] pa/ TkK

9 po’i] po’ī Q 10 spyod] {spyod} Ba 12 /nam mkha’] /namkha’ BgN | med] myed Ba 13 nam mkha’] namkha’ BaBgN

1 sel/] [H52a1] 4 /chags] [Bg387] 8 zhe] [N51a1] 12 /nam] [TkK212b1] MI NUB PA’I RGYAL MTSHAN 172

/mi chags nam mkha’i bsngo ba las/

/rang don chen po nam mkha’ ’byung /

[18] /rtog med mnyam nyid chos kyi sku/

/bzung bas mi zin chu zla ’dra/ 4

/kun tu bzang po’i rol pa yis/

/A li kA li zab tu bstan/

[19] /’di ni a dang mdzes pa’i ta/

/pa dang yan lag spros pa bzhin/ 8

/’jig rten yongs kyi spyod yul la/

/sangs rgyas gsung gi zab mo ’byung /

[20] /e ma’o sangs rgyas spyod yul ’di/

/btsal bas rnyed pa’i gnas med de/ 12

/drug gi chos bzhin yul med pas/

1 bsngo] BaCDDgJKNQTbTkTkK; sngo BgH 2 po] CDDgHJKNQTbTkTkK; po’i BaBg 4 /bzung] BaBgCDDgJKQTkTkK; /gzung HNTb 5 tu] BaBgHKNQTbTkTkK; du CDDgJ || yis/] BaBgCDDgHJKNQTbTk; yi/ TkK 6 /A … li2] CDDgHJKNQTb; /a la ka li BaTk; /a li ka li BgTkK 11 ma’o] CDDgHJNTbTkTkK; ma’i Ba; ma ho BgKQ 12 /btsal] CDDgHJKNQTbTk; /brtsal BaBgTkK

1 /mi] /ma Tk | nam mkha’i] namkha’ Bg 4 zla] om. Ba 6 tu] du Ba 7 /’di] /de TkK | mdzes] ’dzes N | ta/] tha/ Ba; rta Bg 8 spros pa] spro ba TkK 9 spyod] spyol Bg 10 gsung] gzung Ba | gi] ni Ba 12 rnyed] snyed Bg pa’i] ba’i Ba | med de/] myed pa/ Ba; med te/ Tk; med la/ TkK 13 yul] gnas Ba | pas/] bas/ Q

1 /mi] /myi Ba | nam mkha’i] namkha’i BaN 2 nam mkha’] namkha’ BaBgN 3 nyid] nyīd Q 5 po’i] po’ī Q 6 li2] lī Q 13 med] myed Ba

1 bsngo] [C41b1] 3 chos] [D37a1]; [Dg38a1] 8 yan] [Bg388] MI NUB PA’I RGYAL MTSHAN 173

/ldongs pas nam mkha’ bsnyabs pa bzhin/

[21] /gong nas gong du tshangs pa’i lam/

/bya bral chos dang mthun pa min/

/ci ste lam la bgrod ’gyur na/ 4

/nam mkha’i mtha’ bzhin thob pa med/

[22] /de ltar de bzhin de yi phyir/

/de la de bstan de yang thob/

/de ni snying po de bas na/ 8

/de las de byung ngo mtshar che/

[23] /sngon gyi de dang da ltar de/

/de bzhin de yi gnas su che/

/de ltar de yi lam ’dra ba/ 12

1 /ldongs] CDDgJ; /mdongs BaBgHKNQTbTkTkK || bsnyabs] DDgH; rnyeg Ba; snyabs Bg; brnyabs CJKNQ; bsnyags Tb; bsnyag Tk; rnyabs TkK 3 mthun] BaBgDDgHKNQSMTbTkK; ’thun CJTk || min/] BaCDDgHJKNQSM; phyir/ Bg; med/ TbTk; bzhin/ TkK 4 ’gyur] BaBgCDDgHJKNQTkK; gyur SMTbTk 5 thob] BaCDDgHJKNQTbTk; thug BgSMTkK 6 de yi] BaCDDgHJKNQTbTk; de’i BgTkK 9 byung] BaCDDgJKQTbTk; ’byung BgHNSMTkK 11 de yi] CDDgHJKNQTb; de’i BaBgTk; de ni TkK 12 de yi] CDDgHJKNQTbTk; de’i BaBgTkK || ba/] BgCDDgHJKNQ; bar/ BaTbTkTkK

1 nam mkha’] namkha’ BgN 3 min/] myin/ Ba 5 /nam mkha’i] /namkha’i BgN | med/] myed/ Ba 11 de yi] de{’i} Bg | gnas su] gnasu/ N

1 pa] [J37b1] 4 /ci] [Tk47a1] 5 thob] [K51b1] 6 de1] [Tb55a1] 7 thob/] [H52b1] 8 /de] [Q35b1] 11 bzhin] [Ba215a1] MI NUB PA’I RGYAL MTSHAN 174

/de ni de yi rang bzhin no/

[24] /de dang ’dra ba yongs kyi lam/

/zla ba las byung rten dang bcas/

/kun gyi mnyam nyid yin pa la/ 4

/phyogs su bltas pas grub pa med/

[25] /de ltar bde dang phyi mar bde/

/mngon sum pa dang rgyab nas ’byung /

/de yang rnam pa’i skyon yin pas/ 8

/de la brten par mi bya’o/

[26] /dus gsum gcig ste khyad par med/

/sngon med phyis med gdod nas ’byung /

/chos skus khyab pas gcig pa’i phyir/ 12

1 de yi] CDDgHJKNQTbTk; de’i BaBgTkK 4 nyid …l a/] BgCDDgHJKNSMTbTk; nyid yin ba la/ BaQ; pa de nyid yin/ TkK 5 bltas] BaBgCDDgHJKNQTbTk; lta SM; ltas TkK || grub] BaCDDgHJKNQSMTkK; ’grub BgTbTk 6 /de] BaCDDgJKNQ; /da BgHTbTkTkK || ltar] BaCDDgHJKNQTbTk; lta’i BgTkK || mar] BaCDDgHJKNQTbTk; ma’i BgTkK 7 sum] BaCDDgHJNSMTbTkTkK; gsum BgKQ || ’byung] CDDgHJKNQ; byung BaBgSMTbTkTkK 9 brten] BaCDDgHJTbTk; rten BgKNQTkK

1 rang] {rang} Bg 5 /phyogs su] /phyogsu N | med/] myed/ Ba 7 sum] suṃ SM 9 bya’o/] bya ’o/ H 10 med/] myed/ Ba 11 med1] myed Ba | med2] myed Ba

5 med/] [N51b1] 6 /de] [Bg389] MI NUB PA’I RGYAL MTSHAN 175

/che ba’i chen po rang bzhin gnas/

[27] /srid pa gsum na sbyor ba yang /

/ming tsam sgyu mar snang ba ste/

/’khor los sgyur ba’i gnas chen yang / 4

/sgyu ma sbyong ba’i bsti gnas yin/

[28] /rnam spyod dus la ltos pa rnams/

/dus de ’byung bar mi ’gyur te/

/ma bral smon pa spyod pas na/ 8

/stong pa’i mtshan nyid gsungs pa bzhin/

[29] /gcig ste rnam pa yongs kyis med/

/rnal ’byor nam mkha’i bya lam gnas/

/ma byung ma skyes snying po la/ 12

1 ba’i] BaBgCDDgHJKNQTbTk; bas SMTkK || po] DDgHJKNQ; po’i CSM; por BaBgTbTkTkK 2 sbyor ba] BgCDDgHJKNQSM; spyod pa Ba; sbyong ba TbTk; sbyor TkK 3 ste/] BaBgCDDgHJNSMTbTkTkK; dang/ KQ 4 los] BaCDDgHJKNQSM; lo TbTkTkK; om. Bg || sgyur] BaCDDgJKNQTbTkK; bsgyur BgHSMTk 5 /sgyu] BaBgCDDgHJNSMTbTkTkK; /rgyu KQ || sbyong] CDDgHJKNQTbTk; spyod Ba; sbyang Bg; sbyor SMTkK 6 spyod] BaBgCDDgHJKNQTb; dpyod TkTkK 7 de] CDDgHJKNQTkK; der BaBgTbTk 8 pa] BgCDDgHJKNQTkK; pas BaTbTk 10 /gcig] BaCDDgHJNSMTbTkTkK; /cig BgKQ || kyis] BaCDDgHJKQSMTbTkTkK; kyi BgN 12 ma skyes] BaCDDgHJKNQSMTbTk; mi skye BgTkK

3 sgyu] sgyur Bg; rgyu Q 5 ma] mas SM | bsti] gti Ba; sti TkK 6 pa] pa’i TkK 7 ’byung] byung Ba 8 smon pa spyod pas] spyod pas smon pas Ba; smon pa med pas TkK | pas] bas Q 9 bzhin/] yin/ Ba 10 pa] par TkK 11 nam mkha’i] nam mkha’ TkK 12 la/] yang/ Ba

3 /ming] /mying Ba 6 rnams/] rnams/ Bg 7 mi] myi Ba 9 bzhin/] bzhin// Bg 11 nam mkha’i] namkha’i BgN | lam] laṃ Bg

1 chen] [TkK213a1] 11 gnas/] [C42a1] 12 /ma] [Bg390] MI NUB PA’I RGYAL MTSHAN 176

/sgros btags chos kun ga la yod/

[30] /phyi nang gnyis ka phyi nyid nang /

/zab mo cha shas rtogs yul med/

/srid pa ming tsam log pa’i stobs/ 4

/de bas ting ’dzin mnyam dang bral/

[31] /de la tha tshig phyi dang nang /

/rang bzhin phung po khams bzhin gnas/

/dus gsum ’di dang mi ’bral bas/ 8

/tha tshig ming du btags pa med/

[32] /mi g.yo ba ni sku yi rgya/

/mi bskyod pas ni ye shes te/

/mi len pa ni bdag med cing / 12

3 mo] BaCDDgHJKNQTkK; mo’i BgSMTbTk || rtogs] CDDgHJKNQSMTbTkTkK; rtog BaBg 4 log] BaCDDgHJNTbTkTkK; ldog Bg; logs KQ 6 tshig] BaBgDDgHNTbTkTkK; tshigs CJKNQSM 9 tshig] BaBgCDDgHNSMTbTkTkK; tshigs CJKNQ || btags] BaBgCDDgHJNSMTbTkK; brtags Tk; rtags KQ || pa] CDDgHJKNQSMTbTkTkK; ba BaBg 10 sku yi] CDDgHJKNQTbTk; chos sku’i Ba; sku’i BgTkK || rgya/] BgCDDgHJNTbTkTkK; rgyal/ Ba; brgya/ KQ 11 pas] CDDgHJNQ; pa BgTbTkTkK; bas K; om. Ba 12 pa] CDDgHJKNQ; pas BaBgTbTkTkK || ni] CDDgHJKNQ; na BaBgTbTkTkK

1 btags] btab Bg | kun] sku Ba; su TkK 2 ka] ga SM | nang /] na/ TkK 3 shas] shes Q 4 pa] pa’i Tk 5 ting] gting Bg 7 /rang] /de TkK | phung] pung K | khams] rang Ba 8 mi] ma Ba 10 gyo] yo K 11 bskyod] skyong K; skyod Q | pas ni ye shes te/] ye shes de nyid ni/ Ba 12 med] nyid TkK

3 cha] {cha} SM | med/] myed/ Ba 4 ming] mying Ba | stobs/] stobs/ C 7 khams] khams N 8 bas/] bas/ Ba 9 ming] mying Ba | med/] myed/ Ba 10 /mi] /myi Ba 11 /mi] /myi Ba | bskyod] {b}skyod Bg 12 /mi] /myi Ba | med] myed Ba

2 /phyi] [D37b1]; [Dg38b1] 4 ming] [Tb55b1] 5 mnyam] [Tk47b1] 7 bzhin2] [H53a1]; [K52a1] MI NUB PA’I RGYAL MTSHAN 177

/mi ’dor tshig bral mnyam nyid do/

[33] /gang dang gang gi gang du yang /

/kun ’khol kun spyod bdag las byung /

/’di la skyes pa bud med ces/ 4

/mnyam pa’i rgyal pos yong ma gsungs/

[34] /’di la brtul zhugs drag shul gyis/

/gnas su bya ba’i ming med de/

/a dang par ni rnam ldan na/ 8

/sgyu ma’i bde ba ’byung bar ’dod/

[35] /rang bzhin gcig tu ma nges pas/

/ji ltar bltas pa de ltar snang /

2 gi] CDDgHJKNQSMTb; gis BaBgTkTkK || du] BaBgCDDgHJNSMTbTkTkK; tu KQ 3 ’khol] BgCDDgHJKNQSMTk; ’khor BaTb; bkol TkK || byung] CDDgHJKNQTbTkTkK; ’byung BaBgSM 5 yong] BaDDgHJNQTk; yod CK; yongs BgTbTkK 6 /’di] BgCDDgHJKNQSMTbTk; /de BaTkK || zhugs] CDDgHJKNQSMTbTkTkK; shugs BaBg || gyis/] BgCDDgHJKNQTbTkK; kyis/ BaSM; gyi/ Tk 9 /sgyu] BaBgCDDgHJNTbTkTkK; /rgyu KQSM 10 gcig] BgCDDgHJNSMTbTkTkK; cig BaKQ 11 /ji] BgCDDgHJKNQSMTbTk; /ci BaTkK || bltas] CDDgHJKNQTb; ltas BaTkTkK; gnas Bg; brtags SM

1 /mi] /myi Ba 2 /gang dang] /ga{ng dang} Bg | gi] gis Bg 3 ’khol] {’}khol Bg 3– 4 byung / /’di] ’byung ’di Ba 4 med] med/ Bg 6 gyis/] kyi{s}/ SM 7 /gnas su] /gnasu N 8 /a] a Bg

1 /mi] [J38a1] 5 yong] [N52a1] 7 de/] [Bg391] 8 ldan] [Ba215b1] MI NUB PA’I RGYAL MTSHAN 178

/snang ’dod rtsol sems bde ba yang/

/de ni de sgrib skyon chen yin/

[36] /byang chub yan lag kun gyi sgo/

/cha lugs bsgoms pas chu zla bzhin/ 4

/ma gos ma chags ’byung ’gyur yang /

/bsgoms pas byis pa’i spyod yul bzhin/

[37] /dkyil ’khor khro gnyer cha lugs kyis/

/khro bdag chen po’i lus bzung nas/ 8

/yi ge mngon du byung na yang /

/zhi ba de nyid mthong ba min/

[38] /nyon mongs pa yi dbang gis na/

/ta la’i mgo bo bcad pa dang / 12

1 yang/] BgCDDgHJKNQTkK; la/ BaTbTk 6 /bsgoms] BaCDDgHJNTbTkTkK; /bsgom Bg; /sgoms KQ 8 bzung] BaBgCDDgHJNTbTkK; gzung KQTk 9 byung] BgCDDgHJKNQTkK; phyung BaTbTk 11 na/] BgCDDgHJKNQ; ni BaTbTkTkK; om. Bg 12 la’i] CDDgHJKNQTbTkTkK; la BgBa

1 rtsol] brtsol Bg 2 chen] can Ba 4 /cha lugs] /cha lug Bg; /ma chags TkK | bsgoms] bsgom K | pas] pa TkK 5 ’byung ’gyur] cir ’gyur Ba; ’byung gyur Bg 6 byis pa’i] bying ba’i TkK 7 kyis/] bzhin/ Ba; kyi/ TkK 8 lus] lung Ba; cha TkK 11– 1 2 /nyon mongs pa yi dbang gis na/ /ta la’i mgo bo bcad pa dang /] /nyon mongs lta la mgo bcad dang / Bg 11 pa yi] pa’i Ba; om. Bg

4 bsgoms] bsgoms Q 5 yang /] yang TkK 6 /bsgoms] /sgoms K 6– 7 bzhin/ 37 /dkyil] bzhin{/ /}dkyil Ba 7 lugs] lugs Q 10 min/] myin/ Ba

1 /snang] [Q36a1] 5 yang] [TkK213b1] MI NUB PA’I RGYAL MTSHAN 179

/sa bon me yis bsregs pa dang /

/de yi dbang du mi ’gyur ston/

[39] /rnam grangs brgya stong phrag yas pa/

/gang ltar spyad kyang me tog skye/ 4

/mtshan ma med pa’i dbang gis na/

/bsti gnas de las ’byung mi ’gyur/

[40] /gleng bral ’di la gnas pa ni/

/rnal ’byor de ni skal ba bzang / 8

/bdag dang gzhan don mi ’byed pas/

/sgyu ma lhun grub yul la rol/

[41] /lhag ma med pas yongs su rdzogs/

/’gyur ba ma yin drang por gnas/ 12

1 dang] CDDgHJKNQ; yang BaTbTk; bzhin BgTkK 2 /de yi] CDDgHKNQTbTkTkK; /de’i BaBgJ 3 yas] CDDgHJKNQTbTk; yangs BaBgSMTkK || pa/] CDDgHJKNQTbTk; pas/ BaBgSMTkK 5 na/] BgCDDgJKQTbTkTkK; ni/ BaHN 7 la] BaBgCDDgHJKNQTkK; la gang TbTk || pa ni/] BaCDDgHJKNQ; pa’i/ TbTk; pa yi/ BgTkK 8 de] CDDgHJKNQTbTkTkK; ’di BaBg || skal] CDDgHJNQTbTkTkK; bskal BaK; om. Bg || ba] BgCDDgHJQTbTk; pa BaKNTkK 10 /sgyu] BaBgCDDgHJTbTkTkK; /rgyu KNQ || ma] BgCDDgHJKNQTbTk; ma’i BaTkK 11 pas] CDDgHJKNQTkK; par BaBgTbTk

2 mi] myi Ba 3 stong] {stong} Bg 6 mi] myi 7– 8 ni/ /rnal] ni rnal Ba 9 mi] myi/ Ba 11 /lhag] //lhag Bg | med] myed Ba | yongs su] yongsu N

2 ston/] [Tb56a1] 5 gis] [H53b1] 7 /gleng] [Bg392] 8 ba] [Tk48a1] 9 dang] [C42b1] | gzhan] [K52b1] MI NUB PA’I RGYAL MTSHAN 180

/nam mkha’ bzhin te mtha’ mnyam zhing /

/gzhan la ltos ’gyur chos ma yin/

[42] /lhun gyis grub pa’i bde chen te/

/mtshungs pa med pa’i ye shes kyis/ 4

/rang gi mthu yi rig pa las/

/chos ni gzhan las ’byung mi ’gyur/

[43] /sla zhing dka’ la sla phyir dka’/

/mngon sum mi gnas kun tu khyab/ 8

/ming tsam ’di zhes bstan par ni/

/rdo rje sems dpas mtshon du med/

[44] /ngo mtshar rmad byung rol pa ’di/

1 te] CDDgHJKNQTkK; du BaBgTbTk; de SM 3 gyis] BaCDDgHJKNQSMTbTk; gyi BgTkK || pa’i] CDDgHJKNQSMTbTkTkK; pas BaBg || te/] BgTkK; ste Ba; de/ CDDgHJKNQSMTbTk 4 kyis/] BaBgDDgHKQSMTkK; kyi/ CJNTbTk 5 mthu yi] CDDgHJKNQTkK; mthu’i Bg; mthu yis BaSMTbTk 6 gzhan las] BaCDDgHJKNQ; gzhan nas BgSMTbTk; om. TkK 7 /sla] BaCDDgHJNSMTbTkTkK; /bla BgKQ 8 sum] CDDgHJKNQSMTbTkTkK; gsum BaBg || tu] BaBgHKNQSMTbTkK; du CJDDgTk 9 par] BgCDDgHJNQSMTbTkTkK; pa BaK 10 dpas] BgCDDgHJKNQTb; pas Ba; dpa’ SMTkTkK 11 byung] CDDgHJKNQSMTbTkTkK; ’byung BaBg

1 mnyam] snyam K 2 ltos] bstos SM; stobs TkK 3 grub] gnas Bg 4 pa] om. Ba | ye shes] chos nyid TkK 5 gi] rig Ba 6 ’byung] ’byung bar TkK | ’gyur/] ’gyur ro/ TkK 7 dka’] bka’ Bg; dka’ ba K | sla] bla Tk 9 /ming] /mngon TkK | tsam] sum TkK 10 /rdo] /rda N | mtshon] mchon N; ’tshon SM 11 rmad] smad Bg | pa] om. Bg

1 /nam mkha’] /namkha’ BgN | mtha’] mtha’ [ mtha’] Bg 4 /mtshungs] /mtshungs QTb | med] myed Ba 8 mi] myi Ba 9 /ming] /mying Ba 10 med/] myed/ Ba

1 mtha’] [D38a1]; [Dg39a1] 2 yin/] [N52b1] 10 /rdo] [J38b1] MI NUB PA’I RGYAL MTSHAN 181

/bya bral nam mkha’ bzhin du gnas/

/cir yang mi dmigs gti mug las/

/de ma thag tu rang las byung /

[45] /’di ni thams cad mtshungs pa’i lam/ 4

/’gro ba kun gyi rang bzhin gnas/

/bus pas bslad pas ’khrul ba’i phyir/

/sman nyid sman pa tshol ba bzhin/

[46] /go ba’i yul na bde ba che/ 8

/’di ni rnam dag ’jig rten yin/

/de la phyogs kyi ’od ’dus pas/

/phyogs bzhi mtshams dang bla ’og ’grub/

1 bzhin … gnas/] CDDgHJKNQTbTk; ci bzhin gnas/ Ba; ji bzhin te/ BgSMTkK 2 /cir] BaCDDgHJKNQSMTbTk; /ci BgTkK 3 byung] BaCDDgHJKNQTkK; ’byung BgSMTbTk 5 gyi] CDDgHJKNQTkK; la BaBgTbTk 6 /bus] BgCDDgHJKNQTbTk; /byis BaSM; om. TkK || pas1] DgHKNQSMTbTk; bas BaCDJ; pa Bg ; om. TkK || bslad pas] CDDgHJKNQSMTbTk; bslad pa’i BaBg; om. TkK || ’khrul ba’i] CDDgJQ; ’khrul pa’i BaHNSMTbTk; ’khrul pas Bg; ’khrul pi’i K; om. TkK 8 /go] BaCDDgHJKNQSMTbTkK; /’gro BgTk 11 mtshams] BgCDDgHJKNQTbTkK; ’tshams BaTk

1 nam mkha’] rol ba Ba; nam mkha’i K 3 thag] dag SM | rang] nga TkK 4 thams] tham Bg 6 /bus pas bslad pas ’khrul ba’i phyir/] /bus pa ’khrul pas bslad pa’i phyir/ Bg; /thabs mkhas pa’i spyod yul yin/ TkK | phyir/] phyin/ SM; om. TkK 7 pa] pas Tk | tshol] ’tshol Bg | ba] pas K | bzhin/] nis/ K 8 na] la Ba 9 ni] nyid SM | rten yin/] rten/ Ba 11 dang] brgyad Ba

1 nam mkha’] namkha’ BgN 2 mi] myi Ba | dmigs] dmyigs Ba; dmigs CK 4 cad] {cad} Bg 7 bzhin/] bzhin// Bg 8 na] {na} Bg

5 /’gro] [Bg393] 7 /sman] [Ba216a1] 9 /’di] [TkK214a1] MI NUB PA’I RGYAL MTSHAN 182

[47] /ma nges ’ja’ tshon kha dog las/

/rigs kyi khyad par mngon par snang /

/de bzhin g.yo rdul mi g.yo ba/

/’byung ba lnga bas gtso chen yin/ 4

[48] /’das dang ma ’ongs da ltar gyi/

/tha snyad ming la mi gnas te/

/skye ’gag med par rnam brtags shing /

/de nyid dus gsum chen por sbyor/ 8

[49] /mnyam pas rim par bkod pa med/

/gcig pas phyogs bcu bsngo dang bral/

/tshogs kyi rgyan rnams bkram pa yang /

/rang bzhin gnas pas ’grim pa med/ 12

1 ’ja’] BgCDDgHJKNQTbTk; gzha’ BaTkK 2 mngon] BaCDDgHJNTbTkTkK; sngon BgKQ 4 bas] CDDgHJNTbTk; la BaBg; las KQTkK || gtso] BgCDDgHJNTbTk; sogs Ba; mtsho KQ; rtso TkK || yin/] BaBgCDDgHJNTbTkTkK; bzhin/ KQ 8 sbyor/] BgDDgHJKNQTbTkK; spyor/ BaC; gnas/ Tk 10 bcu] BaCDDgHJKNQ; su BgSMTbTkTkK 11 rgyan rnams] BaCDDgHJNTbTkTkK; brgyan rnams BgKQ; brgyan dang SM 12 ’grim] BgCDDgHJNSMTbTkTkK; ’drim Ba; ’grin KQ

3 mi] myi Ba 4 yin/] yin// Bg 5 ’ongs] ’ongs Bg | gyi/] gyi Ba 6 mi] myi Ba 8 gsum] gsuṃ C 9 med/] myed/ Ba 11 rnams] rnaṃs Bg 12 med/] myed/ Ba

1 /ma] [Tb56b1] 5 da] [H54a1] 8 sbyor/] [K53a1] 9 par] [Q36b1] | med/] [Tk48b1] 11 bkram] [Bg394] MI NUB PA’I RGYAL MTSHAN 183

[50] /lhun gyis gnas pas mi bsngo zhing /

/gdod nas dag pas bdud rtsi yin/

/’du mched bcu gnyis khyad par du/

/lhag pa’i bsam pas bzung ba med/ 4

[51] /yid kyis bsam pas yon bdag ste/

/bltas pas tshogs ni bkram pa’o/

/mthong bar gyur pas dngos grub la/

/de nyid mnyam bzhag rdzogs pa’o/ 8

[52] /yud tsam bzung bas sbyor ba yin/

/dga’ bar gyur pas dam tshig ste/

/thabs kyis gar thabs bskyod pa yis/

/gnyis med sbyor ba phul ba yin/ 12

1 gnas] BaBgCDDgHJKNQTb; grup SMTkK; grub Tk || bsngo] BgCDDgHJKNQSMTbTk; sngo BaTkK 4 bzung] BaBgCDDgHJKNQTkK; gzung TbTk || ba] CDDgHJKNQTbTkTkK; du BaBg 5 kyis] BaCDDgHJKNQTbTkTkK; kyi BgSM || bsam] BgCDDgHJKNQSM; bsngos Ba; bsams TbTkTkK || bdag] BaBgDDgKQSMTbTkTkK; gdag CJN;gdags H 6 tshogs ni] BaCDDgHJKNQTbTk; stobs ni BgSMTkK 7 pas] BgCDDgHJKNQTbTk; ba Ba; pa SMTkK 8 /de nyid] BaCDDgHJKNQSMTbTk; /de ni BgTkK || bzhag] BaBgCDDgHJKNQSMTkK; gzhag TbTk 9 bzung] BaBgCDDgHJKNQTkK; gzung TbTk 10 pas] BgCDDgHJKNQTk; pa BaTkK; bas Tb 11 kyis] CDDgHJKNQTkK; kyi BaBgTbTk || bskyod] BgCDDgHJKNQTbTk; spyod BaTkK

1 pas] pa Ba 2 /gdod] /gdong K 3 mched] phyed TkK 5 pas] pa’i Bg | ste/] te/ H 6 /bltas pas] /bltas pa Bg; /ltas pa’i TkK | pa’o/] pa po/ Bg 7 bar] bas Ba | grub la/] las/ Ba 9 bas] ba Ba; bar TkK 10 gyur] ’gyur Tb tshig ste/] tshigs te/ Bg 11 gar] gang C | pa] ba Q | yis/] yang/ TkK 12 med] min TkK

1 mi] myi Ba 3 /’du] /’du[ s] Bg 6 pas] pas N | pa’o/] pa ’o/ H 7 gyur] gyurd BaBg 8 pa’o/] pa’o// Bg 10 gyur] gyurd Ba 11 thabs] thabs Tb | yis/] yis/ Bg 12 med] myed Ba

2 nas] [N53a1] 6 ni] [C43a1] 11 gar] [D38b1]; [Dg39b1] MI NUB PA’I RGYAL MTSHAN 184

[53] /mi bzung gtong bas gtong ma yin/

/bya ba med pa’i las rnams te/

/mi rtog ye shes bgegs bsal nas/

/ma gsungs mnyam bzhag sngags tshig go/ 4

[54] /bla ma mchod dang gtong ba dang /

/de bzhin bsod nams thams cad kyang /

/ma chags mi g.yo’i stobs med na/

/byas na ’ching ba chen por ’gyur/ 8

[55] /de bas de lung de nyid la/

/de la de sbyar sgrib par ’gyur/

/de ltar de la de rtog na/

1 bzung] BgCDDgHJKNQSMTkK; gzung TbTk; om. Ba || gtong bas] BgCDDgHJKNTbTk; gtong pas QSM; rtong bas TkK; om. Ba || gtong2] DDgH; gtor BgCJKNQSMTbTkTkK; om. Ba 2 pa’i] CDHJKNQSMTbTk; ba’i BaCDDg; pas BgTkK 3 bsal] CDDgHJKNQTbTk; gsal BaBgTkK 4 /ma] BaCDDgHJKNQ; /myi Ba; /mi BgTkTbTkK || gsungs] DDgHTkK; gsung BaBgCJKNQTbTk || bzhag] BaBgCDDgJQ; gzhag HNTbTk; bzhag ba K; pas TkK 9 nyid la/] BgCDDgHJKNQTb; nyid las/ BaSM; snyed la/ Tk; om. TkK 10 la] CDDgHJKNQSM; las BaBgTbTk; bas TkK || sgrib] BaBgCDDgHJNSMTbTkTkK; bsgrib KQ

1 /mi bzung gtong bas gtong ma yin/] om. Ba 2 /bya ba] /byed pa TkK | te/] ste/ Ba; zin/ Bg; yin TkK 3 rtog] rtogs K | shes] shas Bg | bgegs] gags TkK 5 ma] om. Ba | gtong] stong Bg | ba] pa Ba 7 gyo’i] g.yo TkK 8 na] kyang SM | ’ching ba chen por] ’chi ba myed par Ba 9 /de bas de lung de nyid la/] om. TkK 10 sbyar] byas Ba; byar SM; bya TkK 11 rtog] rtogs Ba

1 /mi] //mi Bg 2 med] myed Ba 3 /mi] /myi Ba | bgegs] bgegs H 4 sngags] sngags Bg | go/] go BaBgHTbTkK 7 mi] myi Ba | med] myed Ba

5 /bla] [Bg395] 8 chen] [Tb57a1] 9 lung] [J39a1] 10 ’gyur/] [TkK214b1] MI NUB PA’I RGYAL MTSHAN 185

/de la de nyid grub pa med/

1 med/] [Ba216a8]; [Bg395-2]; [C43a4]; [D38b3] [Dg39b3]; [J39a1]; [K53a7]; [N53a5]; [Q36b5]; [Tb57a1]; [Tk48b6]; [TkK214b1]

1 med/ Bg reads: /lhun gyi grub pa’i thig le / bcom ldan ’das lung thams cad kyi bdag nyid/ /dpal rdo rje sems dpas gsungs pa rdzogs sho//, while TkK reads: /byang chub kyi sems mi nub pa’i rgyal mtshan lta ba nam mkha’ che rdzogs so// //rdo rje sems dpa’ nam mkha’ che u rgyan gyi slob dpon dga’ rab rdo rjes dpal a su ra ’od ldan gyi brag phug tu gsungs pa rdzogs so//. All other editions (BaCDDgHJKNQTbTk) read: /ces gsungs so/ /byang chub kyi sems kun byed rgyal po las rdo rje sems dpa’i rang bzhin mi nub pa’i rgyal mtshan gyi le’u ste sum cu pa’o// (Ba reads mi nub ba'i instead of mi nub pa'i; N reads rgyan mchan instead of rgyal mtshan; and BaJK read sum cu as sum bcu). TRANSLATIONS 186

CHAPTER 8

Translations

The five classical rDzogs chen Sems sde texts are Buddhist scriptures. What distinguishes them from so many other Buddhist texts is their form and content.

The texts are in a verse form, and their content is not philosophical but, rather, practical; they are not concerned with philosophical perspectives but “experiential instructions”—instructions that could lead one to “attain” awakening or liberation.

The doctrinal aspect of Buddhism is questioned, deconstructed, and eventually rejected, as “pristine awareness” (the nature of mind and ultimate reality), according to the texts, is only “accessible” in a direct, personal experience, and cannot be appropriated through conceptual thinking or speculative analysis. This seemingly extreme position is codified in a literary form, verse, that seems unusual for a canonical work, and the question is, then, what texts, if any, might the Sems sde be drawing from, when it comes to their form.

The texts belong to a period of great economic prosperity and cultural achievement, as at the end of the eighth century Tibetan Empire reached its apex in terms of extension, Buddhism was proclaimed the state religion of Tibet, and hundreds of Buddhist texts were translated into Tibetan. The Tibetan Empire lasted for more than 200 years, from mid-seventh to mid-ninth century, and competed for

TRANSLATIONS 187 power, influence, and control of the Silk Road, in Central Asia, with the Chinese, the Arabs, and the Turks (see Beckwith 1987). Its political and military power extended from Ch’ang-an, the capital of the T’ang Dynasty that was captured for a brief period of time in 763 (146), and the kingdom of Nan-chao, in the east, to

Ferghana and the kingdoms of the Pamir Region (Kabul, Gilgit/’Bru sha,

Baltistan/sBal ti, Wakhan, etc), in the west; from the Pāla Kingdom of Northern

India (Bengal) and the kingdoms of Nepal and Ladakh, in the south, to Turkistan, the kingdoms of the Tarim Basin (Khotan, Kucha, and Kashgar), and the major

Buddhist oasis-state of Dunhuang, in the north.

It is not known when the Tibetans first came in contact with the Buddhist teachings, however, it is possible that they encountered Buddhism as early as the fourth or the fifth centuries, considering the fact that by this time Tibet was surrounded by many countries where Buddhism was well established and flourishing (India and China, as well as the kingdoms of the Pamir Region and the

Tarim Basin). Traditional Tibetan accounts credit Srong btsan sgam po (c. 605-

650) with the introduction of Buddhism to Tibet (see Stein 2010: 191-230). He succeeded his father, gNam ri srong btsan, to the throne, and was able to consolidate the power of the newly founded dynasty by conquering the kingdoms of Zhang zhung and Nepal, and moving the capital of the Tibetan Kingdom to Ra sa (present day Lhasa), where he build a palace. He also married a Chinese and a

Nepalese princess, both pious Buddhists, and constructed two temples, Jo khang

TRANSLATIONS 188 and Ra mo che, to house the statues of the Buddha that the two princesses brought as a dowry to their marriage (Kollmar-Paulenz 2007: 311-12). Furthermore, in order for the Tibetans to be able to translate Buddhist texts, he sent his minister,

Thon mi saṃ bho ta, to Kashmir and India to develop a script for the Tibetan language. When Thon mi returned with the script, according to the sBa’ bzhed (The

Testament of sBa’), he also brought from India several Buddhist texts, including the

Ratnamegha-sūtra (Jewel Cloud Sutra) and the Saddharmapuṇḍarīka-sūtra (Lotus

Sutra), that were placed in the treasury of Phying pa castle, as “there was nobody to translate them” (Wangdu and Diemberger 2000: 26-27).

The importation of Buddhism to Tibet was a slow process that lasted for several generations, and Buddhism was firmly established only in the reign of Khri srong lde btsan (755-c. 800) when it was proclaimed the state religion of the

Tibetan Empire (c. 779). Khri srong lde btsan converted to Buddhism in 762 (when he was 20 years old), and in 763 initiated the translation of Buddhist texts on a large scale, by inviting Śāntarakṣita (c. 723-787), an erudite Indian monk from

Bengal, as well as other Buddhist masters and translators, to the royal court at

Lhasa. He then proceeded to build bSams yas, the first Buddhist monastery in

Tibet, that was consecrated in 779 and had Śāntarakṣita acting as its first abbot. At the time of the consecration, it is reported that several hundred people (including female nobles) took the monastic vows and were ordained as monks and nuns, thus establishing the first Buddhist monastic community in Tibet (Apple 2016: 132).

TRANSLATIONS 189

A Buddhist council and a committee for the translation of Buddhist scriptures were probably formed as early as 763, when the first Tibetans started to be trained in the art of translation, and an edict—“authoritative decision” (bkas bcad)—that codified the terms, norms, and rules for the translation of Buddhist texts, was issued by the young emperor. This “decision” was apparently based on the work of Śāntarakṣita (assisted by Ananta and Sang shi), who revised the texts that were already circulating in Tibet, such as, Ratnamegha and Laṅkāvatāra (Sutra on the Descent to Lanka), based on the Indian “originals” (Scherrer-Schaub 2002:

313). The edict was accompanied by a “register” (no longer extant),

Svalpavyutpatti (Smal Repertory), a Sanskrit-Tibetan lexicon that contained terms compiled from Ratnamegha and Laṅkāvatāra together with their Tibetan equivalents.

The first edict was later followed and superseded by the second edict, issued sometime in 783. A third edict, that aimed at a major “revision” of the terminology and the translation methodology of the previous edicts and registers, was issued in

814, during the reign of Khri lde srong btsan (800-815), and had affixed to it the

Mahāvyutpatti (Large Repertory) (Scherrer-Schaub 2002: 314-17). The

Mahāvyutpatti is a comprehensive lexicon that provides Tibetan equivalents for almost ten thousand Sanskrit terms and expressions (9492, to be more specific), and was compiled from the vast corpus of manuscripts that reached Tibet in the eighth century (Pagel 2007: 151-53). It became central to any translation project,

TRANSLATIONS 190 for centuries to come, as all translations that post-date it are based on its “revised” terminology.

At this early period, when most Buddhist texts were being translated from

Sanskrit, Tibetans also translated some texts from Chinese. One such text is the

Laṅkāvatāra-sūtra, that was translated from both Sanskrit and Chinese (see Suzuki

1998 [1930]: 12-15), and another work is the Vajrasamādhi-sūtra. The

Vajrasamādhi is to some extent an unusual “scripture,” and is considered by some scholars to be “apocryphal,” as the text was not composed in India, but, rather, in

Korea, around 685, and was translated into Chinese sometime between 695 and 730

(Buswell 1989: 23, 40). To cope with the Buddhist terminology, as in the case of the Svalpavyutpatti, and later the Mahāvyutpatti (the two Sanskrit-Tibetan lexicons), Tibetans also compiled a Tibetan and Chinese lexicon, that was recovered at Dunhuang and is still extant as Pelliot tibétain 1257. This glossary is based on the Old Tibetan translations that date from the eighth century, and was circulated by the Tibetans in order to “learn the Chinese equivalents” of the Tibetan terms, and, thus, help with the translation projects (Apple and Apple 2017: 68).

The earliest evidence for the translations that date to the imperial period is provided by the two extant royally decreed catalogues (dkar chag): lDan kar ma, compiled in 812 (with 737 titles); and ’Phang thang ma, compiled in 842 (with 960 titles) (on the dates see Dotson 2007). The texts that are cited were translated by teams consisting of Tibetan translators (lo tsā ba) such as sKa ba dPal brtseg, Ye

TRANSLATIONS 191 shes sde, and Cog ro kLu’i rgyal mtshan, and Indian scholars (paṇḍita) such as

Jinamitra, Dānaśīla, and Surendrabodhi (Halkias 2004: 63-65). The lDan kar ma includes sūtras (scriptures) as well as śāstras (treatises), from Ratnamegha (no.

89), Laṅkāvatāra (no. 84 and 252), and Saddharmapuṇḍarīka (no. 79), to

Tathāgatagarbha-sūtra (no. 163, Essence of the Buddha Sutra), Vajracchedikā- sūtra (no. 9, Diamond Sutra), and Vajrasamādhi-sūtra (no. 254, Sutra of the

Adamantine Absorption), to name just a few; and treatises such as

Mūlamadhyamakakārikā (no. 573, Fundamental Verses on the Middle Way) by

Nāgārjuna (2nd century), Bodhicaryāvatāra (no. 659, Entering the Path of

Enlightenment) by Śāntideva (685-763), and Madhyamakālaṃkāra (no. 578, The

Ornament of the Middle Way) by Śāntarakṣita (Lalou 1953; Herrmann-Pfandt

2008).

What all these texts have in common is that they all contain verses, in one form or another; the sūtras contain sections (or passages) that are in both prose and verse, and the treatises are entirely in verse. A Buddhist text (whether in Sanskrit or

Chinese) may contain prose, verse, or prose and verse, and scholars are still debating on how a particular text was originally composed. Was it composed in prose and the verses were added later, to summarize the prose? Or, perhaps, it was composed in a verse form, and the prose was added later to provide a narrative that can explain the verses. Present-day research in textual analysis and literary genres seems to suggest that in the ancient times prose form had no preeminence over the

TRANSLATIONS 192 verse form, and while some texts were composed in prose, others were not—they were originally composed in verse.

Jonathan Silk (2009: 184), in his study on the nature of the verses in the

Kāśyapaparivarta-sūtra (The Sutra of Assembled Treasures), a text that existed as early as the second century CE, suggests that the majority of the verses preserved in the sūtra were probably added later, in the following centuries, and only 10 verses

(out of a total of 158) are original. This is an interesting case as the text is preserved in two Tibetan translations: one in bKa’ ’gyur that has verses (9th century); and one at Dunhuang (fragmentary) that does not, is a prose-only version

(mid-to-late 8th century) (Apple 2017). It is possible that Kāśyapaparivarta may have been composed entirely in prose; however, there is evidence to suggest that another text, the Laṅkāvatāra-sūtra, is based on “source materials” that were written in a verse form.

According to Takasaki (1980: 339), Laṅkāvatāra (compiled in the 4th century), is a composite text that consists of two parts. One part is a mix of verse

(432) and prose, the main body of the text, and another part is exclusively in verse

(884), named Sagāthakam. The two parts have 220 verses in common, and based on textual analysis it would seem that in the “beginning” the two parts were two different texts, both in verse form, that were based on the same source materials.

These source materials, that is, the “original” verses, were composed and transmitted within a group of Mahāyāna practitioners to which the author, or the

TRANSLATIONS 193 authors, of the Laṅkāvatāra belonged. The verses were “loosely bound together,” and any practitioner was free to pick up a “certain verse” (or “several verses in a series”), and provide it with a prose explanation (Takasaki 1980: 345).

Development took place in both texts independently, and verses were added to the collections until the two texts were finally combined into one volume, which came to be known as the Laṅkāvatārasūtraṃ Sagāthakam (The Laṅkāvatāra-sūtra that includes the Sagāthakam) (346-47).

The task of translating works that are written in a verse form and have a regular metre is not an easy task and the Tibetans in the second half of the eighth century CE had as many problems in translating them as the Chinese had in the second century CE. Jan Nattier (2008), in his paper presented at the American

Oriental Society, “Rhymed Verse, Unrhymed Verse, or No Verse at All,” points out that both An Shigao (fl. 148-170) and Lokakṣema (fl. 165-185), the earliest known translators of Indian Buddhist texts into Chinese, had problems with the

Indian metric system and translated all verses as prose, with no attempt to

“introduce any metric regularity” or “establish a metric patter” in the Chinese texts.

In order to identify passages and sections that were versified in the original, they used a “formula,” such as, “then the Buddha spoke in verse.” It was only in the third century CE that a Chinese metrically regulated five- and four-character lines were used to translate Sanskrit Buddhist verses (1-8).

TRANSLATIONS 194

Initially, Tibetan translators of Sanskrit verses tried to imitate and strictly follow the Sanskrit syntax, if not the metre. However, there were problems from the beginning as this principle to translated line for line and word for word—to strictly maintain the order of words of the original—forced them, many a time, to ignore basic grammatical rules that govern the Tibetan language, and disregard any attempt to establish metric regularity in the Tibetan texts. In the Tibetan language, to render a sentence comprehensible, it is necessary to place the verb of the sentence at the end of the sentence. As the translators felt uneasy about departing from the Sanskrit manuscripts, in some cases the verb was placed in the middle of a verse (or half-verse), thus, leading to “syntactical monstrosities” and unintelligible passages (Zimmermann 1998: 48-49). In the case of Sanskrit metre, Tibetans were well aware that the Sanskrit metrical system is incompatible with the Tibetan language, and cannot be adapted to it, as Hahn (2007) points out in his interesting study of the early Tibetan translations, “Striving for Perfection: On the Various

Ways of Translating Sanskrit into Tibetan,” and, yet, the principle of adhering strictly to the Sanskrit syntax, that they so faithfully followed, made it impractical to establish (or follow) a metrical pattern.

One problem associated with any attempt of adapting the Sanskrit metre to the Tibetan language, even if one does not follow the Sanskrit syntax, is that the

Tibetan language is primarily bisyllabic (Tucci 1966: 16; Sørensen 1990: 13),

TRANSLATIONS 195 while the Sanskrit is multisyllabic. To illustrate the point, the term “virtue” is a good example:

Skt. kalyāṇa (क쥍याण), Tib. dge ba (+#{-0), Ch. shàn (善)

According to the Mahāvyutpatti (no. 3554), the Tibetan translation for the Sanskrit term kalyāṇa is dge ba (Mahoney 2004), and the Chinese term for the Tibetan dge’ ba’ (dge’ ba’ is an Old Tibetan spelling for dge ba), based on Pelliot tibétain 1257, is shàn (Apple and Apple 2017: 112). As such, the three terms are “equivalent,” but the Sanskrit term has three syllables, the Tibetan has two syllables, and the Chinese has one character (as the Chinese language is character-based), and this becomes a problem in any metre when the number of syllables per line is strictly enforced.

The Sanskrit metrical system has a great number of metres, and all are based on short (⏑) syllables (vowels a, i and u), and long (━) syllables (vowels e, o,

ā, ī and ū). If a short syllable is followed by the anusvāra (ṃ) or a cluster of consonants (two or more), it also is considered to be a long syllable. Furthermore, there is some flexibility in the spelling of words in order to meet the requirement of a particular metre, by allowing consonants and the anusvāra to be added or dropped from a word (e.g., sādhu = sādhuṃ), and the short vowels to be replaced by the long ones, and vice versa (a = ā, o; i = ī, e; u = ū) (Edgerton 1946). Metres are either of the akṣara type and are counted by syllables, or of the mātrā type and

TRANSLATIONS 196

are counted by morae [1 (⏑) = 1 mora; 1 (━) = 2 morae]. The akṣara metric types are based on a combination of eight gaṇas (metrical units of three syllables each): ya-gaṇa (⏑ ━ ━), bha-gaṇa (━ ⏑ ⏑), na-gaṇa (⏑ ⏑ ⏑), ra-gaṇa (━ ⏑ ━), ja-gaṇa (⏑

━ ⏑), ma-gaṇa (━ ━ ━), ta-gaṇa (━ ━ ⏑), and sa-gaṇa (⏑ ⏑ ━); and the mātrā metric types are based on five gaṇas (metrical units of four morae each): bha-gaṇa (━ ⏑

⏑), na-gaṇa (⏑ ⏑ ⏑ ⏑), ja-gaṇa (⏑ ━ ⏑), ma-gaṇa (━ ━), and sa-gaṇa (⏑ ⏑ ━).

Metres define how many syllables or morae a line can have, what type of gaṇa and syllable can be used, in what line, and where in the line they can be used (Upajāti:

⏓ ━ ⏑ | ━ ━ ⏑ | ⏑ ━ ⏑ | ━ ⏓ ||). The best known and most widely used metres in the Buddhist texts are the syllable-based, Anuṣṭubh (Śloka) (4 x 8 syllables),

Upajāti (4 x 11 syllables), and Triṣṭubh (4 x 11 syllables); and the morae-based,

Āryā (two halves, 30 + 27 morae) (Hahn 2014; Ollett 2013).

Because of the Sanskrit syntax, script, and metrical requirements, and the fact that the Tibetan script has only five vowels (a, i, u, e and o) that are neither short nor long (stressed or unstressed), the Tibetan translators had a major hurdle to overcome. Some of them (early translators of the Tathāgatagarbha-sūtra), focused on the Sanskrit syntax and the quatrain format, a verse by verse approach, and ignored the number of syllables in a line or a verse (Zimmermann 2002: 275). They managed to produce a translation that was “faithful” to the original, but one that was neither easy to read in Tibetan, nor was it “elegant” in terms of continuity of

TRANSLATIONS 197

lines and sentences. Others (later translators of the Subhāṣitaratna), tried to maintain the order of words as found in the original and a fix number of syllables per line (7 in this case), and produced a translation that is largely unintelligible in terms of its syntax, and a “riddle,” according to Hahn (2007), on “how could they assume that the work would be intelligible” to anyone who reads it (124-27).

The five Sems sde texts are nothing like some of these early Tibetan translations, and for a good reason: they are written by the Tibetans who are not trying to imitate or follow a Sanskrit work in terms of its syntax and metrical system. Lines and sentences of the texts flow naturally, and that is why, most probably, the texts were appealing to the Tibetan readers in the first place. Tibetan metres are based on “units” and “accents.” A “metrical unit” (foot) provides the form of a metre, and consists of either 1-syllable (monosyllabic), 2-syllables

(bisyllabic), or 3-syllables (trisyllabic); while the “metrical accent” determines the beat and the rhythm of the metre, and always falls on the first syllable of each unit.

A 4-syllable unit, even though possible, is unusual and extremely rare (Beyer 1992:

421). When the units are combined, they form a metrical pattern, a metre, which can consist of two or more units (e.g., 1 + 2 + 2 + 3), that can also be expressed in

“feet” (e.g. - / - - / - - / - - -) (see Jinpa and Elsner 2000: 13-14). Don grub rgyal, in his study of the Tibetan metrical system, Bod kyi mgur glu, has documented thirty metres (Pang 2011: 105-107; Sujata 2005: 213-31), and his list is not exhaustive, as

TRANSLATIONS 198

he takes into consideration only combinations of up to six units, and more combinations are still possible. Some Western scholars have attempted to represent the “metrical accent” in a graphical form, such as, Beyer (1992: 408-23), who depicts syllables in terms of stressed (1) in a “strong position” (1st syllable), and unstressed (0) in “weak positions” (2nd and 3rd syllables) (e.g., 1-10101010010).

This approach may have its merits; nevertheless, it is debatable whether it is more practical, or even more accurate than that based on “units.”

A Tibetan verse consists of metrical lines that can vary from 4 to more than

17 syllables, predominantly of the same length, which are grouped together in stanzas of two (couplets), three, four (quatrains), or more lines per verse. The Sems sde texts, as most other Tibetan Buddhist texts that contain verses, are written in odd numbers of syllables per line, from 7 (2 + 2 + 3) to 13 (2 + 2 + 2 + 2 + 2 + 3).

In terms of their form and metrical pattern, we have the following structures: Rig pa’i khu byug consists of 3 couplets, of 7-syllables per line; rTsal chen sprugs pa has 5 quatrains, of 9-syllables per line (with the exception of 3.3 and 4.1 in 11 syll.; and 4.2 in 13 syll.); Khyung chen lding ba has 28 verses, with stanzas ranging from

4 to 6 lines (1-23, 25 and 28 in 4 ll.; 27 in 5 ll.; and 24 and 26 in 6 ll.), that consist of a variable number of syllables per line (1-14 in 11 syll.; 15-16 in 9 syll.; and 17-

28 in 7 syll.); rDo la gser zhun is primarily written in a versified prose (pars. 1-10 are written in a combination of 9, 11, and 13 syll.), with a single verse attached at

TRANSLATIONS 199

the end of the text (6 ll. in 7 syll.); and Mi nub pa’i rgyal mtshan consists of 55 quatrains of 7-syllables per line.

The Tibetan metrical system seems to be based on the “beat” and the

“rhythm” of the Tibetan folk songs and poems. It does not have much in common with the Sanskrit form, and it most definitely is not a “Tibetanized Sanskrit form,” as some scholars tend to suggest (Sujata 2005: 112, n. 1). It is likely that while some Tibetans were translating Sanskrit Buddhist texts that had verses (or were entirely in verse), as best as they could, others were experimenting with metrical patterns that were based on the Tibetan language, and the Tibetan syntax. Literary documents from Dunhuang testify to the popularity of the versified form, and the recovered materials abound in written poems, hymns and odes, as well as traditional songs, written in a bisyllabic (2 + 2 + 2) and trisyllabic (3 + 3) six- syllable metre (Beyer 1992: 409, 412-13). In order to set the Buddhist texts apart from the folk literature, the Tibetans opted to reserve metres of odd number of syllables for religious works, and this soon became the fixed norm in versification

(Sørensen 1990: 13).

The rDzogs chen Sems sde texts may well have been some of the early

Tibetan works, written in a verse form, that led the Tibetan king Khri lde srong btsan to proclaim the third “imperial edict” in regard to the translation terminology and methodology. It must have been puzzling for the Tibetans to read a text

TRANSLATIONS 200

translated from the “sacred” language of Sanskrit and find out that the text was mostly unintelligible. The “edict” and the Mahāvyutpatti were accompanied by a compendium, the sGra sbyor bam po gnyis pa ([Principles of] Literary

Composition in Two Fascicules), which provides specific guidelines on how to translate from Sanskrit. It states: “if good [Tibetan] and intelligibility can occur only by changing [the Sanskrit syntax],” then one needs to make the change and

“translate to the full advantage [of Tibetan] by [having in view] both the expression and the meaning” (Scherrer-Schaub 1999: 72). It is only after they adopted, and followed these guidelines, that the Tibetans were able to produce some of their best translations, that could rival the rDzogs chen Sems sde texts, and the works of such great masters of composition as Nāgārjuna, Śāntideva, and Śāntarakṣita.

In terms of their form, there might not have been a specific work that the five Sems sde texts were drawing from, but they do belong to a rich Buddhist tradition of versification that for centuries was using verse as a literary medium to transmit the Buddhist teachings, and in that sense they are no exception. As canonical texts, they contain the original germ of the rDzogs chen teachings, and serve as the basic structure on which later texts are built, expanded, and elaborated.

In my translation I attempted to follow the sGra sbyor’s advice and make the texts as readable as possible: they speak for themselves (if only we would listen to them).

THE CUCKOO’S SONG OF AWARENESS 201

8.1 The Cuckoo’s Song of Awareness (Rig pa’i khu byug)

[1] The true nature of infinite diversity is non-dual,

Since each phenomenon is free of conceptual elaboration.

[2] The so-called state of “just as-it-is” is non-discursive,

Since all manifested forms are ultimately self-perfected.

[3] As everything is already accomplished, by giving up the affliction of striving

One finds oneself naturally abiding in the effortless state.

THE GREAT POTENTIALITY 202

8.2 The Great Potentiality (rTsal chen sprugs pa)

[1] From the very beginning, the sphere of reality consists of everything that is

outer and inner.

In this dimension of the original perfect purity

There is no distinction between the Buddhas and the sentient beings.

So, how could there be anything that needs to be corrected with paths and

antidotes?

[2] As there is no desire or striving, there is nothing to be attained;

The state of reality itself is spontaneously self-perfected and free from activity.

In the pure field of reality, concepts and analyses are non-dual.

So, how could this dimension be conditioned by the behavior of some foolish

people and their wrong views?

[3] The non-dual great bliss can be experienced by all sentient beings.

Even the wrong path as conceived by the deluded followers

Is not different from the universal path, as explained above.

Whoever understands this equality is the lord of the Buddhas.

THE GREAT POTENTIALITY 203

[4] Thinking in terms of “I” and “mine” is the mistaken path of the heretics.

As the foolish [practitioners] are deceived, they enter the path of conceptual

activities

And never reach their goal, or attain any understanding.

So, how could they attain reality by searching for reality as-it-is?

[5] If one follows the teachings of the monkey-like masters who are devoid of

authentic knowledge,

One will certainly end up on the wrong path that is conditioned by concepts.

The one who can extract gold from minerals

Is an authentic master whose teachings are most precious;

A treasure worthy to be bought for any price.

THE GREAT GARUDA IN FLIGHT 204

8.3 The Great Garuda in Flight (Khyung chen lding ba)

[1] The teaching of the non-discursive state is without abode or support.

To generate a dimension of subtle aspects and intentions

Is a pointless investigation—a conceptual meditation on the manifestation of truth;

Self-originated pristine awareness abides just as-it-is, entirely free of discursive

thinking.

[2] Beyond action, it does not dwell as an object, nor does it need an antidote to be

corrected.

Even if one were to seek the realm of the fundamental essence by means of diverse

phenomena,

One can only enjoy it through its non-conceptual aspects;

As the essence manifests naturally, the true nature of reality cannot be found

anywhere else.

[3] Indivisible and beyond the ten directions is the ultimate real condition.

Pristine awareness, that self-originates and does not abide in anything,

Is the very essence of the sublime, non-conceptual direct experience.

THE GREAT GARUDA IN FLIGHT 205

Those who enter this pure path will most certainly attain the supreme equality.

[4] As this state is unchanging and free of conditions, there is nothing to be

attached to;

Likewise, as there is no object to be grasped, there is no place for the grasping

mind.

Those who meditate on causes to reach a direct experience

Will never attain equality if they are happily attached to meditation.

[5] As the one dimension is all-pervading, there is nothing to be added to it,

As this dimension is endless, there is nothing to be taken away from it.

There is no other hidden state that dwells beyond the manifest reality;

The dimension of the great self-origination always abides just as-it-is.

[6] The eye which sees that there is no object to be seen sees the wonder;

It transcends all definitions as there is nothing specific to be heard.

What is right and what is wrong are always mixed up and equal;

A higher dimension named “ultimate reality” cannot even be described.

[7] The path of perfect purity cannot be conditioned by intellectual illusions;

The self-originated pristine awareness transcends the limitations of words.

THE GREAT GARUDA IN FLIGHT 206

In the direct experience of the timeless perfect essence

Conceptual thought arises just like a shadow of a material form.

[8] Without existing, it is not non-existent; inner reality manifests as an absence.

Emptiness is not empty; it abides in the dimension of emptiness.

Awareness arises from the nature of the open space:

Without even wishing for it, one attains the bliss that is already free from any

activity.

[9] Pristine awareness cannot manifest as an object that can be conceived.

By establishing a mind attached to ancient sages,

One ends up utterly tormented by striving and struggles.

Omniscience arises only when one enters the path of the intrinsic reality.

[10] For those who conceptualize the authentic condition, meditation becomes a

mere reflection;

They become sick of attachment by desiring the great bliss.

If they do not apply the medicine of abiding in the state of unchanging equality,

Then, even the cause of progress to higher realms becomes infected by delusions.

[11] The worst sickness of all those who enter a path when there is no path

THE GREAT GARUDA IN FLIGHT 207

Is that they reach for a goal, just like a deer pursuing a mirage.

The goal is not an object that can be attained, nor does it arise from the threefold

world;

Even the state of dependence on the ten stages is an obstacle to perfect purity.

[12] Devoid of concepts, ever-present pristine awareness

Is like a precious jewel that arises amidst the spiritual friends;

Independent of change and without apprehending anything,

By its very nature, it fulfills all desires.

[13] Examined, there is nothing to be found; if left as-it-is it gives rise to great

qualities.

It gratifies all forms of needs, even though it is not visible;

It is a master teacher devoid of self and other—a precious treasure.

As the realm of perfection, it is revealed through selfless compassion.

[14] Unmoving, there is nothing to be found within:

It is not an object that one can get attached to, or that can be apprehended.

Selfless-compassion is not a state that one can enter into, or that can emerge;

It is ever present, without arising and without the illusion of otherness.

THE GREAT GARUDA IN FLIGHT 208

[15] Those who desire bliss turn their backs on bliss;

As bliss is already present, they seek out bliss by bliss.

Mistaken about perfect purity, they aim for an outer object;

However, these grasping perceivers never experience enlightenment.

[16] As there is no enlightenment, even the name “enlightenment” does not exist;

It is an error to assign a name in order to point out enlightenment.

Hoping to attain enlightenment from others is a mistaken path

As there is not even the slightest trace of a teaching on the formless experience.

[17] Already peaceful and free of attachment,

Immaterial and entirely formless,

The nature of the great nectar

Cannot be grasped by concepts.

[18] The vast, grand, incomparable teaching

Is an antidote to everything that is small;

When equality is established as the object of greatness,

It transcends the concepts of lower and higher.

[19] The teaching, the expanse, the observation, and the appearance,

THE GREAT GARUDA IN FLIGHT 209

Are like scenes created by a magician;

By obscuring the manifestation of the pristine awareness,

They lead to a further rebirth.

[20] This is the supreme vehicle:

A fundamental nature which posits everything and does not hold to anything;

It cannot be grasped, nor can it be desired,

And it does not produce even the slightest experience of eagerness.

[21] Just like the great garuda that is soaring in the open space,

Without elaboration and any comprehension,

It does not aim for anything, nor is it afraid to lose anything.

[22] Ever present, like an ocean,

It gives rise to all phenomena.

Its qualities are similar to that of the open space;

They have no place where to originate.

[23] Suddenly, the essence of enlightenment

Manifests as the supreme contemplation;

Its appearance is similar to that of a great ocean

THE GREAT GARUDA IN FLIGHT 210

As its non-conceptual state is as wide as the infinite space.

[24] The dimension of the Samantabhadra (the all-good)

Is unborn and unchanging;

The twelve links of causes and conditions

Are nothing but explanations conceived by those who are afraid

For the benefit of those who are deluded.

Accordingly, the wise should keep this in mind and remember it.

[25] Again, even though the six classes of beings do appear

One should recognize this as the original path;

Those endowed with compassion, who pursue desires,

May also pursue perfect purity by any means whatsoever.

[26] Butchers, prostitutes, and those who have committed the five deadly sins,

As well as those involved in the perverted actions, are abandoned by the world;

But, the pious, the fully perfected ones, know

That even these acts are not that different from the great bliss.

Thus, they know all phenomena

Through the inherent nature of all phenomena.

THE GREAT GARUDA IN FLIGHT 211

[27] Searching for the true nature by depending on the true nature,

It would be like the sky searching for the sky;

And expecting to discover the true nature by depending on others,

It would be similar to extinguishing fire with fire.

These, are very difficult things to do indeed.

[28] This pure essence of the non-discursive state

Is not hidden in one’s personal experience;

All those who live without grasping, in accord with perfect purity,

Always dwell in the intrinsic reality.

THE REFINING OF GOLD FROM ORE 212

8.4 The Refining of Gold from Ore (rDo la gser zhun)

[1] The mind of perfect purity, indescribable and beyond all thoughts or expressions, is much praised as the light of the teachers. Being the essence of every teaching, it is the embodiment of the youthful Mañjuśrī. It abides in the natural bliss, spontaneously self-perfected and free from activities.

[2] It is taught to be the path of every form of liberation; the common basis of all the innumerable practices, such as the moral discipline, and so forth. It is the mother of all the buddhas, [the enlightened ones], and the universal path to liberation. If it did not exist, there would be no path, therefore, this is the supreme path to a complete liberation.

[3] Hard to follow and difficult to understand, the universal path is beyond thinking and not thinking. Non-abiding, imperceptible, and free of concepts, it transcends reflection and deliberation. Words cannot express it. Without shape and color, it is not an object that can be experienced by the senses. [The universal path] is difficult to investigate and to explain as there is nothing that can be said about it.

THE REFINING OF GOLD FROM ORE 213

[4] Those who follow sages of the past end up being afflicted by the disease of attachment to the path of meditation. When they base themselves on the final words and the precepts of the teachers, they merely follow streams of thoughts and are similar to those who pursue a mirage. It would be a mistake to try to define the true reality: the true path cannot be expressed in words.

[5] [As concepts], pure and impure are integrated and inseparable. The light of purity that shines unobstructed, and the ignorance that does not distinguish between forms, are both beyond thinking. Pristine awareness dwells as the supreme meditation, obscured [only] by the unchanging state of its own nature.

[6] [Pristine awareness] is the eye that directly sees that there is nothing to be seen. Because of that, it is called “the eye of omniscience.” Wide and broad, with no limit and center, it is the very essence that dwells as the supreme equality, not accepting or rejecting anything.

[7] The mind and its karmic predispositions are mixed and non-dual. All subjectively conceived phenomena appear as ornaments [of the spontaneously self- perfected state], neither renounced nor abandoned. One rejoices in this state by the method of non-thinking.

THE REFINING OF GOLD FROM ORE 214

[8] When one enters “into” the path of perfect purity, one attains the mastery of the supreme equality, [even if he/she has committed] negative deeds that are entirely abandoned by all as harmful, such as the five obscurations and the five deadly sins. On this path, nothing is avoided or given up, like mental formations, and so forth.

[9] Those who rely on logic and the two traditional forms of benefit [to oneself and to others], and apply the doctrinal views to stabilize the three meditative contemplations, remain deluded. They deviate from the indisputable teaching: to abide effortlessly, in the spontaneously self-perfected bliss, free from activities.

[10] The essence of the great, self-originated pristine awareness is unchangeable and unshakable, beyond all designations. It is the nectar of the natural perfection that eliminates the misery of exertion. “Abide just like that in the already accomplished dimension, free from activities!”

[11] As all phenomena have the nature of the mind of perfect purity, the great sphere [without limitations],

there is neither elaboration nor abbreviation,

neither arising nor ceasing;

THE REFINING OF GOLD FROM ORE 215

unobstructed, [pristine awareness] abides just like that.

This intrinsic reality, free of discursive thinking,

abides timelessly, like space,

beyond the conceptual realm and the conceptual designations.

THE UN-WANING VICTORY BANNER 216

8.5 The Un-waning Victory Banner (Mi nub pa’i rgyal mtshan)

[1] The great space of Vajrasattva (the indestructible nature of mind)

Is the ever-perfect dimension of reality.

As it is all-liberating, this pure and universal sphere

Is unproduced, unobstructed, and inconceivable.

[2] As its true meaning is already accomplished through love,

Great compassion is not practiced (either);

Supremely vast and deep,

Its attainments are beyond praise.

[3] All phenomena, immutable in their natural condition,

Are liberated through deedless liberation.

The self-originated pristine awareness is free of striving:

Self-liberated, it also teaches the path of liberation.

[4] The great elements, as the Bhagavan,

Abide intrinsically in all beings.

THE UN-WANING VICTORY BANNER 217

Even though false notions are conceived,

Liberation is self-originated and not dependent on others.

[5] The great pristine awareness is difficult to grasp:

It depends on wisdom and is realized through it.

Other-dependent is just a label;

The true bliss spontaneously self-originates.

[6] The great miracles are not difficult to see:

The various powers and attainments,

Subtly understood as the natural state,

Instantly self-manifest.

[7] As the true nature is beyond manifestation,

It is to be contemplated by letting it be.

If one searches for it in different places,

It will never be found.

[8] This unsurpassed and secret true nature

Cannot be learned by hearing about it.

Likewise, the faculty of speech

THE UN-WANING VICTORY BANNER 218

Cannot express it.

[9] The suffering of beings is the mind of perfect purity.

When it manifests as fully perfected,

Unmoving and unstirring,

It abides equally in all, just like the infinite space.

[10] The equality in all distinctions

Is sometimes designated as “karma.”

However, as long as one is under the sway of karma

Self-originated pristine awareness does not exist.

[11] The cause itself, as the indestructible condition,

Never having been born it cannot be destroyed.

The timeless, fundamental state of perfect purity,

The immutable sphere, cannot be disclosed by thought.

[12] Contemplation of the great attainment

Is a non-reflective contemplation.

Beyond the experience of reflection and purification,

Pristine awareness arises from the discursive thought itself.

THE UN-WANING VICTORY BANNER 219

[13] Coining the expression “a subtle gate,”

Some seek out the path devoid of mental events

By grasping at emptiness in a silent place.

However, when examined, this is a conceptual meditation.

[14] By applying designations such as “cause” and “effect,”

Some believe that they can overcome both virtue and vice,

And transcend this world.

However, they only generate a great complacency for acceptance and rejection.

[15] “Attachment” and “detachment” are only words,

Just like something in-between—like an echo,

While pleasure and pain have the same cause:

Vajrasattva has proclaimed to all sentient beings.

[16] Attachment, anger, and ignorance

Arise from the path of the great enlightenment;

While the five ordinary enjoyments

Adorn the true nature of existence—he also said.

THE UN-WANING VICTORY BANNER 220

[17] As both space and the concept of space are un-originated,

The conceptual thought itself is like the open space;

When space like intention is examined without attachment,

The manifest self-benefit is as great as the open space.

[18] Non-conceptual equality is the fundamental dimension of reality,

Like the moon reflected in the water it cannot be grasped;

Samantabhadra’s manifest energy,

Is revealed as the profound system of vowels and consonants.

[19] With A and TA as adornments,

All phenomena manifest as PA.

The sphere of activity of the transient world

Arises as the profound teaching of the Buddha.

[20] How wonderful! Buddha’s sphere of experience

Cannot be found through seeking and striving;

As it is not an object of the sensory perception,

To search for it is like a blind man reaching for the sky.

[21] The path of purification with higher and higher levels

THE UN-WANING VICTORY BANNER 221

Is not in accordance with the teachings of no-action;

If there is a path conducive to progress,

It is beyond attainment, like the end of the sky.

[22] Since the authentic condition is just like that,

“That” is taught, so that one realizes it as “that.”

As “it” is the very essence,

Everything arises from “it.” How wonderful!

[23] Present time and past time

Abide in the state of the authentic condition;

This, likewise, is also the path to “that,”

And the inherent reality of “that.”

[24] The here and now is the universal path for all,

It manifests as both the moon and its reflection.

As its nature is all-pervading,

It cannot be realized by those looking for specific features.

[25] Present pleasure and future pleasure

Arise from a direct perception and its consequences;

THE UN-WANING VICTORY BANNER 222

As this is a defect of the conceptual mind,

One should not rely on it.

[26] The three times are one without any distinction;

The past never arisen and the future never arising.

Because the fundamental dimension of reality pervades everything,

It rests in the natural condition, in the greatness of the great.

[27] Religious practice in the three-fold world

Is just a name and a magical illusion;

Even the great place of a universal monarch

Is a dwelling place conditioned by illusions.

[28] Practices of those who depend on time

Can never reach an outcome in time:

If one’s practice does not transcend desire,

It can only be described as an “empty” action.

[29] Beyond any particularity whatsoever, as one,

A practitioner’s path is like a bird’s flight across the sky;

In the un-originated and unborn essence,

THE UN-WANING VICTORY BANNER 223

How could there be any signs of his passing?

[30] Inner and outer are one, the outside itself is the inside

So there is no hidden depth to discover.

Worldly existence is just a label caused by a mistaken view;

Consequently, one is separate from the equality of contemplation.

[31] As to the outer and inner designations,

They abide in the elements and the five aggregates;

Since they are never separated from the three times,

It is improper to apply such names and designations.

[32] Immovable, it is the seal [of a pure being].

Unshakable, it is pristine awareness.

When not grasping anything, there is no self;

When not rejecting anything, there is equality that transcends words.

[33] Whatever, and whenever comes forth,

All beings and their behavior, originates in the state of pure mind.

The distinction between man and woman

Has never been taught by the King of equality.

THE UN-WANING VICTORY BANNER 224

[34] By means of forced and wrathful conduct

There is nothing at all to be attained;

When one joins the A to the PA,

One aims to experience the bliss of illusion.

[35] As the ultimate nature cannot be defined,

It may appear in the way it is perceived;

Striving for and delighting in appearances

Is a great defect and an obstacle [to pristine awareness].

[36] [To claim that] the door to all the aspects of enlightenment

Is by meditating on a deity, is like reaching for the moon in the water;

Even if one attains a state that is undefiled and detached,

Such a meditation is the experience of foolish people.

[37] Even if one visualizes oneself with wrathful attributes in a mandala,

By taking the form of the lord of wrath, Heruka,

And recites the seed syllables of the mantra,

One will not experience the tranquility of reality itself.

THE UN-WANING VICTORY BANNER 225

[38] By being under the power of afflictions,

As soon as one cuts off the top of the palm tree

Or burns the seeds by fire,

One will not fall under their influence: so it is taught.

[39] Each of the hundreds and thousands of teachings

Produces its characteristic flavor—whichever one practices;

But, as the real condition is without attributes,

It has no place to arise from.

[40] Dwelling free of words,

That yogi is very fortunate;

By not distinguishing between self and others,

One delights in the self-perfected magical illusions.

[41] Without a remainder, it is fully perfected.

Unchanging, it always remains whole.

The boundless equality is like the open space,

It does not depend on anything else.

[42] The great bliss arises spontaneously self-perfected

THE UN-WANING VICTORY BANNER 226

From the incomparable pristine awareness;

Solely and exclusively as pure presence,

Reality cannot originate from anything else.

[43] It is easy and difficult, difficult because it is [too] easy.

Not apparent, it pervades everything.

When practiced only as “this is it,” it becomes a mere name

And even the great Vajrasattva cannot show it.

[44] This wonderful and extraordinary manifestation

Abides like space, beyond activities;

Out of non-conceptual ignorance,

It arises naturally and spontaneously.

[45] This is the only path for all,

Naturally present in all living beings;

Ignorant people, conditioned by delusions,

Search for a medicine although the mind’s own nature is the cure.

[46] In the field of understanding lies the great bliss,

As the pure dimension of the world;

THE UN-WANING VICTORY BANNER 227

When the light of the cardinal directions gathers,

The four main directions, the intermediary one, and the above and below are

perfected.

[47] From different colors of the rainbow light,

The attributes of the [Buddha] families directly manifest;

Likewise, the animate and the inanimate world

Originates from the five elements.

[48] Past, present, and future,

As conventional designations, are mere labels;

To comprehend that there is neither origination nor cessation,

It is to know reality as-it-is—the great unity of the three times.

[49] In sameness, there are no successive stages.

In oneness, there is no progressive dedication.

Even if one prepares the offerings as the adornments,

There is nothing to distribute, as they already exist naturally.

[50] Being self-perfected, there is nothing to be desired.

Being pure from the beginning, it is already nectar.

THE UN-WANING VICTORY BANNER 228

By the twelve sense bases, in particular,

There is nothing to be grasped, however noble the intention.

[51] Mind, the benefactor of the offerings,

Manifests various phenomena through the power of the gaze;

Attainment that derives from seeing,

Is the perfect, non-conceptual primordial awareness—the reality itself.

[52] Grasping even for an instant is the union;

Blissful satisfaction is the commitment.

Performing the dance movements of the skillful-means

Is the offering of the non-dual union.

[53] Letting go without grasping is the sacrificial offering,

As all activities are already completed without acting;

When non-conceptual pristine awareness removes obstacles,

Remaining silent in meditation is the speech of mantra.

[54] Generosity and the offering to the Guru,

As well as any other meritorious act,

If performed without an unwavering detachment,

THE UN-WANING VICTORY BANNER 229

All become a great bondage.

[55] Therefore, the great meaning of the teaching

Is veiled if it is structured;

If conceptualized,

Reality itself can never be attained.

CONCLUSION 230

CHAPTER 9

Conclusion

The primary aim of this dissertation has been to call into question the ahistorical origin of the rDzogs chen Sems sde tradition, by situating it in its proper historical context. It represents an attempt to put the study of the Sems sde teachings on a firmer critical footing by dispensing with some of the assumptions that inform the modern history of this ancient Tibetan Buddhist tradition. The modern conception of the rDzogs chen, as an object of historical investigation, has evolved out of the traditional legendary accounts that were designed to provide some “legitimacy” to an emerging tradition, by arguing that rDzogs chen originated in India and that the earliest texts preserved in Tibetan are translations from

Sanskrit. That being the case, I have challenged the “mythical origin” based on the fact that this particular claim is not grounded on any historico-critical evidence. It derives, rather, from a conception of the early rDzogs chen that post-dates the formative period of the Sems sde tradition, and rests on the hagiographies of mythological masters, such as, dGa’ rab rdo rje.

This fictive model of an Indic “origin,” even though it lacked a historical basis, did function as a legitimizing tool in a period when texts and teachings produced in India, or by Indians, were considered as “authentic,” while those

CONCLUSION 231 produced by Tibetans, as “inauthentic” (Davidson 2002: 203-205). It also allowed the rDzogs chen scholars to argue that the tradition was deeply rooted in the past, and, thus, to reject the claims that it was a Tibetan innovation. The problem with this model is that neither rDzogs chen Sems sde teachings nor texts have been found in India. Rat na gling pa (1403-1479), in his attempt to “explain” the absence of Sanskrit rDzogs chen texts, suggests that it was Vimalamitra (8th century) who removed “all rDzogs chen texts” that were hidden by the Indian pundits under a

“vase-shaped pillar” at Vajrāsana (Bodh Gaya), when he departed India for Tibet.

He also claims that the Indian pundits had “no authority to explain the texts,” and were not teaching them, as it was not “the right time” to do so (Neumaier-Dargyay

1992: 21-22). This is the kind of “evidence” that scholars have to rely on in order to support their claim that the rDzogs chen tradition “originated” in India, but, somehow, no rDzogs chen text has ever been found in India, and no rDzogs chen teaching is known to have been taught in India.

I have argued that, perhaps, it is time to move away from the ahistorical

“origin,” and the mythological accounts of textual transmission and translation, and acknowledge that the rDzogs chen Sems sde tradition is a Tibetan tradition. The primary impetus for its emergence may not have been the Chinese and/or the Indian textual sources, but the oral teachings of the Tibetan Ch’an and the Tibetan

Mahāyoga that were circulating in Tibet. As the traditions “emerged” and developed at approximately the same time, late eighth and early ninth century, it

CONCLUSION 232 can be said that the three traditions co-emerged. This would explain the similarities and affinities between them. As oral traditions, they all shared some common terminology, propounded some form of sudden enlightenment, espoused a non-dual view of reality, and gave primacy to the non-conceptual. In other words, at the early stages of emergence and development, the boundaries between the Sems sde, the Tibetan Ch’an, and the Tibetan Mahāyoga were not fixed, and the distinctions between their teachings were not firmly established. This was an interlinked set of traditions, and the relationship between them was one of mutual interaction and interdependence.

By providing accessible translations of the five “earlier” Sems sde texts, and situating them in their proper literary context, the dissertation also calls into question the idea that the rDzogs chen Sems sde texts are repositories of rDzogs chen doctrine and philosophy. The Sems sde texts are practical in their concerns and are centered on the experience of awareness; awareness understood as a natural state of being in the present moment. The texts are not descriptive or prescriptive, but, rather, suggestive. They hint and allude, inspire and evoke the experience of

“pristine awareness” and “presence.” They provide a rDzogs chen vision of reality as-it-is—a reality that is not colored by human prejudice, habit, and preconception—through heuristic devices such as metaphors, analogies, and paradoxical language.

CONCLUSION 233

To understand the Sems sde texts as philosophical treatises that are based on philosophical arguments is to miss the point of what the texts are all about.

Karmay (1988), in his ground-breaking study on rDzogs chen—his revised dissertation (1986), Origin and Early Development of the Tibetan Religious

Traditions of the Great Perfection (rDzogs Chen)—is mainly concerned with the theoretical exposition of the rDzogs chen doctrine and philosophy. Thus, when he talks about the formative period of rDzogs chen, he is focusing on the formation of

“a new philosophical doctrine,” and when he discusses the rDzogs chen texts, his focus is on the rDzogs chen philosophy. Furthermore, rDzogs chen scholars, such as kLong chen pa (1308-1363), according to him, are rDzogs chen philosophers who have codified the rDzogs chen philosophy and brought together the “various philosophical views” within the rDzogs chen tradition (1988, 211). Karmay’s study has its value and its merits, however, studies that focus exclusively on the doctrine and philosophy tend to misrepresent the rDzogs chen texts.

The Sems sde texts are considered by the traditional accounts to be based on experiential understanding, and their primary concern is with “pristine awareness” and “presence” as a lived experience. In a sense, therefore, the texts provide experiential instructions. They point to a reality that is beyond the conceptual and the discursive and allow us to become aware of the fact that we have access to this reality. To have an experiential understanding of this condition, a practical knowledge, is to discover it directly, that is, to find oneself in it as a lived

CONCLUSION 234 experience. Thus, when one discovers the state of “presence” and is in that

“presence” only then it can be said that one has a “real” understanding of this experience; only then one has a personal, direct experience that is not based on words, ideas, or points of view that constantly change and contradict one another.

The Sems sde texts are pragmatic in their approach and serve practical purposes; they evoke the experience of “pristine awareness” and “presence” and, thus, disclose to the readers unfamiliar possibilities that are not merely conceivable, but also attainable. The focus of the texts is on the readers, that is, on how to

“bring” the readers as close as possible to the actual experience. The texts, therefore, do not seek to change minds, through philosophical arguments, but aim to “draw” the readers toward the experience, and, thus, “lead” them to a personal, direct experience (on indirect communication see Wang 2000, 2003; and D’Amato

2009). That being the case, if we focus exclusively on the doctrine and the philosophy of the rDzogs chen Sems sde texts, we can easily miss the point that the texts are not “about” abstract, philosophical speculations that are based on philosophical arguments, but a lived experience: they aim to evoke the lived experience of “pristine awareness” and “presence.”

It is possible that future work will refine, modify, or even refute some of the conclusions of this dissertation; however, I hope that I have outlined some basic aspects of the emergence and the development of the Sems sde tradition. The rDzogs chen Sems sde tradition is a Tibetan tradition, that emerged in Tibet, and

CONCLUSION 235 was developed by the Tibetans—there is no question about that. In order to avoid gross misunderstandings and distortions of the formative period, we must distinguish between the legendary and the historical accounts. This is not to say, though, that the legendary accounts have no value. The legendary role as a model, of a patriarch, such as dGa’ rab rdo rje, may have been far more important in the development of the Sems sde tradition than his historical role.

With the discovery of several new bKa’ ’gyur collections, in Western Tibet, the future of Tibetan studies in general, and that of rDzogs chen Sems sde in particular has never looked better. Basgo and Hemis are closely affiliated editions that preserve older readings. It is possible that perhaps one day, out of the extant fragments, scholars will be able to reconstruct a version of the Kun byed rgyal po, and thus the five “earlier” Sems sde texts, that predates both the rNying ma rgyud bum and the Tshal pa bKa’ ’gyur extant editions. This would allow us to get a better picture of the Sems sde tradition in the tenth century, when the Kun byed rgyal po was compiled, and better understand the process of editing and transmitting texts in the Tibetan Buddhist traditions. BIBLIOGRAPHY 236

BIBLIOGRAPHY Tibetan Sources

Primary Tibetan Sources Rig pa’i khu byug = rDo rje tshig drug — TkK vol. 1, fol. 209a1-209a4; Bg vol. 5, p. 306; Dg vol. 5, fol. 39b3-39b5; Tb vol. 1, fol. 57a2-57a5; Tk vol. 1, ff. 48b7-49a3; Ba vol. 20, ff. 216b1-216b3; C vol. 19, fol. 43a5-43a7; D vol. 97, fol. 38b3-38b5; H vol. 97, fol. 54b1-54b4; J vol. 102, fol. 39a2-39a4; K vol. 20, fol. 53a8-53b3; N vol. 98, ff. 53a6-53b2; Q vol. 20, fol. 36b5-36b8. rTsal chen sprugs pa — TkK vol. 1, ff. 211a3-211b1; Bg vol. 5, pp. 306-308; Dg vol. 5, fol. 34b1-34b7; Tb vol. 1, ff. 49b7-50b1; Tk vol. 1, ff. 42b1-43a2; Ba vol. 20, ff. 210b7-211a7; C vol. 19, ff. 37b7-38a6; D vol. 97, fol. 33b1- 33b7; H vol. 97, ff. 47a6-47b7; J vol. 102, fol. 33b4-34a3; K vol. 20, ff. 46b2-47a3; N vol. 98, ff. 46a6-46b7; Q vol. 20, ff. 32a4-32b2.

Khyung chen lding ba — TkK vol. 1, ff. 209a4-211a3; Bg vol. 5, pp. 308-314; Dg vol. 5, ff. 30a6-32a1; Tb vol. 1, ff. 44a2-46a7; Tk vol. 1, ff. 37a6-39a7; Ba vol. 20, ff. 206a4-206b8 (fragment); C vol. 19, ff. 33b1-35a6; D vol. 97, ff. 29a6-31a1; H vol. 97, ff. 41b3-43b7; J vol. 102, fol. 29a7-31a3; K vol. 20, ff. 41a3-43a7; N vol. 98, ff. 40b4-42b7; Q vol. 20, ff. 28a7-29b8. rDo la gser zhun — Dg vol. 5, ff. 33b3-34b1; Tb vol. 1, ff. 48b4-49b7; Tk vol. 1, ff. 41b1-42b1; Ba vol. 20, ff. 209b7-210b7; C vol. 19, ff. 37a1-37b7; D vol. 97, ff. 32b3-33b1; H vol. 97, fol. 46a3-47a6; J vol. 102, fol. 32b6-33b4; K vol. 20, ff. 45a8-46b2; N vol. 98, ff. 45a3-46a6; Q vol. 20, ff. 31a7-32a4.

BIBLIOGRAPHY 237

Mi nub pa’i rgyal mtshan = Nam mkha’ che — TkK vol. 1, ff. 211b1-214b2; Bg vol. 5, pp. 384-395; Dg vol. 5, ff. 37a2-39b3; Tb vol. 1, ff. 53b1-57a2; Tk vol. 1, ff. 45b4-48b7; Ba vol. 20, ff. 214a1-216b1 (fragment); C vol. 19, ff. 40b2-43a5; D vol. 97, ff. 36a2-38b3; H vol. 97, ff. 50b7-54b1; J vol. 102, ff. 36a7-39a2; K vol. 20, ff. 50a2-53a8; N vol. 98, ff. 49b6-53a6; Q vol. 20, ff. 34b1-36b5.

Editions of the Bai ro’i rgyud ’bum, rNying ma’i rgyud ’bum, and bKa’ ’gyur Independent Ancient Texts (gTing skyes) [TkK]: rNying ma’i rgyud ’bum. Reproduced from manuscripts preserved at gTing skyes dGon pa byang monastery in Tibet under the direction of Dingo Khyentse Rinpoche, Thimpu, 1973. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, rNying ma rgyud ’bum (gTing skyes), W21518.]

Bai ro’i rgyud ’bum [Bg]: The rGyud ’bum of Vairocana (Bai ro’i rgyud ’bum). In 8 vols. (Smanrtsis Shesrig Spendzod 18-23). Leh: S. W. Tashigangpa, 1971. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, Bai ro’i rgyud ’bum, W21519.] sDe dge [Dg]: rNying ma rgyud ’bum (sDe dge par ma). sDe dge: sDe dge par khang chen mo, 2000-?. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, rNying ma rgyud ’bum (sDe dge par ma), W21939.]

BIBLIOGRAPHY 238 mTshams brag [Tb]: The mTshams brag manuscript of the rNying ma rgyud ’bum. Thimpu: National Library, Royal Government of Bhutan, 1982. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, rNying ma rgyud ’bum (mTshams brag dGon pa’i bris ma), W21521.] gTing skyes [Tk]: rNying ma’i rgyud ’bum. Reproduced from manuscripts preserved at gTing skyes dGon pa byang monastery in Tibet under the direction of Dingo Khyentse Rinpoche, Thimpu, 1973. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, rNying ma rgyud ’bum (gTing skyes), W21518.]

Basgo (Ladakh) [Ba]: Basgo bKa’ ’gyur, 17th century. Basgo scans from the University of Vienna, Austria.

Co ne [C]: Co ne bKa’ ’gyur. Co ne rdzong: Co ne dgon, 1721-31. Reprinted in 1926. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, bKa’ ’gyur (Co ne), W1PD96685.] sDe dge [D]: The sDe dge mtshal par bKa’ ’gyur: A facsimile edition of the 18th century redaction of Si tu Chos kyi ’byung gnas prepared under the direction of H.H. the 16th rGyal dban Karma-pa, Delhi, 1976-1979. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, bKa’ ’gyur (sDe dge par phud), W22084.]

BIBLIOGRAPHY 239

lHa sa [H]: lHa sa bKa’ ’gyur. lHa sa: Zhol bKa’ ’gyur par khang, 1934. Reprinted in 1994. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, bKa’ ’gyur (lHa sa), W26071.]

’Jang sa tham (Lithang) [J]: ’Jang sa tham/Lithang bKa’ ’gyur, 1621. ’Jang sa tham/Lithang scans from the University of Vienna, Austria.

Kangxi (Peking) [K]: Kangxi bKa’ ’gyur, 1669. Taipai Manuscript Kanjur. Kangxi ‘Dragon Sutra’ Kanjur, National Palace Museum, Long-Kuang Digital Culture Publishers, Taiwan, 2011. sNar thang [N]: sNar thang bKa’ ’gyur. sNar thang: sNar thang dgon pa, 18th cent. [An electronic version is now available from the Buddhist Digital Resource Centre (www.tbrc.org), under the title, bKa’ ’gyur (sNar thang), W22703.]

Qianlong (Peking) [Q]: Qianlong bKa’ ’gyur, 1737. The Tibetan Tripiṭaka, Peking edition. Reprint, ed. by D.T. Suzuki. Tōkyō: Tibetan Tripiṭaka Research Institute, 1957.

Secondary Tibetan Sources Ka thog Rig ’dzin Tshe dbang nor bu. 1977. rGya nag hwa shang gi byung tshul grub mtha’i phyogs snga bcas sa bon tsam smos pa in The Collected Works (gSung ’bum) of Ka thog Rig ’dzin chen po Tshe dbang nor bu (6 vols.). Dalhousie, H.P, India.

BIBLIOGRAPHY 240

Nyang Ral nyi ma’i ’od zer. 1988. Chos ’byung me tog snying po sbrang rtsi’i bcud, Gangs can rig mdzod, series no.5, Lhasa, Bod ljongs mi dmangs dpe skrun khang. gNubs chen Sangs rgyas ye shes. 1974. rNal ’byor mig gi bsam gtan or bSam gtan mig sgron, Smanrtsis shesrig spendzod Series. Vol. 74. Leh. kLong chen rab ’byams pa dri med ’od zer. 1983. Chos dbyings rin po che’i mdzod kyi ’grel pa lung gi gter mdzod, in mDzod bdun, Sherab Gyaltsen and Khyentse Labrang, Gangtok, Sikkim. kLong chen rab ’byams pa dri med ’od zer. 1983. Grub mtha’ mdzod, in mDzod bdun, Sherab Gyaltsen and Khyentse Labrang, Gangtok, Sikkim.

Non-Tibetan References

Achard, Jean-Luc. 2013. “The Tibetan Tradition of the Great Perfection.” The Vajrayana Research Resource Website: Nyingma Studies. Accessed November 01, 2013. http://vajrayana.faithweb.com/rich_text_6.html.

Adamek, Wendi L. 2007. The Mystique of Transmission: On an Early Chan History and Its Contexts. New York: Columbia University Press.

BIBLIOGRAPHY 241

Anspal, Sten. 2005. The Space Section of the Great Perfection (rDzogs chen klong sde): A Category of Philosophical and Meditative Teachings in Tibetan Buddhism. MA Thesis, University of Oslo, Norway.

Apple, James B. 2014. “Fragments and Phylogeny of the Tibetan Version of the Mañjuśrīvihārasūtra: A Case Study in the Genealogy of Tibetan Kanjurs.” Annual Report of The International Research Institute for Advanced Buddhology (ARIRIAB) at Soka University for the Academic Year 2013 no. 17:293-335.

Apple, James B. 2016. “The Lotus Sutra in Tibetan Buddhist History and Culture, Part 1.” Bulletin of the Institute of Oriental Philosophy no. 32:130-143.

Apple, James B. 2017. “The Old Tibetan Version of the Kāśyapaparivarta Preserved in Fragments from Dunhuang (1).” Annual Report of The International Research Institute for Advanced Buddhology (ARIRIAB) at Soka University for the Academic Year 2016 no. 20:205-230.

Apple, James B. and Shinobu A. Apple. 2017. “A Re-evaluation of Pelliot tibétain 1257: An Early Tibetan-Chinese Glossary from Dunhuang.” Revue d’Études Tibétaines no. 42:68-180.

Baroetto, Giuseppe. 2010. La dottrina dell’ atiyoga nel bSam gtan mig sgron: Traduzione del settimo capitolo. Vol. 1. LaVergne: Lulu.

BIBLIOGRAPHY 242

Battezzato, Luigi. 2009. “Textual Criticism.” In The Oxford Handbook of Hellenic Studies, edited by George Boys-Stones, Barbara Graziosi, and Phiroze Vasunia, 773-787. Oxford: Oxford University Press.

Beckwith, Christopher I. 1987. The Tibetan Empire in Central Asia: A History of the Struggle for Great Power among Tibetans, Turks, Arabs, and Chinese during the Early Middle Ages. Princeton, New Jersey: Princeton University Press.

Beyer, Stephan V. 1992. The Classical Tibetan Language. Albany: State University of New York.

Broughton, Jeffrey. 1983. “Early Ch’an Schools in Tibet.” In Studies in Ch’an and Hua-yen, edited by Robert M. Gimelo and Peter N. Gregory, 1-68. Honolulu: University of Hawaii Press.

Broughton, Jeffrey L. 1999. The Bodhidharma Anthology: The Earliest Records of Zen. Berkeley: University of California Press.

Buswell, Robert E. 1989. The Formation of Ch’an Ideology in China and Korea: The Vajrasamādhi-Sūtra, a Buddhist Apocryphon. Princeton: Princeton University Press.

Cantwell, Cathy and Rob Mayer. 2006. “Two Proposals for Critically Editing the Texts of the rNying ma’i rGyud ’bum.” Revue d’Études Tibétaines no. 10:56-70.

BIBLIOGRAPHY 243

Capriles, Elías. 2014. Buddhism and Dzogchen: The Doctrine of the Buddha and the Supreme Vehicle of Tibetan Buddhism. Enlarged, revised and corrected version of chapter I of Budismo y dzogchen. La doctrina del Buda y el vehiculo supremo del budismo tibetano. Vitoria: Ediciones La Llave, published in 1999.

D’Amato, Mario. 2009. “Why the Buddha Never Uttered a Word.” In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, edited by Mario D’Amato, Jay L. Garfield and Tom J.F. Tillemans, 41-55. Oxford; New York: Oxford University Press.

Dargyay, Eva K. 1985. “A Nyingmapa Text: The Kun byed rgyal po’i mdo.” In Soundings in Tibetan Civilization, edited by Barbara N. Aziz and Matthew Kapstein. Delhi: Manohar.

Dargyay, Eva M. 1977. The Rise of Esoteric Buddhism in Tibet. Delhi: Motilal Banarsidass.

Davidson, Ronald M. 1981. “The Litany of Names of Mañjuśrī: Text and Translation of the Mañjuśrīnāmasaṃgītī.” In Tantric and Taoist Studies in Honour of R. A. Stein, edited by Michael Strickmann, 1-69. Brussels: Institut Belge des Hautes Études Chinoises.

Davidson, Ronald M. 2002. “Gsar ma Apocrypha: The Creation of Orthodoxy, Gray Texts, and the New Revelation.” In The Many Canons of Tibetan Buddhism: Proceedings of the Ninth Seminar of the International

BIBLIOGRAPHY 244

Association for Tibetan Studies, Leiden 2000, edited by Helmut Eimer and David Germano, 203-224. Leiden: Brill.

De Rossi Filibeck, Elena. 2007. “The Fragmentary Tholing bKa’ ’gyur in the IsIAO Library.” In Pramāṇakīrtiḥ. Papers Dedicated to Ernst Steinkellner on the Occasion of his 70th Birthday. Part I, edited by B. Kellner, H. Krasser, H. Lasix, M.T. Much and H. Tauscher, 53-62. Wien: Arbeitskreis für tibetische und buddhistische Studien.

DeAngelis, Gary D. 2008. “Mysticism in the Daode Jing.” In Teaching the Daode Jing, edited by Gary D. DeAngelis and Warren G. Frisina, 61-73. Oxford; New York: Oxford University Press.

Demiéville, Paul. 1952. Le concile de Lhasa. Une controverse sur le quiétisme entre bouddhistes de l’Inde et de la Chine au VIIIe siècle de l’ère chrétienne, Bibliothèque de l’Institut des Hautes Études Chinoises. Paris: Imprimerie Nationale de France.

Derbac, Mihai. 2007. rNying ma’i rgyud ’bum: A Tibetan Buddhist Canon. MA Thesis, University of Alberta, Canada.

Doctor, Andreas. 2005. Tibetan Treasure Literature: Revelation, Tradition, and Accomplishment in Visionary Buddhism. Ithaca: NY: Snow Lion.

Dotson, Brandon. 2007. “‘Emperor’ Mu-rug-btsan and the ’Phang thang ma Catalogue.” Journal of the International Association for Tibetan Studies no. 3:1-25.

BIBLIOGRAPHY 245

Dowman, Keith (trans. & com). 2006. Eye of the Storm: Vairotsana’s Five Original Transmissions. Kathmandu, Nepal: Vajra Publications.

Downing, Francis G. 2007. “Redaction Criticism.” In Dictionary of Biblical Criticism and Interpretation, edited by Stanley E. Porter. London and New York: Routledge.

Epp, Eldon J. 1993. “The Claremont Profile Method for Grouping New Testament Minuscule Manuscripts.” In Studies in the Theory and Method of New Testament Textual Criticism, edited by Eldon J. Epp and Gordon D. Fee, 211-220. Grand Rapids, Michigan: William B. Eerdmans Publishing Company.

Epp, Eldon J. 1993. “The Eclectic Method in New Testament Textual Criticism: Solution or Symptom?” In Studies in the Theory and Method of New Testament Textual Criticism, edited by Eldon J. Epp and Gordon D. Fee, 141-173. Grand Rapids, Michigan: William B. Eerdmans Publishing Company.

Esler, Dylan 2012. “The Exposition of Atiyoga in gNubs-chen Sangs-rgyas ye- shes’ bSam-gtan mig-sgron.” Revue d’Études Tibétaines no. 24:81-136.

Felsenstein, Joseph. 1989. “PHYLIP-Phylogeny Inference Package (Version 3.2).” Cladistics no. 5:164-166.

BIBLIOGRAPHY 246

Garson, Nathaniel DeWitt. 2004. Penetrating the Secret Essence Tantra: Context and Philosophy in the Mahaāyoga System of rNying-ma Tantra. Ph.D. diss., University of Virginia.

Germano, David. 1995. “Architecture and Absence in the Secret Tantric History of the Great Perfection (rdzogs-chen).” Journal of the International Association for Buddhist Studies no. 17 (2):203-335.

Germano, David. 2002. “The Seven Descents and the Early History of rNying ma Transmissions.” In The Many Canons of Tibetan Buddhism: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, Leiden 2000, edited by Helmut Eimer and David Germano, 225-263. Leiden: Brill.

Germano, David. 2005. “The Funerary Transformation of the Great Perfection (Rdzogs chen).” Journal of the International Association of Tibetan Studies no. 1:1-54. www.thdl.org?id=T1219.

Germano, David. 2007. “Food, Clothes, Dreams, and Karmic Propensities.” In Religions of Tibet in Practice, edited by Donal S. Lopez, 293-312. Princeton , NJ: Princeton University Press.

Gomez, Luis O. 1983. “The Direct and the Gradual Approaches of Zen Master Mahāyāna: Fragments of the Teachings of Mo-ho-yen.” In Studies in Ch’an and Hua-yen, edited by Robert M. Gimelo and Peter N. Gregory, 69-167. Honolulu: University of Hawaii Press.

BIBLIOGRAPHY 247

Gomez, Luis O. 1983. “Indian Materials on the Doctrine of Sudden Enlightenment.” In Early Ch’an in China and Tibet, edited by Whalen Lai and Lewis R. Lancaster, 393-434. Berkeley, CA: Asian Humanities Press.

Guenther, Herbert V. 1983. “‘Meditation’ Trends in Early Tibet.” In Early Ch’an in China and Tibet, edited by Whalen Lai and Lewis R. Lancaster, 351-366. Berkeley, CA: Asian Humanities Press.

Gyatso, Janet. 1996. “Drawn from the Tibetan Treasury: The Gter-ma Literature.” In Tibetan Literature: Studies in Genre, edited by Jose I. Cabezon and Roger R. Jackson, 147-69. Ithaca, NY: Snow Lion.

Hahn, Michael. 2007. “Striving for Perfection: On the Various Ways of Translating Sanskrit into Tibetan.” Pacific World: Journal of the Institute of Buddhist Studies no. 6:123-149.

Hahn, Michael. 2014. “A Brief Introduction into the Indian Metrical System for Students.” https://uni-marburg.academia.edu/MichaelHahn. Accessed November 25, 2015.

Halkias, Georgios T. 2004. “Tibetan Buddhism Registered: A Catalogue from the Imperial Court of ’Phang Thang.” The Eastern Buddhist no. 36 (1-2):46- 105.

Hanson-Barber, A. W. 1986. “The Identification of dGa’-rab rdo-rje.” Journal of the International Association of Buddhist Studies no. 9 (2):55-63.

BIBLIOGRAPHY 248

Harrison, Paul. 1992. Druma-kinnara-rāja-paripṛcchā-sūtra: A Critical Edition of the Tibetan Text (Recension A) based on Eight Editions of the Kanjur and the Dunhuang Manuscript Fragment. Tokyo: The International Institute for Buddhist Studies.

Herrmann-Pfandt, Adelheid. 2002. “The Lhan kar ma as a Source for the History of Tantric Buddhism.” In The Many Canons of Tibetan Buddhism: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, Leiden 2000, edited by Helmut Eimer and David Germano, 129-149. Leiden: Brill.

Herrmann-Pfandt, Adelheid. 2008. Die Lhan kar ma: Ein früher Katalog der ins Tibetische übersetzten buddhistischen Texte. Wien: Verlag der Österreichischen Akademie der Wissenschaften.

Holder, Mark and Paul O. Lewis. 2003. “Phylogeny Estimation: Traditional and Bayesian Approaches.” Nature Reviews Genetics no. 4:275-284.

Howe, Christopher J., Ruth Connolly, and Heather F. Windram. 2012. “Responding to Criticisms of Phylogenetic Methods in Stemmatology.” SEL no. 52 (1):51-67.

Huson, Daniel H. 1998. “SplitsTree: A Program for Analyzing and Visualizing Evelutionary Data.” Bioinformatics no. 14:68-73.

Imaeda, Yoshiro. 1975. “Documents tibétains de Touen-houang concernant le concile du Tibet.” Journal Asiatique no. 263:125-146.

BIBLIOGRAPHY 249

Jinpa, Thupten and Jaś Elsner (trans.). 2000. Songs of Spiritual Experience: Tibetan Buddhist Poems of Insight and Awakening. Boston: Shambhala.

Kapstein, Matthew T. 2008. “The Sun of the Heart and the Bai ro rgyud 'bum.” Revue d’Études Tibétaines no. 14/15:275-288.

Karmay, Samten G. 1980. “An Open Letter by Pho-brang Zhi-ba-’od to the Buddhists in Tibet.” The Tibetan Journal no. 5 (3):3-28.

Karmay, Samten G. 1985. “The rDzogs-chen in Its Earliest Text: A Manuscript from Tun Huang.” In Soundings in Tibetan Civilization, edited by Barbara N. Aziz and Matthew Kapstein, 272-282. Delhi: Manohar.

Karmay, Samten G. 1986. Origin and Early Development of the Tibetan Religious Traditions of the Great Perfection (Rdzogs Chen). Ph.D. diss., School of Oriental and African Studies (SOAS), University of London.

Karmay, Samten G. 1988. The Great Perfection (rDzogs Chen): A Philosophical and Meditative Teaching in Tibetan Buddhism. Leiden: E.J. Brill.

Klein, Anne Carolyn. 1994. Path to the Middle: Oral Mādhyamika Philosophy in Tibet. The Spoken Scholarship of Kensur Yeshey Tupden. Albany: State University of New York Press.

BIBLIOGRAPHY 250

Kloppenborg, John S. 2007. “Source Criticism.” In Dictionary of Biblical Criticism and Interpretation, edited by Stanley E. Porter. London and New York: Routledge.

Kollmar-Paulenz, Karénina. 2007. “The Buddhist Way into Tibet.” In The Spread of Buddhism, edited by Ann Heirman and Stephan Peter Bumbacher, 303- 340. Leiden: Brill.

Lainé, Bruno. 2009. “Canonical Literature in Western Tibet and the Structural Analysis of Canonical Collections.” Journal of the International Association of Tibetan Studies no. 5:1-27. http://www.thlib.org?tid=T5694.

Lalou, Marcelle. 1953. “Les Textes Bouddhiques au Temps du Roi Khri-srong-lde- bcan.” Journal Asiatique no. 241:313-353.

Lantin, Anne-Catherine, Philippe V. Baret, and Caroline Macé. 2004. “Phylogenetic Analysis of Gregory of Nazianzus’ Homily 27.” In Le poids des mots: Actes des 7es Journées Internationales d’Analyse Statistique des Données Textuelles, edited by Gérald Purnelle, Cédrick Fairon, and Anne Dister. Louvain-la-Neuve: Presses Universitaires de Louvain.

Liljenberg, Karen. 2009. “On the History and Identification of Two of the Thirteen Later Translations of the Dzogchen Mind Series.” Revue d’Études Tibétaines no. 17:51-62.

Liljenberg, Karen. 2012. A Critical Study of the Thirteen Later Translations of the Dzogchen Mind Series. Ph.D. diss., University of London.

BIBLIOGRAPHY 251

Liljenberg, Karen. 2012. “On the Thig le drug pa and the sPyi chings, Two of the Thirteen Later Translations of the rDzogs chen Mind Series.” Revue d’Études Tibétaines no. 24:137-156.

Liljenberg, Karen. 2018. “From Treatise to Tantra: the Pure Golden Ore (rDo la gser zhun) and the Tantra of Meditation on Bodhicitta (Byang chub sems bsgom pa’i rgyud).” Revue d’Études Tibétaines no. 44:31-108.

Maas, Paul. 1958. Textual Criticism. Translated by Barbara Flowers. Oxford: Clarendon Press.

Maas, Philipp A. 2008. “A Phylogenetic Approach to the Transmission of the Tibetan Kanjur — The Akṣayamatinirdeśa Revisited.” In Bauddhasāhityastabakāvalī. Essays and Studies on Buddhist Sanskrit Literature: Dedicated to Claus Vogel by Colleagues, Students, and Friends, edited by Dragomir Dimitrov, Michael Hahn and Roland Steiner, 229-243. Marburg: Indica et Tibetica Verlag.

Maas, Philipp A. 2010. “Computer Aided Stemmatics — The Case of Fifty-Two Text Versions of Carakasaṃhitā Vimānasthāna 8.67-157.” Wiener Zeitschrift für die Kunde Südasiens / Vienna Journal of South Asian Studies no. 52/53:63-119.

Mahoney, Richard, ed. 2004. Sanskrit, Tibetan and Chinese Buddhist Terminology: Mahāvyutpatti and Yogācārabhūmi. Oxford, North Canterbury: Indica et Buddhica.

BIBLIOGRAPHY 252

Marwick, Ben. 2012. “A Cladistic Evaluation of Ancient Thai Bronze Buddha Images: Six Tests for a Phylogenetic Signal in the Griswold Collection.” In Connecting Empires and States, edited by Dominik Bonatz, Andreas Reinecke and Mai Lin Tjoa-Bonatz, 159-176. Singapore: National University of Singapore Press.

Mayer, Robert. 2006. “Textual Criticism of the rNying ma’i rgyud ’bum Tradition.” In Tibetan Buddhist Literature and Praxis: Studies in its Formative Period 900-1400 (PIATS 2003), edited by Ronald Davidson and Christian Wedemeyer, 95-121. Leiden: Brill.

McRae, John R. 1986. The Northern School and the Formation of Early Ch’an Buddhism. Honolulu: University of Hawaii Press.

Meinert, Carmen. 2002. “Chinese Ch’an and Tibetan rDzogs chen: Preliminary Remarks on Two Tibetan Dunhuang Manuscripts.” In Religion and Secular Culture in Tibet: Tibetan Studies II, PIATS 2000, edited by Henk Blezer, 289-307. Leiden: Brill.

Meinert, Carmen. 2006. “The Legend of Cig car ba Criticism in Tibet: A List of Six Cig car ba Titles in the Chos ’byung me tog snying po of Nyang Nyi ma’od zer (12th Century).” In Tibetan Buddhist Literature and Praxis. Studies in Its Formative Period 900-1400 (PIATS 2003), edited by Ronald Davidson and Christian Wedemeyer, 31-54. Leiden: Brill.

BIBLIOGRAPHY 253

Metzger, Bruce M. and Bart. D. Ehrman. 2005. The Text of the New Testament: Its Transmission, Corruption, and Restoration. 4th ed. New York and Oxford: Oxford University Press.

Mink, Gerd. 2002. “The Coherence-Based Genealogical Method - What is it about?” http://egora.uni-muenster.de/intf/projekte/gsm_aus_en.shtml. Accessed November 25, 2015.

Mink, Gerd. 2004. “Problems of a Highly Contaminated Tradition: The New Testament.” In Studies in Stemmatology 2, edited by Pieter Van Reenen, August den Hollander, and Margot van Mulken, 13-85. Philadelphia, PA: John Benjamins Publishing Company.

Nattier, Jan. 2008. “Rhymed Verse, Unrhymed Verse, or No Verse at All? The Treatment of Indian Poetic Passages in Chinese Buddhist Translations.” Paper read at the American Oriental Society (Western Branch), at Portland State University, Portland.

Neumaier, Eva K. 1998. “The bSam gtan mig sgron and its rDzogs chen Quotations: A Study in the Production of Tibetan Texts.” Paper read at Proceedings of the 8th Seminar of the International Association for Tibetan Studies, at Bloomington, Indiana.

Neumaier-Dargyay, Eva K. 1992. The Sovereign All-Creating Mind, the Motherly Buddha. A translation of the Kun byed rgyal po’i mdo. Albany, NY: State University of New York Press.

BIBLIOGRAPHY 254

Norbu, Namkhai. 1986. The Cristal and the Way of Light. Compiled and edited by John Shane. London: RKP.

Norbu, Namkhai. 1989. Dzogchen: The Self-Perfected State. Translated by John Shane. Compiled and edited by Adriano Clemente. Ithaca, NY: Snow Lion.

Norbu, Namkhai. 2006. Dzogchen Teachings. Ithaca, New York: Snow Lion Publications.

Norbu, Namkhai and Adriano Clemente. 1999. The Supreme Source: The Fundamental Tantra of the Dzogchen Semde Kunjed Gyalpo. Translated from Italian into English by Andrew Lukianowicz. Ithaca, New York: Snow Lion Publications.

Norbu, Namkhai and Kenard Lipman (trans.). 1987. Mañjuśrimitra. Primordial Experience: An Introduction to rDzogs-chen Meditation. Boston & London: Shambhala.

Odin, Jaishree. 1986. “Suggestiveness: Poe’s Writings from the Perspective of Indian ‘Rasa’ Theory.” Comparative Literature Studies no. 23 (4):297-309.

Ollett, Andrew. 2013. “The Gaṇacchandas in the Indian Metrical Tradition.” In Puṣpikā: Tracing Ancient India, through Texts and Traditions. Contributions to Current Research in Indology. Vol. I, edited by Nina Mirnig, Péter-Dániel Szántó and Michel Williams, 331-364. Oxford & Philadelphia: Oxbow Books.

BIBLIOGRAPHY 255

Orofino, Giacomella. 1994. Sekoddeśa: A Critical Edition of the Tibetan Translations. With an Appendix by Raniero Gnoli on the Sanskrit Text. Roma: Instituto Italiano per il Medio ed Estremo Oriente Roma.

Pagel, Ulrich. 2007. “The Dhāraṇīs of Mahāvyutpatti #748: Origin and Formation.” Buddhist Studies Review no. 24 (2):151-191.

Palmo, Ani Jinba (trans.). 2004. The Great Image: The Life Story of Vairochana the Translator. Boston & London: Shambhala.

Pang, Rachel Hua-Wei. 2011. Dissipating Boundaries: The Life, Song-Poems, and Non-Sectarian Paradigm of Shabkar Tsokdruk Rangdrol (1781-1851). Ph.D. diss., University of Virginia.

Parker, David C. 2008. An Introduction to the New Testament Manuscripts and Their Texts. Cambridge, New York: Cambridge University Press.

Reynolds, John M. 1996. The Golden Letters: The Three Statements of Garab Dorje. Ithaca, NY: Snow Lion Publications.

Ronquist, Fredrik and John P. Huelsenbeck. 2003. “MrBayes 3: Bayesian Phylogenetic Inference under Mixed Models.” Bioinformatics no. 19:1572- 1574.

Roos, Teemu and Tuomas Heikkilä. 2009. “Evaluating Methods for Computer- Assisted Stemmatology Using Artificial Benchmark Data Sets.” Literary and Linguistic Computing no. 24 (4):417-433.

BIBLIOGRAPHY 256

Ruegg, David Seyfort. 1989. Buddha-Nature, Mind and the Problem of Gradualism in Comparative Perspective: On the Transmission and Reception of Buddhism in India and Tibet. Jordan Lectures in Comparative Religion no. 13. London: School of Oriental and African Studies, University of London.

Scherrer-Schaub, Cristina. 1999. “Translation, Transmission, Tradition: Suggestions from Ninth-Century Tibet.” Journal of Indian Philosophy no. 27:67-77.

Scherrer-Schaub, Cristina. 2002. “Enacting Words: A Diplomatic Analysis of the Imperial Decrees (bkas bcad) and Their Application in the sGra sbyor bam po gñyis pa Tradition.” Journal of the International Association of Buddhist Studies no. 25 (1-2):263-340.

Silk, Jonathan A. 1994. The Heart Sūtra in Tibetan: A Critical Edition of the Two Recensions Contained in the Kanjur. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universitat Wien.

Silk, Jonathan A. 2009. “The Nature of the Verses of the Kāśyapaparivarta.” Bulletin of the Asia Institute no. 23:181-190.

Sørensen, Per K. 1990. Divinity Secularized: An Inquiry Into the Nature and Form of the Songs Ascribed to the Sixth Dalai Lama. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien.

BIBLIOGRAPHY 257

Soulen, Richard N. and R. Kendall Soulen. 2001. Handbook of Biblical Criticism. 3rd rev. ed. Louisville, Kentucky: Westminster John Knox Press.

Sparks, Kenton L. 2007. “Form Criticism.” In Dictionary of Biblical Criticism and Interpretation, edited by Stanley E. Porter. London and New York: Routledge.

Stein, Rolf A. 2010. Rolf Stein’s Tibetica Antiqua: With Additional Materials. Translated and edited by Arthur P. McKeown. Leiden: Brill.

Studstill, Randall. 2005. The Unity of Mystical Traditions: The Transformation of Consciousness in Tibetan and German Mysticism. Leiden: Brill.

Sujata, Victoria. 2005. Tibetan Songs of Realization: Echoes from a Seventeenth- Century Scholar and Siddha in Amdo. Leiden: Brill.

Suzuki, Daisetz Teitaro. 1989. Studies in the Laṅkāvatāra Sūtra: One of the most important texts of Mahāyāna Buddhism, in which almost all its principal tenets are presented, including the teaching of Zen. First Published in 1930. New Delhi, India: Munshiram Manoharlal.

Swofford, David L. 2002. PAUP. Phylogenetic Analysis Using Parsimony (and Other Methods). Version 4. Sunderland, Massachusetts: Sinauer Associates.

Takahashi, Kammie Morrison. 2009. Lamps for the Mind: Illumination and Innovation in dPal dbyangs Mahāyoga. Ph.D. diss., University of Virginia.

BIBLIOGRAPHY 258

Takasaki, Jikido. 1980. “Analysis of the Laṅkāvatāra: In Search of Its Original Form.” In Indianisme et Bouddhisme: Mélanges offerts à Mgr Étienne Lamotte, edited by Daniel Donnet, 339-352. Louvain-la-Neuve: Institut Orientaliste.

Tanaka, Kenneth K. and Raymond E. Robertson. 1992. “A Ch’an Text from Tun- huang: Implications for Ch’an Influence on Tibetan Buddhism.” In Tibetan Buddhism: Reason and Revelation, edited by Steven D. Goodman and Ronald M. Davidson, 57-78. Albany, NY: State University of New York Press.

Tanselle, G. Thomas. 1972. “Some Principles for Editorial Apparatus.” Studies in Bibliography no. 25:41-88.

Tanselle, G. Thomas. 1983. “Classical, Biblical, and Medieval Textual Criticism and Modern Editing.” Studies in Bibliography no. 36:21-68.

Tov, Emanuel. 1992. Textual Criticism of the Hebrew Bible. 2nd rev. ed. Fortress Press, Minneapolis: Royal Van Gorcum, Assen.

Tucci, Giuseppe. 1958. Minor Buddhist Texts. Part 2. Serie Orientale Roma, Vol. IX. Rome: Istituto Italiano per il Medio ed Estremo Oriente.

Tucci, Giuseppe. 1966. Tibetan Folk Songs from Gyantse and Western Tibet. Second, Revised and Enlarged ed. Ascona, Switzerland: Artibus Asiae.

BIBLIOGRAPHY 259

Tulku Thondup. 1986. Hidden Teachings of Tibet: An Explanation of the Terma Tradition of Tibetan Buddhism. Edited by Harold Talbott. London: Wisdom Publications.

Ueyama, Daishun. 1983. “The Study of Tibetan Ch’an Manuscripts Recovered from Tun-huang: A Review of the Field and its Prospects.” In Early Ch’an in China and Tibet, edited by Whalen Lai and Lewis R. Lancaster, 327-349. Berkeley, CA: Asian Humanities Press.

Vaganay, Lion and Christian-Bernard Amphoux. 1991. An Introduction to New Testament Textual Criticism. Translated by Jenny Heimerdinger. 2nd rev. and updated ed. Cambridge, New York: Cambridge University Press.

Valby, James. 2012. “Five Principles of rDzogs chen Transmission in the Kun byed rgyal po.” Revue d’Études Tibétaines no. 24:157-163.

Valby, James M. 1983. The Life and Ideas of the 8th Century A.D. Indian Buddhist Mystic Vimalamitra: A Computer-Assisted Approach to Tibetan Texts. PhD diss., University of Saskatchewan, Canada. van Schaik, Sam. 2004. “The Early Days of the Great Perfection.” Journal of the International Association for Buddhist Studies no. 27 (1):165-206. van Schaik, Sam. 2012. “Dzogchen, Chan and the Question of Influence.” Revue d’Études Tibétaines no. 24:5-19.

BIBLIOGRAPHY 260 van Schaik, Sam. 2015. Tibetan Zen: Discovering a Lost Tradition. The Stories Told by Dunhuang Cave Manuscripts. Boston: Snow Lion. van Schaik, Sam. 2018. The Spirit of Zen. New Haven and London: Yale University Press.

Wang, Youru. 2000. “The Pragmatics of ‘Never Tell Too Plainly’: Indirect Communication in Chan Buddhism.” Asian Philosophy no. 10 (1):7-31.

Wang, Youru. 2003. Linguistic Strategies in Daoist Zhuangzhi and Chan Buddhism: The Other Way of Speaking. London and New York: RoutledgeCurzon.

Wangchuk, Dorji. 2002. “An Eleventh-Century Defence of the Authenticity of the Guhyagarbha Tantra.” In The Many Canons of Tibetan Buddhism: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, Leiden 2000, edited by Helmut Eimer and David Germano, 265-291. Leiden: Brill.

Wangdu, Pasang and Hildegard Diemberger. 2000. dBa’ bzhed: The Royal Narrative Concerning the Bringing of the Buddha’s Doctrine to Tibet. Wien: Verlag der Österreichischen Akademie der Wissenschaften.

West, Martin L. 1973. Textual Criticism and Editorial Technique: Applicable to Greek and Latin Texts. Stuttgart: B. G. Teubner.

BIBLIOGRAPHY 261

Wilkinson, Christopher. 2012. “The Mi nub rgyal mtshan Nam mkha’ che and the Mahā Ākāśa Kārikās: Origins and Authenticity.” Revue d’Études Tibétaines no. 24:21-80.

Windram, Heather F., Christopher J. Howe, and Matthew Spencer. 2005. “The Identification of Exemplar Change in the Wife of Bath’s Prologue Using the Maximum Chi-Squared Method.” Literary and Linguistic Computing no. 20 (2):189-204.

Windram, Heather F., Prue Shaw, Peter Robinson, and Christopher J. Howe. 2008. “Dante’s Monarchia as a Test Case for the Use of Phylogenetic Methods in Stemmatic Analysis.” Literary and Linguistic Computing no. 23 (4):443- 463.

Wynne, Alexander. 2006. “The Oral Transmission of Early Buddhist Literature.” Journal of the International Association of Buddhist Studies no. 27 (1):97- 124.

Ying, Chinghui Jianying. 2010. Being and Knowing in Wholeness: Chinese Chan, Tibetan Dzogchen, and the Logic of Immediacy in Contemplation. Ph.D. diss., Rice University.

Yorav, Avishai, Tal Dagan, and Dan Graur. 2005. “An Exploratory Study on the Use of a Phylogenetic Algorithm in the Reconstruction of Stemmata of Halachi Texts.” Hebrew Union College Annual no. 76:273-288.

BIBLIOGRAPHY 262

Zimmermann, Michael. 1998. “A Second Tibetan Translation of the Tathāgatagarbhasūtra in the Newark Manuscript Kanjur from Bathang: A Translation of the Early Period (snga dar).” Transactions of the International Conference of Eastern Studies no. 43:33-50.

Zimmermann, Michael. 2002. A Buddha Within, The Tathāgatagarbhasūtra: The Earliest Exposition of the Buddha-Nature Teaching in India. Tokyo: The International Research Institute for Advanced Buddhology.

Zou, Yuan. 2010. Structural EM Methods in Phylogenetics and Stemmatology. M.Sc. Thesis, University of Helsinki, Finland.