1. Yoga, History-Text-and-Context-of-the-Yoga-Upanisads

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directiy from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps.

ProQuest Information and Learning

300 North Zeeb Road, Ann Arbor, MI 48106-1346 USA

800-521-0600

UMI®

UNIVERSITY OF CALIFORNIA

Santa Barbara

History, Text, and Context of the Yoga Upaniṣads

A Dissertation submitted in partial satisfaction of the

requirements for the degree of Doctor of Philosophy

in Religious Studies

by

Jeffrey Clark Ruff

Committee in charge:

Professor David Gordon White

Professor Barbara A. Holdrege

Professor William F. Powell

December 2002

UMI Number: 3073645

UMI

UMI Microform 3073645

Copyright 2003 by ProQuest Information and Learning Company.

All rights reserved. This microform edition is protected against

unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company

300 North Zeeb Road

P.O. Box 1346

Ann Arbor, MI 48106-1346

The dissertation of Jeffrey C. Ruff is approved.

William F. Powell

Barbara A. Holdrege

David G. White, Committee Chair

August 2002

History, Text, and Context of the Yoga Upanisads

Copyright @ 2002

by

Jeffrey Clark Ruff

iii

VITA OF JEFFREY CLARK RUFF

SEPTEMBER 2002

EDUCATION

Bachelor of Arts, Southwest Missouri State University, May 1990 (magna cum laude)

Master of Arts, Southwest Missouri State University, May 1997

Doctor of Philosophy, University of California, Santa Barbara, September 2002 (expected)

PROFESSIONAL EMPLOYMENT

1988, 1989: Editorial Internship, Hendrickson Publishers

1990-1993: Teaching Assistant and Research Assistant, Department of Religious Studies, University of North Carolina, Chapel Hill

1995-1996: Teaching Assistant, Department of Religious Studies, Southwest Missouri State University

1996-1997: Instructor, Department of Religious Studies, Southwest Missouri State University

1997-1999: Teaching Assistant, Department of Religious Studies, University of California, Santa Barbara

2001-2002: Director, University of California Education Abroad Program in India, Delhi

2002-present: Assistant Professor, Department of Religious Studies, Marshall University

PUBLICATIONS

"Tantra and Gender: Symbol, Polarity, and Unity," Excursus: A Review of Religious Studies (North Carolina, 1992)

"The Use of Skulls as Ritual Objects In Prehistoric and Historic Global Context," Excursus: A Review of Religious Studies (North Carolina, 1993)

Abstracts for Science of Religion (London, 1997)

Review Note: In Search of the Sacred: Anthropology and the Study of Religions, by Clinton Bennett (London, 1996). Religious Studies Reviews (1997)

Review Note: The Cultural Dialectics of Knowledge and Desire, by Charles W. Nuckolls (Madison, 1996). Religious Studies Reviews (1999)

Review Essay: Savage Systems: Colonialism and Comparative Religion in Southern

Africa, by David Chidester (Charlottesville and London, 1996). Method and Theory in the Study of Religion (1999)

"Lecture Supplements" for Instructor's Manual to Accompany Experiencing the World's Religions, by Michael Molloy (Mountain View and London, 1999)

"Healing," in Contemporary American Religion, ed. Wade Clark Roof (New York, 2000)

AWARDS

Regents Scholarship, Southwest Missouri State University, 1985-1989

Antiquities Program Scholarship, Southwest Missouri State University, 1987-1988, 1989-1990

Walter O. Cralle Memorial Scholarship for Anthropology, Southwest Missouri State University, 1988-1989

Religious Studies Scholarship, Southwest Missouri State University, 1988-1989

Religious Studies Memorial Award for Academic Excellence, 1989-1990

Rose Marie and Martin H. Boone, Jr. Scholarship for Religious Studies, Southwest Missouri State University, 1989-1990

Merit Assistantship, University of North Carolina, Chapel Hill, 1990-1991

Graduate Assistantship, Southwest Missouri State University, 1995-1996

Rowny Assistantship, University of California, Santa Barbara, 1997-2000

Graduate Division Dissertation Fellowship, University of California, Santa Barbara, 2002

FIELDS OF STUDY

Major Field: South Asian Religions

Studies in Ancient South Asia with Barbara Holdrege

Studies in Yoga, Tantra, and History of Religions with David G. White

Studies in East Asian Traditions with William Powell

v

ABSTRACT

History, Text, and Context of the Yoga Upaniṣads

by

Jeffrey Clark Ruff

The Yoga Upaniṣads represent two different canons of minor Upaniṣads

whose primary subjects are mantra yoga and hatha yoga. The history of these

texts is examined in terms of their roots in classical Vedanta and the

development of the genre to which they belong, minor Upaniṣads, or

Atharvanic Upaniṣads. The texts are examined according to the sources and

subject matter of each text in their North Indian versions, and in the later

South Indian expanded versions. A detailed synopsis, with selected

translation, is provided for each text. The study also includes an

investigation of the notion of the yogic body in the Brahmanical yoga

traditions associated with these texts. Special attention is given to the

unsystematic descriptions of mantra yoga and the yogic body in the Northern

Yoga Upaniṣads, and hatha yoga and the standardized presentation of the yogic

body in the Southern texts. These discussions of the yogic body include

especially the texts' notions of the cakras, or energy centers within the human

vi

body. The context of the canons is explained relative to the larger contexts of

Classical Yoga, Vedanta, Tantra, and the yoga traditions among practitioners

of Tantra and among the Nath Siddhas. Taken together, all the proceeding

evidence is employed to present a reconstruction of the milieu of the texts

and of the communities that produced them.

vii

TABLE

OF

CONTENTS

Chapter

1:

Yoga

and

the

Yoga

Upanisads

.................................................................................

1

Chapter

2:

The

Classical

and

Minor

Upanisads:

Secret

teachings

on

Hidden

Connections

................................

41

Chapter

3:

The

Yoga

Upanisads:

Textual

Synopses

............................................................

107

Chapter

4:

The

Traditions

of

the

Yoga

Upanisads:

Mantra

Yoga

and

Hatha

Yoga

............................................................

204

Chapter

5:

Social

World

of

the

Yoga

Upanisads:

History,

Text,

and

Context

....................................................

281

Selected

Bibliography

.........................................................................

314

viii

Abbreviations

Sanskrit texts

ATU

Advayatāraka Upaniṣad

AU

Aitareya Upaniṣad

AP

Amaraugha Prabodha of Gorakhnath

ABU

Amṛtabindu Upaniṣad (Brahmā-Bindu)

ANU

Amṛtānada Upaniṣad (Amṛtabindu, Amṛtanādabindu)

AV

Atharvaveda

BĀU

Bṛhādāraṇyaka Upaniṣad

BG

Bhagavadgītā

BBU

Brahmabindu Upaniṣad (earlier title of the ABU)

BVU

Brahmanvidya Upaniṣad

CU

Cūlikā Upaniṣad (see also Mantrikā)

ChU

Chāndogya Upaniṣad

DhBU

Dhyānabindu Upaniṣad

DU

Darśana Upaniṣad (Yogadarśana)

GP

Gorakṣapaddhati

GS

Gheraṇḍa Samhitā

GŚ

Gorakṣāṣataka

HU

Haṃsa Upaniṣad ( Haṃsanāda)

HYP

Haṭhayogapradīpikā

IU

Īśā Upaniṣad

KaṭhU

Kāṭha Upaniṣad

KJN

Kaulaññānanirnaya

KSU

Kauṣītaki Upaniṣad

KeU

Kena Upaniṣad

KU

Kṣurika Upaniṣad

KMT

Kubjikāmata Tantra

LYV

Laghu yogavāsiṣṭha

MaitU

Maitrāyaṇīya Upaniṣad (also Maitri, et al.)

MBU

Maṇḍalabrāhmaṇa Upaniṣad

ManU

Mantrikā Upaniṣad (southern recension of the CU)

MBh

Mahābhārata

MVU

Mahāvākya Upaniṣad

MaU

Maṇḍukya Upaniṣad

MU

Muktikā Upaniṣad

MuṇḍU

Muṇḍaka Upaniṣad

NBU

Nādabindu Upaniṣad

ix

PP

Padukāpañcaka

PBU

Pāśupatabrahmā Upaniṣad

PraśU

Praśna Upaniṣad

RG

Rbhuḡitā

RV

Rgveda

SV

Sāmaveda

ŚU

Śāṇḍilya Upaniṣad

ŚTT

Śāradātilaka Tantra

SCN

Saṭcakranirupaṇa

SSS

Saṭṣahasra Saṃhitā

SSP

Siddhasiddhāntapaddhati (attrib. to Gorakhnāth)

ŚS

Śiva Saṃhitā

ŚvetU

Śvetāśvatara Upaniṣad

TaitU

Taittirīya Upaniṣad

TŚBU

Triśikhibrahmaṇa Upaniṣad

TBU

Tejobindu Upaniṣad

UG

Uttaragītā

VU

Varāha Upaniṣad

YV

Yajurveda

YBV

Yogabhāṣya of Vyāsa

YB

Yogabhāṣya of Gorakhnāth (see also YSA)

YCU

Yogacūḍāmaṇi Upaniṣad

YKU

Yogakuṇḍali Upaniṣad (Yogakuṇdalini)

YSD

Yogāśāstra of Dattātreya

YSA

Yogisiddhāmrta (variant of the Yogabīja of Gorakhnāth)

YŚU

Yogaśikha Upaniṣad

YS

Yoga Sūtra of Patañjali

YTU

Yogatattva Upaniṣad

YVVB

Yogavārttika of Vijñāna Bhikṣu

YV

Yogavāsiṣṭha

YYV

Yogayājn̄āvalkya

YYS

Yogayājn̄āvalkya Saṃhitā

Collections

ALS

Adyar Library Series

ĀSS

Ānandāśrama Sanskrit Series

BI

Bibliotheca Indica

SU

Saṃnyāsa Upaniṣad(s)

YU

Yoga Upaniṣad(s)

X

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

Chapter 1:

Yoga and the Yoga Upaniṣads

1.1. Introduction

The Yoga Upaniṣads are a set of medieval Indian texts composed by students

and masters of mantra yoga and haṭha yoga originating within Brahmanical

communities.1 No Indian or Western scholar has systematically analyzed these

medieval texts. Consequently, many scholars have consistently made comments

and claims about the texts that my more detailed research does not support. One

example of how thorough the misunderstandings have been concerns the number

of Yoga Upaniṣads and contents of different recensions of the texts in this category.

Scholars have generally treated these texts as of one body and of similar historical

development. On the contrary, the Yoga Upaniṣads represent at least two textual

traditions. Eleven texts of this “yoga” genre were composed between the ninth

and thirteenth centuries in North India. These texts expound a form of mantra

yoga that employs theory and practice of recitation of the om mantra. These eleven

texts were substantially expanded in the seventeenth to eighteenth centuries in

South India. At the same time ten additional texts were added that brought the

1 Yoga Upaniṣad is an artificial designation for these texts, as discussed below. Most of the

earliest texts (or parts of the texts) date to approximately the twelfth to thirteenth centuries, others

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

number of conventionally recognized Yoga Upaniṣads to twenty-one.2 Thus, the

traditions of medieval Upaniṣads on the subject of yoga were continually evolving

ones.

The southern traditions expanded the older northern recensions, drawing

heavily from the North Indian Siddha traditions, especially the Nāth Siddhas, and

various additional yogic and tantric traditions. The adoption of tantric and Nāth

materials did not alter the conventionally Brahmanical tone of the texts. The

Upanisadic authors rejected any form of extreme libertine practices and symbols

of these groups as well as the Siddhas' search for power-in-this-world. Instead,

these texts are non-dual Vedānta texts, primarily employed by brahmin males

during their life as yogins, whether in monastic settings, as hermit-yogins, or

possibly as renunciant yogins.3

Unlike the Tantras, these texts mark a return to the moods and voices of the

classical Upaniṣads and Yoga Darśana in their concentration on kaivalya, mokṣa, and

date as late as the eighteenth century. The analyses of and difficulties of dating will be discussed

throughout, see especially chapter 2 and 3.

2 The Sanskrit edition of the text from the Adyar Library Series contains twenty texts. I

have analyzed twenty-one texts. The issues involved in classification are explained in chapter 3.

3 The texts do not provide enough information to be conclusive. However, there are

intellectual trends in the southern corpus that suggests an urban and possibly monastic setting.

Further research is necessary to examine specific institutions in Tamil Nadu to substantiate such

hints within the texts. Hermit-yogins are those who are described as the vānaprasthins or forest-

dwellers, in the Brahmanical scheme of four āśramas, or stages of life. The relationship of these

texts to "renunciants", or samnyāsus, is unclear. The texts do not have an overly "renunciant" tone,

māyā.4 In other words, if the Tantras are seen as reorienting praxis away from

liberation (mokṣa) in favor of power, then the Yoga Upaniṣads and related texts are a

return to the alternate formula of “the world is illusion” and the yogin must

exclusively seek liberation. However, the story or history of this return is

temporally after tantra had already gained permanent place within South Asian

spirituality, theology, and religious institutions. Thus, the Upanisadic “revival” of

the Yoga Upaniṣads is also not “ole time” Indian religion but something new. This

new creation is not identical to the classical Upaniṣads and Yoga Darśana nor is it

tantric, but instead it is a late medieval recombination and continuation of all these

traditions. As such, the Yoga Upaniṣads do not represent an isolated movement.

Śaṅkara, and others like him, during the same period were reasserting

Brahmanical interpretations of their classical texts and adding new texts, ideas,

and practices of their own.

This is no new pattern in the evolution of South Asian religious practice

and thought. Even tantra, in the broadest meaning of the word, is a combination

especially compared with the “renunciant” texts of the same medieval genre, the Saṃnyāsa

Upaniṣads.

4 Kaiṅvalya has the specialized meaning for yoga traditions of “liberation” and “isolation”

from material nature, mokṣa is the more generally used term for “liberation” from the phenomenal

world. Māyā is a complex term with several meanings. Primary among these are “the creative

matrix of reality” or the “creative power that sustains material nature.” Through interpretation by

the schools of non-dual Vedānta this term gains the additional connotation of “illusion.” These

traditions interpret the “power of material nature” as conditional, temporary, and in fluctuating

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

of the Vedas, Darśanas, Āgamas and several other intellectual traditions and

cultural movements, such as those of Buddhists, Jains, and also includes many

shamanic elements from extra-Vedic village traditions. Unfortunately it is exactly

this mimetic recombination of older complex modes of practice and thought that

have left the Yoga Upaniṣads under-studied, or more accurately phrased, poorly

explained. To offer an example, the attitude of scholars concerning the derivative

character of these texts leads Mircea Eliade to write:

The majority of these [Yoga Upaniṣads] merely repeat the traditional

clichés, and either adhere to or summarize the schemata of the most

important yogic Upaniṣads—Yogatattva, the Dhyanabindu, and the

Nādabindu. Only these three are worth examining more thoroughly.5

Other scholars assess the texts more seriously, but none have offered a systematic

discussion of the corpus.6

The derivative and Brahmanical characters of these texts exhibit several

streams of textual confluence. These textual and philosophical antecedents are of

two types. First, they include the Brahmanical traditions of Śaṅkara and post-

Śaṅkara interpretations of the classical Upaniṣads and Vedānta Sūtras, Patañjali’s

Yoga Sūtra and its commentary traditions, and medieval systems of theistic yoga.

and therefore, not “True, eternal, or Absolute” as is the case with the “Absolute,” Brahman, or the

essential Self, the ātman.

5 Eliade (1973), p. 129.

6 Jean Varenne and Georg Feuerstein offer a more positive assessment about the

importance of these texts. They do not provide comprehensive analysis or histories for the texts.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

These latter systems are often Vaiṣṇava and are associated with the names of sages

and demigods, such as Yajñavalkya, Dattātreya, and Ṛbhu. Second, significant

influences come from Āgamas, various Tantras, texts attributed to Gorakhnāth and

the late haṭha yoga texts derived from the works of the Nāth Siddhas, especially the

Haṭhayogapradīpikā. This second group is often Śaiva (but seldom Śākta) in

character. These antecedent traditions assert considerable influence on the shape

of the post-sixteenth century South Indian tradition of Yoga Upaniṣads. The earlier

northern texts bear closer resemblance to the classical Upaniṣads. Their intellectual

and philosophical influences are more impressionistic and therefore more difficult

to trace and analyze.

The Yoga Upaniṣads are not overly theological. Although gods are

mentioned they are the composite and generic gods of the Vedas and Purāṇas, not

the specialized gods of yoga and tantra. If present at all, gods usually appear in the

frame verses at the beginning and end of the texts. Goddesses are not prominent

in these texts, although goddesses present in the cakras and kundalini śakti are

important to some of the southern expanded texts. Most of these Upaniṣads focus

their teachings and practice on repetition of the mantra om. Some of the Yoga

Upaniṣads include a similar series of nāda yoga concerns and descriptions,

especially focused on recitation of the haṃsa mantra. In addition to the practice of

the mantra, haṃsa imagery, symbolism, and theory appears throughout virtually

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

all of the expanded southern texts. Northern and southern recensions of the texts

employ many common classical Upanisadic metaphors for the essential Self, the

ātman. Some of these metaphors included the ātman as the empty space in a jar,

ātman as a chariot, ātman as the sun, ātman as the migratory gander (then haṃsa),

ātman as a fire, and other metaphors and symbols as well. The southern texts

expanded their presentation of Vedānta to include many of the ideas and images

of Śaṅkara's non-dualism.

Although internal evidence (quotations from datable texts) and manuscript

histories provide relative dates for these texts, specific dating and authorship are

uncertain. Heretofore, few scholars have suggested specific geographic origins of

the texts. Herein, I synthesize evidence from both primary and secondary sources

to conclusively demonstrate that the most of the Yoga Upaniṣads are dateable to

within a hundred year period (in two phases, northern and southern), the regional

geographic location is estimable, and that the textual and intellectual sources of

the texts are partially known.

In addition to these arguments and analyses that largely describe and

contextualize this body of related but “artificially” collected7 texts, I will pose and

attempt to answer the following question: “What is yoga according to this

particular set of texts?” Yoga, as a term, has such an expansive semantic field in

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

Indian cultural practices and textual history, that it has meant almost anything

and everything. This Indian historical reality has led to the generation of

generalizations and universals among scholars and practitioners of yoga. In order

to redress this particular terminological difficulty, I will demonstrate that a more

useful approach is to analyze the term yoga in particular contexts and textual

traditions, in this case the Yoga Upaniṣads. Answering the question, "What is yoga

in the Yoga Upaniṣads," builds upon the general history of yoga provided below yet

simultaneously suggests that such general histories be continuously nuanced by

and ultimately replaced by specific textual and cultural histories. This argument

follows the models of recent works: for example, Patrick Olivelle's analysis of

another "artificial" but logical set of minor Upaniṣads, the Samnyāsa Upaniṣads.8 In

this way this study provides the foundation for a subsequent, more thorough

analysis, such as those found in Ian Whicher's detailed textual and philosophical

analyses of the Yoga Sūtra and its commentary traditions, or David G. White's

7 See discussion below, p. 19.

8 Concerning the categories of text and how they were collected in the previous century,

see discussion below, p. 16, and in chapter 2. The Samnyāsa Upaniṣads are a set of sister texts to the

Yoga Upaniṣads: they share similarities in dating and development. My study provides more

discussion and explanation of the yoga texts than Olivelle's work does for the renunciation texts,

yet it does not provide a complete original translation of the entire group of texts. Selected

translations from the texts are included throughout the study for the purpose of example,

description, and argument. Olivelle translated from Otto Schrader's critically edited texts. The

Yoga Upaniṣads have never been critically edited.

7

History, Text, and Context of the Yoga Upanisads: Chapter 1

comprehensive study of yoga, alchemy, and tantra.9

1.2. Method and Structure

This is a work about yoga. As such it is one voice in a chorus of thousands.

Unlike categories such as religion, myth, or asceticism, yoga is not singly an imagined category. Yoga is not solely the creation of the scholar’s study.10 In the

words of Jean Varenne,

[Yoga] is a “world view,” a Weltanschauung that comprehends reality in its totality—material as well as spiritual—and provides the foundation for certain practices intended to enable those worthy of it to integrate themselves totally into that reality, if not transcend it.11

However, there are many yogas and even the yoga that is the imagined re-creation of scholars, Asian and Western.12 These yogas are both the institutions and behaviors of human beings and the second-order categories of the human sciences. There are the human beings who practice these yogas, and the larger

9 Although it is beyond the scope of the present research, analysis of the Telugu character manuscripts, more in-depth manuscript histories of each text, and a comparison with more Āgamas, yoga texts, and Tantras would be the next step in a thorough presentation of the history and meaning of these texts. Christian Bouy has provided significant textual analysis of the source texts of many of the southern expanded texts. A comprehensive study would continue this process. It would additionally provide cultural and historical analysis of these sources.

10 For categories as products of scholarly activity, see J. Z. Smith (1982), p. xi.

11 Varenne (1976), p. ix.

12 Throughout the study I have differentiated uses of the term, “yoga” according to the following pattern: capitalized and singular, ‘Yoga’ to apply to the scholarly category; lowercase and singular for the South Asian worldview signaled by the term; and lowercase and plural to signal the many different traditions associated with specific gurus, textual traditions, or different geographic regions.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

number of human beings who do not practice. There are scholars who seek to

describe and understand yoga practices, and others who attempt to elucidate the

many accompanying philosophies, cosmologies, and epistemologies. Thus, there

are insiders and outsiders and etic and emic approaches, yet even these two pairs

are not the same.13 Insiders to the yogic traditions are those who believe in a

more-or-less coherent system and practice its rituals and technologies. Outsiders

are the remainders of humanity who neither practice any forms of yoga nor accept

its philosophies and ideologies. Etic and emic are scholarly viewpoints or

approaches. These are second-order activities removed from the lived behaviors

of insiders and outsiders. The etic viewpoint studies human behavior from

outside the tradition of its subject. The emic viewpoint examines human behavior

according to the tradition's own values, or from within the system. These

approaches are ways to study something, not descriptions of the actors who are

either inside or outside a given tradition.

Yoga and the yoga traditions are complex and temporally longstanding. The

topic begs attention to the many varied actors and their words, for yogins have

long exercised their own voices in constructing their traditions and writing their

13 The technical terms etic and emic were coined by Kenneth Pike in 1967, derived from the

suffixes of the words “phonetic” and “phonemic” and have the general usage in the humanities

and social sciences for methodologies that approach human behavior from outside or inside a given

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

texts. Even so, the once-removed stance of both etic and emic examination is also

necessary for understanding (interpreting and articulating the traditions) and

explanation (reducing, generalizing, and theorizing about the traditions). I have

sampled multiple voices inside and out and employed both etic and emic

approaches in order to present a well-rounded discussion of yoga and the yoga

traditions of South Asia.

My analysis involves three steps: description and interpretation of the texts

and their history, consideration of yoga as an intellectual category, and application

of the category yoga to the texts of the Yoga Upaniṣads. When clarifying what the

yogas of the Yoga Upaniṣads are I have maintained a close connection with the texts

themselves. Only after articulating the Yoga Upaniṣads traditions based on the

texts, their known sources, and the histories and traditions of their interpretation

do I attempt to view the tradition from the outside.14 Following J. Z. Smith, the

second step is to weigh and measure the exemplum of yoga of the Yoga Upaniṣads

in critique, or service, of the broader category of yoga. In the third step of this

process, I employ a series of methods (discussed below) for explicitly relating the

system. See Pike (1967). For a comprehensive assessment of the scholarly discussion concerning

these approaches see Russell McCutcheon, ed., (1999).

14 See J. Z. Smith (1982), p. xi, for the suggestion of an historian of religion following the

three-step method described in this paragraph.

10

History, Text, and Context of the Yoga Upanisads: Chapter 1

yoga of the Yoga Upanisads to the general category of yoga, and methods for

evaluating each in terms of the other.

This last step leads to my reconstructed history of these texts, and this is an

act of imagination or re-imagination. Unlike many Latin, Hebrew, or Chinese

genres, the South Asian authors of the Yoga Upanisads neither value the individual

merit of named authorship nor do they value the talismanic value of dating.

Reconstructing the history of the Yoga Upanisads is, in many ways, the formation

of a context. In addition to the technical aspects of dating (datable quotes, relative

dating), the imagined context and history are constructed by drawing parallels

between the yoga of the Yoga Upanisads and other yogas, as measured against the

broader category yoga. Remarkable continuity exists across time and traditions.

Descriptions of some of the many yogas and articulation of the general category

yoga reveal this continuity. To draw these many parallels and to construct a

history serve to demonstrate the persistence of meanings and practices within

altered expressions of the mysterious and the commonplace. However "it is an

abstract continuity for likeness is not sameness. In history everything wears its

own dress and raises images peculiar to itself."15 The yogas encompassed by the

category Yoga are not identical traditions; they are not the same. "The continuity

History, Text, and Context of the Yoga Upanisads: Chapter 1

among these various traditions is in their human motive." The human motive

consistent among all of these yogas is the yogins' attempt to follow a system of

body-mind exercises intended to transform an individual into something powerful

and different from other human beings.16

In this chapter, I postulate a definition of yoga as a category in a way that is

useful for asking questions and revealing information about South Asian practice

and ideology. In chapters 3 and 4 I describe the yogas from the Yoga Upanisads in

detail, interpreting how they fit the category and explaining what that reveals.

Comparisons are discussed and presented in chapters 2 and 4; wherein I compare

the yogas of the Yoga Upanisads to the yogas of other South Asian texts, traditions,

and time periods, demonstrating how they are similar and different. These

similarities reveal the strengths of the scholarly category of yoga, while the

differences reveal the unique nature of a specific tradition of practices and beliefs

as distinct from the broader category.

Throughout the study, and in particular in chapter 5, I explore different

theoretical interests. First, with specific theoretical concern for the Yoga Upanisads,

15The words and mood of the previous and following sentences are quoted or adapted

from Jacques Barzun's discussion of parallel themes and continuities across 500 years of Western

history. Jacques Barzun (2000), p.31.

16Whether the resulting state be that of a perfected one, enlightened one, awakened one, a

liberated-while-alive, or as a simulacra of the absolute reality or the god-source of the cosmos, all

of these states or identities are dependent on the yogin being rebom, remade, reengineered into

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

I offer explanations concerning why peoples practice this yoga, drawing from

metaphor theory, learning theory, and psychology, and attempting to explain

whose interests are at work (political, social, economic, and so on). Second, I

address certain questions and issues of theory: why I chose the approaches that I

employ, and how they are different or better than previous approaches.17

1.3. Approaches to Yoga

Examinations of the term yoga have a long history of study in both Indian

and Western popular and academic circles. These investigations have generated

many useful approaches to yoga and its accompanying literature. Many of these

studies are syncronic and comprehensive (for example, Eliade and Varenne).18

something less iike other human beings and more like the imagined gods, demigods, or primeval

forces of the cosmos.

17 Deconstructing discourses about yoga is neither of great interest to me, nor the primary

focus of my research. Herein, these theoretical concerns are addressed throughout the work,

where the issues arise, in order to demonstrate the place of my study in the larger body of

scholarship.

18 Scholarship on yoga has primarily been of two types: 1) general, comprehensive, a-

historical works; 2) specific textual studies devoted to philosophical or linguistic analysis.

In general approaches to yoga, the major authors that explore yoga do so

phenomenologically and synchronically, not historically. There are several but the following

authors are exemplary: Julius Evola: Italian scholar of the mid-twentieth century; Jean Filliozat:

French scholar 1930s-60s; Mircea Eliade: works on yoga in his French and American periods

(1950s-70s), and Jean Varenne, French scholar writing from 1970s to 90s. Guy Beck's work called

Sonic theology (1993) address several issues important to this discussion but misinterprets the place

of the YU in the medieval period. Georg Feuerstein is a prolific popular-academic writer who

appreciates historical sequence but does not delve deeply into any of the traditions.

In many ways all of these works are 'up in the air'—Evola has a subtle understanding of

kundalini and tantric yoga but generalizes across centuries and is singly interested in tantra not the

broader multiple traditions of yoga. Eliade relies on timeless categories: karma, yoga, nirvana and

History, Text, and Context of the Yoga Upanisads: Chapter 1

Although these works are helpful for elucidating different aspects of Indian

religious texts and practices associated with the category, yoga, considerable work

remains concerning the translation and elucidation of specific textual traditions.

Recent studies, such as those by Feuerstein, Whicher, and White, address some of

these issues and traditions in a more diachronic, historical fashion, but many

textual traditions remain unstudied. Feuerstein's work is chronologically

organized but is introductory, descriptive, and includes some inaccuracies in

history and translation. Whicher's work focuses on Patañjali and the commentary

traditions related to the classical Yoga Darśana. His introductory chapter provides

others that mix and match traditions and texts separated by geographic boundaries, local history,

and time. Varenne presents a completely Brahmanical yoga praising the early (2nd century) system

of Patanjali and treating everything else as degeneration from the 'golden age' of classical yoga.

Beck presents the chronology of texts exactly backwards and therefore attempts to say that yoga's

sonic theology influenced tantra's sonic theologies when it is certainly the other way around. In

general, his study makes numerous mistakes precisely because he is following the works of the

previously mentioned scholars instead of looking at the texts themselves. In many ways these

general approaches obscure the history of the traditions of yoga and they obscure an accurate

description of the traditions themselves.

There are numerous Indian authored works within this category of general approaches to

yoga. Surendranath Dasgupta (1920s and 1930s) and more recently Sures Chandra Banerji (1995)

have contributed what I would describe as emic approaches. These authors' primary concerns are

philosophical and descriptive. Their works are in-depth and incredibly detailed but lack western

historical methods or other etic, reductive, or explanatory analysis.

Alternative studies of yoga exist in specific studies usually in the form of linguistic

analyses detailed in specialized journals: for example, Boris Oguibénine's study of the term yoga in

the Vedas, and numerous studies by Hélène Brunner. These approaches are intellectually rigorous

contributions to the understanding of South Asian traditions but are not histories or

comprehensive presentations of a tradition or a category. One needs dozens of such detailed

studies to knit together a history and most of these authors are not attempting any kind of

theoretical enterprise. Unlike the hermeneutic I outline above they never do anything more than

define and describe. They don't really try to explain anything about humanity, behavior, history,

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

a good historical outline of the earliest periods but understandably does not assess

medieval traditions. White's work is comprehensive for the Siddha (and related)

traditions. The Yoga Upaniṣads are partially derivative of these two traditions, so

while neither Whicher nor White attempt extensive analyses of this latter set of

texts they do provide background materials that help to explain them. Christian

Buoy is the only scholar to devote special research to the relative dating of some

the Yoga Upaniṣads.19

1.4. Defining Yoga

Concepts and categories of yoga and yogic practice proliferate in the

religious traditions of India. The word yoga has several meanings. From the root

yuj, the term has considerable semantic field in the Sanskrit language. Some

usages parallel such English cognates as "union" and "yoke." Other valences

include "team," "sum," "equipment," "conjunction," "mathematical calculus,"

"harness," along with many other meanings. Early in the Indian traditions'

histories, the term yoga took on the meaning "spiritual endeavor," especially when

related to control, harnessing, or yoking of the mind, the senses, the breath, and

society, politics, and so on. These are the kind of studies that provide the foundations for general

statements and presentations, but as they are, are of little use to the general reader.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

the body. The term has continually maintained this general meaning of spiritual

praxis or sādhana.20 The valences, both as noun and verb, of “union” (unify,

unification) are employed as the action or result of these harnessing practices:

unification of the self, union with God; union with Brahman; or unification as

synonymous with liberation from suffering, karma, and rebirth.21

By the third century BCE, the term is in general use as a term for spiritual

disciplines (praxes) associated with several approaches (theoria) to self-realization

and self-empowerment.22 Although the term has continuously maintained these

varieties of meaning, in the medieval period a particular type of yoga, haṭha yoga,

was developed that informed all later uses of the term yoga. This evolution begins

in the yogas of the Tantras and alchemy (Rasa) texts of Buddhist, Jain, and Hindu

19 See Christian Buoy (1990a, 1990b, 1994, 1995). Buoy’s scholarship is meticulous and of

high quality. His relative dating analyses concentrates on dating and contextualizing several

minor Upaniṣads by means of quotations from other datable text.

20 Gerald J. Larson (1978). Review: Gaspar M. Koelman (1970), Pātañjala Yoga: From Related

Ego to Absolute Self. Philosophy East und West 28.2: 236-239.

21 For extended examinations of all the various meanings and implications of the word

yoga, see Dasgupta (1922); K. S. Joshi (1965); Eliade (1973); Whicher (1998); Feuerstein (1996, 1998).

22 A “canonical” list of the yogas of the Hindu traditions would list the following: jñāna yoga

(“the yoga of experiential gnosis”), bhakti yoga (“yoga of devotion to god”), karma yoga (“yoga of

acting in the world while avoiding the fruits of that action”), haṭha yoga (“forceful yoga” of the

physical body), mantra yoga (“yoga of the recitation of sound”), laya yoga (“yogu of dissolution”),

kuṇḍalini yoga (“yoga of raising the serpent power”), and nada yoga (“sound yoga”). Kuṇdalini yoga

and haṭha yoga are independent but associated approaches. Nada yoga is a subset of mantra yoga.

Although classical examples from the seminal and authoritative Bhagavad Gītā, bhakti and karma

yogas are yogas in the weaker, more general use of the term as “disciplines.” The others in this

category are more specifically mind-body, meditation technologies. There are many additional

emic and etic classifications: rāja yoga (“Royal yoga,” Patañjali’s yoga as interpreted by

Vivekananda), Buddhist yoga, Vedānta yoga, Jaina yoga, Sāṃkhya yoga, Integral yoga, Taoist yoga,

Tibetan yoga, and more.

16

sects and traditions, in which yoga is a prerequisite for proper ritual practices and

for meditation. These traditions culminated in the development of hatha yoga in

renunciant, monastic, and ascetic communities. Haṭha yoga, with its elaborate

postures, exercises, subtle physiology, hygiene and purifications, and meditation

systems, was then re-employed in the yoga systems of tantra and alchemy and

among devotional and Brahmanical sects. Haṭha yoga becomes part of theoria and

praxis in various sectarian movements—Śaiva, Śākta, Vaiṣṇava, as well as

Buddhist and Jain movements—and as the orthopraxes of philosophical

viewpoints both dualist and non-dualist.23

In Patañjali’s Yoga Sūtra (second to third century CE), the author posits that

yoga is employed to prevent thought from whirling fluctuations.24 Although

numerous texts, sects, and traditions have elaborated yogic practices, much of the

history of yoga is specifically devoted to elaboration of this apparently simple, but

ultimately quite difficult, process: individual control of the mind and mental

fluctuations through systematic embodied practice. This control is not solely a

23 Non-Hindu sects are not analyzed in this work. Haṭha yoga practices exist among

Buddhists, Jains, Sikhs, and Muslims.

24 Yogah citta-vritti-nirodhah: “Yoga is the cessation of the turnings of thought,” (YS I.2).

According to Barbara Stoler Miller (1995, p. 30), Citta is best translated as “thought”—the sensitive,

subtle aspects of mental capacity. Thought exists in the form of its activity, or “turning” (vritti).

"The turning of the thought" (citta-vritti) refers to the totality of mental processes—conscious,

subconscious, and hyperconscious—not simply to the faculties of intellect, recollection, or emotion.

... [The] thought process [citta] is a composite of mind (manas), intelligence (buddhi), and ego

History, Text, and Context of the Yoga Upanisads: Chapter 1

mental exercise. Instead, it is situated firmly in practices related to the body.

Body posture, breath, bodily function, social constructions of body, and

embodiment informed by psychosomatic and neurological realities all provide the

contexts of mental exercises and meditation. Mind and body control, although

fundamentally grounded in human neurological experience, evolve within

complex South Asian cosmological, ideological, and epistemological

environments.

Meditation and reflective practices are described from the Vedas onward in

various philosophical and religious schools. Yoga traditions undergo more

systematic development in their classical form with the appearance of Patañjali’s

Yoga Sūtra. However, the yoga traditions attain their greatest elaboration and

development in the haṭha yoga schools of the Nāth Siddhas of the medieval period.

The traditions associated with haṭha yoga’s elaboration and practice are beginning

to be understood, in terms of dating, geographic context, and historical

antecedents.25 There is another body of important yoga texts that require

considerable study and research: the Yoga Upaniṣads. These twenty-one short texts

present a corpus of mantra yoga and haṭha yoga. The Yoga Upaniṣads require a more

detailed analysis of the social world that produced them and an explanation of

(ahamkāra), the three mental evolutes of material culture (prakṛti). For detailed analysis, see Miller

(1995), pp. 29-31, and B. K. S. Iyengar (1996), pp. 45-48.

18

their ideas and practices.

The history of the Yoga Upaniṣads, when considered as a group of texts, is a

history of scholarly convention. These texts are called “Upaniṣads” but they post-

date the classical (or “major”) Upaniṣads by several centuries. Although the

classical Upaniṣads have received considerable attention from both South Asian

and Western scholars, the later (“minor”) Upaniṣads have been seldom studied

outside of India.

The Yoga Upaniṣads are grouped together in a somewhat arbitrary fashion.

Upaniṣads is a classification of texts meant to ascribe authority and to connect them

with the larger tradition of Vedānta. Since Indian traditions have historically

tended to avoid rigid closure of their canonical traditions, Upaniṣads have been

composed continuously from the first millennium BCE up to the present. Lists of

60 to 108 Upaniṣads are often considered canonical, although there are hundreds of

texts in this broad category.26 “Yoga Upaniṣads” is not an indigenous or sectarian

division of texts. Instead, this division was first suggested by Albrecht Weber and

later elaborated by Paul Deussen. The texts are thus named because of they share

common characteristics and themes: Deussen’s divisions include Samnyāsa

Upaniṣads, Vedānta Upaniṣads, Yoga Upaniṣads, Śiva Upaniṣads, and Viṣṇu Upaniṣads.

25 See White (1996), Feuerstein (1998).

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

The division is useful in two ways: first, it has existed long enough to be adopted

as a scholarly convention, both in the West and in South Asia, and second, the

divisions are based on themes within the texts that differentiate them from the

larger body of Upaniṣads. One drawback of these divisions is that they tend to

suggest a single or purposeful voice or producing-community, behind the texts,

that is problematic. Even so, just as the Saṃnyāsa Upaniṣads focus on the life of

renunciation and ascetics, so too, the Yoga Upaniṣads deal specifically with yoga

and are more like each other than they are like texts found in the other

conventional divisions.

Of the several approaches available, three techniques of analysis will be

employed in the discussion and explanation of this material: textual analysis,

historical analysis, and a contextual interpretation of the texts. Adyar Library

published a Sanskrit edition of the Yoga Upaniṣads with commentary by Upaniṣad

Brahman yogin in 1920. This edition is the only widely available Sanskrit edition

of the texts. The Adyar Library edition represents the seventeenth to early

eighteenth century South Indian redaction of the texts. These are the South Indian

texts that were expanded (often to many times their original size) by the inclusion

of various additional yogic and tantric materials. The Adyar Library edition has

26 In North India lists include approximately twelve classical titles and forty-eight minor

Upaniṣads, southern lists include classical and minor titles together in a list of 108.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

not been critically edited, and it does not contain critical apparatus. There is no

readily available Sanskrit edition of the shorter northern recensions.27

Many of the Yoga Upaniṣads exist in more than one English translation.

These previous translations suffer from specific difficulties and problems. The

single complete translation by T. R. Śrīnivāsa Ayyangār contains many errors. It

also uses archaic English that is often misleading or unclear unless read along with

the Sanskrit (1938).28 Ayyangār often interpolates much of Upaniṣad

Brahmayogin’s commentary (written 1751) into the translation without notice to

the reader. Other works do not translate all of the Yoga Upaniṣads and include

archaic English and some mistakes and inaccuracies, such as the 1914 translation

by K. Narayanasvami Aiyar.29 Aiyar also translated from South Indian expanded

texts. His translations were written before the Adyar Library edition was

prepared, apparently from a single set of manuscripts. A few other translations of

select Yoga Upaniṣads exist, some of quality, yet these are in different languages

and fail to provide a comprehensive view of the materials. Georg Feuerstein

27 There are at least two editions available, the Calcutta Bibliotheca Indica edition and the

Poona Anandāśrama Sanskrit Series. Neither of these is easily available to the average reader.

28 Often spelled without diacriticals, Ayyangar. His translation was prepared from the

Adyar Library edition of the texts.

29 This text has been recently reprinted with some corrections and the addition of Sanskrit

versions of the text. The Sanskrit texts differ in minor ways from the Adyar Library edition. These

Sanskrit texts are not critically edited, and they contain no critical apparatus. It is likely that his

original translations were prepared from the 1883 Telugu characters printed edition, but this is not

known with any certainty.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

translates a few of the Yoga Upaniṣads in his general survey of the yoga traditions of

India (1998). These translations were made from the Adyar Library edition. Jean

Varenne translated nine of the Yoga Upaniṣads into French (1971 and 1973).30

Deussen translated the eleven Yoga Upaniṣads of the northern tradition

(1897). Deussen’s translation is adequate in German, but suffers from double

translation (German translation: 1897; English translation of German 1980,

reprinted in 1997). Deussen’s work is instructive but limited to the shorter

northern recensions of the eleven oldest Yoga Upaniṣads. Deussen comments in

several places that he translated from Sanskrit manuscripts that were often corrupt

or unreadable. As described before, the northern recensions differ considerably

from the texts in the Adyar Library edition. Deussen’s manuscripts are often 10 to

20 verses, whereas the Yoga Upaniṣads with the same titles in the Adyar Library

edition may have as many as 50 to 400 additional verses.31

In the course of the following study, I have sometimes quoted from extant

30 The Yoga Darśana Upaniṣad translated by Varenne, has been retranslated into English in

the English edition (1976) of his 1973 work. Varenne translated the southern expanded texts,

probably from the Adyar Library edition.

31 The differences are instructive as in most cases the brief verses translated by Deussen

also appear in the longer southern editions. Deussen did consult the 1883 Sanskrit in Telugu

characters printed edition. The longer Adyar Library editions correspond closely with the Telugu

versions of the texts. It is not clear how many manuscripts Deussen consulted, but it is likely that

the Bibliotheca Indica (Calcutta) edition of the texts was the basis of his tranlsations. In addition,

Eknath Easwaran (1996) includes translations of the Amṛtabindu and the Tejabindu [sic] in his work.

These translations are similar to Deussen’s in their length and composition but either originate

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

English translations. When the extant translations have been found to be too

inaccurate or obscure, I have provided my own original translations from the

Adyar Library edition. This collection was not critically edited, but does include

the commentary by Upaniṣad Brahmayogin. I have also consulted the revised

edition of Aiyar's translations (original 1914; revised 1997) that provides the

Sanskrit for eleven of the texts.32 This revised edition is particularly useful because

it contains variants from the Adyar edition in some verses.

The specific historical contexts of these texts provide the most obstacles to

research. It is unclear who wrote them and when they were written. Individual

authorship is not as important as assessing the nature of the communities that

produced the texts and the practical uses of the texts within their communities of

origin. The texts are pervasively Vedāntin and non-dual in their metaphysical

sections. The texts do not possess overtly tantric rites or themes, although the

southern recensions quote Nāth texts and Tantras. Such details as these suggest

that these texts were produced in sects that did experiment with mantra yoga and

haṭha yoga but were not Siddhas nor were they tāntrikas. Instead, Brahmanical

communities, or other groups more concerned with purity than with power, likely

from yet another set of manuscripts or are very free in translation. Sanskrit is not included and

manuscripts are not cited.

32 In 2001 I was able to examine northern recensions of some of the texts from Calcutta.

None of these texts contained critical apparatus, nor were they edited.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

produced the texts. That they are titled "Upaniṣad" suggests a self-conscious

desire on the part of their authorial communities to tie their texts to the authority

and orthodoxy of the Brahmanical tradition.33 These types of details provide some

of the basis of both questions and analyses.

Two other techniques of dating analysis provide clues to communities and

dates: textual content, style, and quotation of other datable texts; and consultation

of manuscript catalogs. First, the style of the Sanskrit, and the use of technical

vocabulary and philosophical concepts assist the reader in determining

chronology (and in some cases, geographic location as well). The quoting of

known extant sources provides clues to authorship and dates: if all other

approaches yield scant results, relative dating against known texts provides a

general chronology. Furthermore, consultation of manuscripts and manuscript

catalogs provides a means of tracking of extant texts: locations, dates, and number

of copies.

Interpretation and explanation of the texts are dependent on the textual and

historical stages of research. Several initial questions and comments are already

possible from extant scholarly reflection. These initial considerations provide a

basis for the following questions: How do the Yoga Upaniṣads relate to the theory

33 For extensive discussion of the self-designation of upaniṣads in late medieval and early

modern period, see Brooks (1992), pp. 11ff.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

and practice of mantra in the Āgamas and Tantras? How do the Yoga Upaniṣads

relate to the haṭha yoga literature and particular religious communities, such as the

Nāth Siddhas and their texts, and the larger context of yoga, tantra, and alchemy?

The Yoga Upaniṣads are pervasively Vedāntin in style and in philosophical content.

What does this say about their communities relative to the larger yogic, tantric,

and alchemical context? Do these texts represent an independent but related

tradition of emergent haṭha yoga, or are they solely inheritors of the Nāth Siddha

materials? Within the Vedāntin framework, what are the specific theologies and

philosophies of these texts: Are they unified? For example, although many gods

are mentioned in the texts, there is little dominant sectarian emphasis or overt

theological concern.

It is these types of questions that this project addresses. In the following

chapters, I demonstrate that there are related groups within the larger category

"Yoga Upaniṣad." Several of the northern texts, especially the bindu titles, appear to

represent a unified Upanisadic tradition of mantra yoga. Among the southern

recensions, another subset of the texts extensively quotes Nāth materials,

providing elaborate descriptions of haṭha yoga and kuṇḍalini theory. There is a

third group extensively from southern yoga traditions that has a Vaiṣṇava

orientation. There is some overlap among the second and third groups.

The project has three primary goals: selected translation, assessment of the

historical and social contexts, and interpretation and explanation of selected

contents of the texts. One theoretical goal—a brief history of the textual uses of the

term yoga—is presented in the following analysis and will be reconsidered and

critiqued in the conclusion.

1.5. History of Yoga

There is no universal system of classification of yogic history. I present the

following system based on broad temporal divisions. These divisions begin with

the prehistory of yoga and then correspond to the appearances of self-conscious

schools, sects, or traditions.34 This historical summary is selective. As such, it

provides an outline to help the reader place the many texts discussed in the

following chapters in general temporal relationships with each other.

1.5.1. Proto-Yoga

Reconstruction of the period of proto-yoga is speculative because it is based

on uncertain and indirect evidence. Three sources exist for speculations on proto-

34 These categories follow general historical terminology (“classical,” “Medieval,”

“Modern”), which function adequately if the reader considers them as not exact matches for the

same terminology applied to Western history. In addition, employing these terms can lead to the

“classical” bias that some historians and philosophers display. I am not employing the term

classical to signal that Patañjali’s yoga is the seminal, original, or “best” of Indian yogas. On the

contrary, I will argue throughout that the Medieval yogas are novel, elaborate, and in many ways

original. They are not simple repetitions of nor degenerations from classical systems.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

yoga. These sources include speculative analysis of the seals from Indus Valley

civilization (3000 to 1800 BCE). The four Vedic Saṃhitās (ca. 1500 to 800 BCE),

Brāhmaṇas, early major Upaniṣads, and other Vedic texts provide some references

to yogic practice.35 Finally, comparative data concerning human ecstatic

experience and shamanic practices (both historical and modern) provide

additional clues for explaining the data of the previous two sources. These

materials are difficult to analyze, although the term yoga (and related words) is

used in the Vedic corpus. It is premature to call these reconstructed practices and

beliefs yoga, but they do provide context for the later developments.

Arguments for the existence of a proto-yoga in the Indus Valley are entirely

speculative. The evidence that does exist is based on seals and small statuary.

Some of these artifacts depict humans in postures that resemble yoga postures.

Indeed, this is scant evidence. Analysis of this type groups these clues that

indicate proto-yoga with various other Hindu images to form a quantitative

argument. There is a buffalo god, in seated posture, whom animals surround.

This image is often described as the “Proto-Śiva” or “Master of Animals” seal.

There is evidence of goddess worship, a characteristic of village worship and later

elite Hinduism. It is likely that there was a bull or buffalo cult that also possess

35 See Boris Oguibenine article on the root yuj in the RV. Oguibenine, Boris. “Sur le terme

yoga, le verbe yuj- et quelques-uns de leurs dérivés dans les hymnes védiques.” Indo-Iranian Journal

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

pan-Indic parallels. The prominence of water tanks in the largest cities of the river

valley are often interpreted in terms of later Hindu views of purity and practices

of bathing. These and other examples are employed to suggest a “Hindu” Indus

Valley. The argument follows that if all of these examples together suggest the

roots of Hindu practice and culture in the Indus Valley, then the images shaped in

yoga-like postures may (and should) be seen as evidence of yoga. Nevertheless,

there is a tremendous temporal gap between these suggestive clues and the

practices, stories, and beliefs of later yoga, tantra, and Puranic Hinduism.36 The

data is simply too ambiguous to have much explanatory value.

Proper sacrificial action, cosmology, and hymns (mantras) to the deities

(devas) are the primary focus of the Vedas. However, esoteric and meditative

concepts and clues also pervade these earliest repositories of speculation and

practice.

27 (1984): 85-101.

36 See Parpola (1983), p. 56. Archaeologists of Indus Valley and village India are less

disturbed by the gap of time then are textualists. The archaeological record, especially at the

village level, is consistent across the centuries for much of northern India. Different cultures

occupy the larger area, but much of the material culture is similar. This suggests that village

lifestyle in India has been marked by millennia of continuity and its ethos and worldview has

likewise maintained a certain consistency. Textual evidence for this kind of analysis is limited to

the Epics. The Mahābhārata holds numerous clues to the continuity between India and the larger

Indo-European world, as well as ample evidence for the cult of the mothers and the importance of

Rudra-Śiva and other “Hindu” gods. It is for Vedism and Brahmanism that Indus Valley offers

weaker connections. Evidence for some forms of continuity are greater than others. It is wishful

thinking to find yoga in the Indus Valley. Speculation on the widest selection of available evidence

might speculate that Indus Valley had some kind of ecstatic practices in its varieties of religious

behavior. There is no evidence to suggest meditation traditions or yoga.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

In the Vedas there are clear indications that the Vedic "seers" (r̥ṣis) were familiar with various methods that, upon being followed faithfully, were known to bring about a state of self-transcendence, a transformation of consciousness and self-identity beyond the limitations of the ego-personality, resulting in an exalted and expanded sense of identity and being. These methods or techniques were spoken of variously as dhyāna (meditation), dīkṣā (spiritual initiation), tapas (asceticism), and so on.37

Prajāpati, high god of the Brāhmaṇas, heats himself through austerities to create the world.38 The long-haired ecstatic flies through the air and drinks poison.39 Soma-willing seers are called "ecstatics," "shakers," "visionaries," and "singers."40 These and many other suggestive references provide at least enough evidence to ensure that the Vedic tradition was steeped in ecstatic and visionary behaviors. Ecstatics, visionaries, ascetics, and priests experimented with and elaborated practices that were later systematized and institutionalized as yoga.

Scholarship on comparative ecstatic behaviors and beliefs offers some

37 Whicher (1998), p. 9. Cf. AV XIX.43.1; RV IV.1.1; and YV II.2. See also Joshi (1965) p. 55, where he writes: We find examples of Vedic seers aspiring to reach the heavens or even for attaining Brahman, through dhyāna, tapas, etc. But, in all probability, these practices were in the beginning in a more or less fluid form, lacking elaborate classification and differentiation. Later on, they were organized into a system, and it was possibly then that the name "Yoga" came to be associated with it. The word "Yoga" is thus older than the discipline or system of philosophy which goes by that name.

38 As found throughout the Brāhmaṇas. See B. K. Smith (1989), pp. 50ff.

39 Rgveda 10.136.

40 muni, vipra, r̥ṣi, kavi. See Holdrege (1996), pp. 227ff. See also Gonda (1963), pp. 36ff. Monier-Williams (1988).

evidence for analyzing early South Asian traditions.41 The symbolic and

embodied description of fire and heat are one example. Although ecstatic internal

heat is not a universal characteristic of religious experience, the phenomena of

heat generation inside the human body is described as experienced in many

cultural settings from Northern Europe and across Asia and the Pacific, in Africa

and the Americas. Throughout South Asian traditions the phenomenon of

internal heat reaches considerable development.42 Although heat generation as

asceticism and as part of ecstatic experience occur as early as the Vedas (especially

as tapas), it reaches an elaborate and specialized form in the medieval kundalini

theory of the Tantras and the haṭha yoga texts of the Nāths. In their broadest

configurations, the descriptions of the generation of internal heat that appear

throughout the ascetic and yogic traditions of India are also characteristic of

healing trance experiences found among peoples across the globe as a product of

both enstatic and ecstatic body practices.

1.5.2 Preclassical Yoga

The second stage in the development of yoga can be termed the preclassical

41 Analysis of these materials is beyond the scope of this study. The following sources are

suggested to readers who want to examine this topic in greater details. Pentikainen (1997), Luckert

(1975, 1981), Wedenoja (1990), Staal (1975), Eliade (1964), Goodman (1998), Turner (1992), Taussig

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

yoga period. Preclassical or unsystematic yoga is depicted in the classical Upaniṣads

the early writings of the Buddhists and Jains, and the earliest materials of the

Sanskrit Epics (800 B.C.E - 200 CE). The Upaniṣads expound ascetic and mystical

practices conceptualized as an internalization of the Brahmanical rituals. These

texts provide considerable information on theories of ascetic practice concerning

the generation of internal heat through austerities (tapas) and offer speculation on

sonic theology as demonstrated in reflections on the sacred syllable om. Other

yogic concepts and practices are detailed, usually according to non-dual

metaphysics (later referred to as Vedānta).43 The epics, in particular the

Mahābhārata (composed in its present form between 400 BCE and 400 CE), present

details of yogic practice, especially in references to prānāyāma (breath control), and

evidence the growing shift of yoga toward the more elaborate metaphysics of the

Sāṃkhya system. This period is best characterized as a time in which yogic

practices appear in unsystematic form. The lastest of the classical Upaniṣads—in

particular the yoga of six-limbs found in the Maitri Upaniṣad—show the beginnings

of systemization similar to that found in the Yoga Sūtra. During this period, extra-

Vedic traditions, such as the Buddhists and Jains, experiment extensively with

(1987), Marusich (1991). See selected bibliography of secondary sources for additional authors and

titles.

42 See Kaelber (1989).

43 See chapter 2 for summary of yoga in the classical Upaniṣads.

31

History, Text, and Context of the Yoga Upanisads: Chapter 1

breath-control, concentration, and meditation practices that contribute to the later

developed traditions of yoga.

1.5.3. Classical Yoga

The third division in the history of yoga can be called the classical yoga

period. This phase recognizes the first systematization of yoga, as demonstrated in

Patañjali's Yoga Sūtra (written in the second to third centuries CE, but representing

an elaboration and systematization of ideas that precede the text).44 This period

sees yoga accepted as a recognized school of Indian philosophy, termed a Darśana.

Patañjali's text spawned a long textual tradition of additional literature and

commentaries. These yoga schools developed in close connection with the

Sāṃkhya system of metaphysics. This period begins around the second to third

century and continues until approximately the sixth century. Yoga in the classical

period is particularly associated with meditation techniques and breath control.

Yoga, from this point forward, is truly a complex body of philosophical

conceptions and practices particularly focused on control of the behavior, body,

44 The Yoga Sūtra contains a system of eight parts or limbs. These are yama, niyama, āsana,

prānāyāma, pratyāhāra, dhāranā, dhyāna, and samādhi. These may be presented in three groups and

rendered in English as behavioral disciplines: ethical principles such as nonviolence, disciplined

observances such as commitment and contentment. Next are body disciplines: posture or the

physical preparation of the body for practicing meditation, the control or exercise of the breath,

and the withdrawal of the senses from external objects and distractions. Last are the mental

disciplines: concentration, meditation, and the integrative trance of pure contemplation.

History, Text, and Context of the Yoga Upanisads: Chapter 1

breath, and mind for the purpose of spiritual liberation.

Although Patañjali’s work is generally understood as crucial to the classical

formation of the tradition, its connection to Sāṃkhya, rather than Vedānta,

disqualifies it as the exclusive origin of the system. Although Sāṃkhya

metaphysics is often the “physics” of many schools of thought in South Asia,

Vedānta characterizations become normative for many later developments. These

mixtures of metaphysical and conceptual systems in the history and developments

of yoga suggest that yoga did not originate in any one system. The Yoga Sūtra is not

an Ur-text. Instead, yoga practice is part of a shared South Asian ethos.45

1.5.4. Medieval Yoga

The fourth period in yoga’s long development saw the vast elaboration of

yogic ideology and practice: the medieval yoga period (from approximately

seventh to seventeenth centuries). This period brings developments in which the

classical formulations are transformed by institutional evolution, as well as

indigenous and non-Indic influences.46 These new developments include the

elaboration of medical theory and the introduction of alchemy, tantra (from the

45 Bronkhorst (1990 and 1996).

46 This is the period in which India had increased contact with China, South East Asia, and

the West via trade and cultural contacts overland and through maritime routes. Although it is

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

sixth to sixteenth centuries), and new yoga movements, such as the haṭha yoga of

the Nāth Siddhas. The Yoga Upaniṣads were composed during this period. Much

of the content of the Yoga Upaniṣads is dependent on mantra yoga and haṭha yoga

practices and concepts. The practices and ideas described in the Yoga Upaniṣads

were widely shared yogic systems by the fifteenth century. It is in the medieval

period that the elaborate internal geography and physiology of yoga are

systematized. Breath control, postures, and many other practices are defined in

detail and elaborated in complexity and number.

1.5.5. Early Modern and Modern Yoga

Yoga traditions undergo further transformations in the early modern period

(seventeenth to nineteenth centuries) and the modern period (beginning in the

nineteenth century and continuing to the present time). The early modern period

is characterized by the spread of the medieval traditions throughout India, but

also their beginning process of decline in competition with Islam, the Sant

traditions of Sikhism, and modern bhakti forms of Hinduism.

The modern period is characterized by a continuing disappearance of

viable traditions of tantra and alchemy. Haṭha yoga has continued to gain

outside the focus of this project, these contacts brought significant ideas and practices into the

laboratories (real and figuratively) of the esoteric schools of Indian yoga.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

prominence in India and has been exported globally. Vivekananda's "Raja Yoga"

powerfully asserted a Vedāntin interpretation of Patañjali that is likely the most

widely known interpretation of yoga in the world today.47 Further examples of the

ongoing developments of yoga are Sri Aurobindo's Integral Yoga and the Siddha

Yoga of Muktānanda. These modern systems generally follow the Tantras by

departing from world-rejection and asceticism. They often teach modified

systems of kuṇdalinī yoga and meditation. Aurobindo's system even combines his

modified tantric yoga with Western evolutionary theory. There are dozens of other

modern traditions of yoga that are globally known. These various yogas represent

diverse worldviews with diverse systems of practice that continue to evolve as

cultures and contexts evolve.

1.6. Periodization and the Yoga Upaniṣads

The discussion and analysis of the Yoga Upaniṣads properly belong to the

medieval period, but there is a wide range of possible dates within this period.

The texts themselves are not all from the same period, and some of them are

clearly composite texts whose component parts have different authors and

therefore divergent dates. The contents of the Yoga Upaniṣads borrow from texts

written between 800 and 1750, virtually the entire medieval period. The northern

47 Miller (1995), p. xi.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

textual tradition of Yoga Upaniṣads is well attested by the twelfth to thirteenth

centuries. The southern tradition is not attested until the seventeenth to

eighteenth centuries and its expanded contents include texts of various dates.

One of the fundamental problems concerns the connection between these

Upaniṣads and the larger medieval context. For example, the texts' conservative

and explicitly Vedāntin tone, suggests that they were produced community other

than the Nāth Siddhas. Many of the texts appear to have conscientiously stripped

much of the alchemical and overtly tantric elements from their characterizations of

mantra yoga and haṭha yoga.48 By testing this hypothesis I will demonstrate that a

conservative, academic, intellectual, Brahmanical community produced the Yoga

Upaniṣads. This conclusion is supported by the self-characterization of these

documents as Upaniṣads and not as Tantras or haṭha yoga text-manuals.

Several important topics appear repeatedly throughout the Yoga Upaniṣads.

(1) Two texts primarily describe a system called tāraka yoga that especially explores

visualization practice and a theory of lights and visions. (2) The most pervasive

system in the texts is mantra yoga, especially focused on the theory and practice of

om, hamsa, bindu, and nāda. (3) Breath theory and breath-control are fundamental

to these texts' yogas. (4) The texts express special concern for the nāḍīs (internal

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

pathways, called veins, arteries, or nerves). The late southern texts that borrow

extensively from tantrism and the Nāths show special interest in the meditation

practice called khecarī mudrā.

In my analysis I will be concerned with characterizations of esoteric

physiology that are not the same as the subtle physiology found in the systems of

the Nāth Siddhas. This topic is the primary focus of chapter 4. Brief consideration

will also be given to the different systems of divisions of the yoga path.

Traditionally in India spiritual paths are described in terms of divisions or folds

(aṅga, literally, "limbs"). For example, one of the best known of such

characterizations is the Buddha's eightfold path. Patañjali's classical formulation

likewise has eight limbs. The Yoga Upaniṣads offer several different schemes:

eightfold, sixfold, and even fifteenfold. The divisions are often presented as a

hierarchy of behaviors: beginning with moral and ethical observances and

practical issues, such as dietary restrictions, they progress to bodily practices and

culminate in meditation practices that lead to liberation and human perfection.

48 Tantric materials are especially included in the southern expanded texts. The Yoga

Upaniṣads' redactors reshaped these materials in many ways, and even direct quotations are no

longer in their original contexts.

History, Text, and Context of the Yoga Upanisads: Chapter 1

1.7. Concluding Remarks

Diachronic historiography of yoga provides a better description of the

traditions than are found in synchronic or essentialist histories. Diachronic

historiography also begins the analytical process—it has built-in questions,

complexities, and variables that often claim less and reveal more.49 It has often

been asserted and argued that India is timeless and without history. One of the

reasons that India has no “history” is because it served Brahmanical, colonial, and

at times, various philosophical interests for India to be timeless and a-historical.50

Eliade and Varenne and other scholars have thus been complicit in

processes that erase history and hide political-social interests. First, these authors

boldly put forward the myth of a golden age: yoga was perfect in the beginning

and degenerated through history, and therefore good practice should go back to

the original tradition expounded by Patañjali. This golden age myth exists in the

49 “Deconstruction” has long been an important theoretical code word for the human

sciences. Consistent with it's etymology it is a series of method's and theories devoted to seeking

discontinuities and challenging essentialist and synchronic theories and histories. A diachronic

presentation of the Yoga Upanisads and their history avoids some of the pitfalls of synchrony and

therefore requires less de-constructing. This is not to say that any construction of history is

immune to obscuring whatever facts there may be or that any history can be written that does not

serve some parties interests (thus the truism that all histories are political in one way or another).

Instead it is a matter of asserting that some ideas and constructs are better than others because they

account for more variables, are more intellectually rigorous, more open to revision, and do not

claim to offer a total explanation. Therefore in the following pages I attempt to present the best

possible history of the Yoga Upanisads, all the while not claiming to have presented “The” history of

the Yoga Upanisads.

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

history of Western philosophy and scholarship as well as in the Indian tradition

itself. These two myths reinforce each other and lead to the second problem: Who

"owns" the tradition? Investigating this second problem, "ownership" of the

tradition, involves an examination of how India has repeatedly experienced waves

of universal-ization, Brahmin-ization, Sanskrit-ization, Vedanta-ization, and

Hindu-ization. These processes are themselves neither synchronic nor completely

successful, but they always have political motives (although there are other factors

and motives as well).51

One goal of writing about the Yoga Upaniṣads is to articulate the co-option

of tantric yoga into Brahmanical Hinduism. There are complex arguments about

how and why the elite traditions of tantra disappeared from India. No one has

explored the role of co-option of tantric yoga by Brahmanical Hinduism as one of

the factors in this disappearance. White's "Introduction: Tantra in Practice:

Mapping a Tradition" provides a careful and articulate assessment of definition,

history, and analysis of the near total disappearance of elite tantra from India.52

My analysis and explanation of the history and context of the Yoga Upaniṣads add

50 Even today the "Hindutva" politics of modern India have employed the complex

metaphor and myth of a timeless and "Hindu" South Asia to perpetuate their own very modern

issues and interests.

51 There are parallels in the study of both Taoism and East Asian Buddhism. The

preference for the original, pure, philosophical Taoism compared and employed to judge religious

and folk Taoism offers a good example of what is still going on in the study of yoga.

52 White (2001), pp. 3-38.

39

History, Text, and Context of the Yoga Upaniṣads: Chapter 1

an important chapter to this broader discussion. In the late medieval and early

modern periods of South Asian history the spiritual and intellectual map of India

was being redrawn. Although admittedly a small and elite set of texts, the Yoga

Upaniṣads offer the occasion to examine and consider the forces in this redrawing

of the religious and ideological landscapes of the Indian subcontinent.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Chapter 2

The Classical and Minor Upaniṣads

Secret teachings on Hidden Connections

The thoughts of the Vedānta . . . became for India a

permanent and characteristic spiritual atmosphere, which

pervades all the products of later literature.1

2.1. The Categories of Upaniṣad and Vedānta

Like all texts classified as Upaniṣads, the Yoga Upaniṣads are authoritative in

the Brahmanical Hindu tradition. This authority derives from their classifications

as Upaniṣads and as Vedānta.2 In Brahmanical Hinduism “Upaniṣad” is the

category of Vedic texts that contain esoteric, mystical, and philosophical

speculation.3 According to traditional South Asian characterizations, the

Upaniṣads are the concluding sections of the corpora of Vedas. In this way, these

texts are literally the “end of the Vedas,” the Veda-anta. “The Upaniṣads, however,

came to be viewed in many traditions not merely as the last books of the Vedas but

also as the most important. “The very term Vedānta was understood to mean not

1 Paul Deussen (2000), p. vii.

2 See Patrick Olivelle (1992), pp. 3–11, for a similar introduction to the Saṃnyāsa Upaniṣads

(SU). The Saṃnyāsa Upaniṣads are a sister set of texts to the Yoga Upaniṣads and share many

characteristics in development and dating.

3 For a survey of the major philosophical themes of the Upaniṣads see the English

translation of Deussen’s Die Philosophie der Upanishads (1899): The Philosophy of the Upanishads,

(1906), reprinted in 2000.

41

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

just the end but also the summit and crown of the Veda."4 These texts are the

summit of Veda because they contained the secret (meta)physics5 and

soteriological knowledge of the Vedic schools and lineages (śākhā, literally

"branches"). Thus, the terms Vedānta and Upaniṣad connote ancient pedigree,

practical philosophy, and some of the most long-standing and influential

scriptures of Hindu traditions.6

2.2. Hidden Connections and Secret Teachings

The word upaniṣad is a technical term in early Indian esotericism. The

primary meaning of upaniṣad is "connection," "homology," or "equivalence."7

Since, according to the metaphysics of the Brahmanical schools, esoteric

4 Olivelle (1992), p. 3.

5 Metaphysics is both correct and incorrect in this usage. It is correct as far as (since

Aristotle) philosophy has employed the term to refer to the study of ultimate causes and the

underlying nature of things. For Indic thought, physics and metaphysics are not strictly separate

in the nondualistic systems of Vedānta. Technically, the sacrificial (Brahmanical) worldview is

dualistic; the Upaniṣads include both dualist and nondualist formulations. Thus, sometimes

physics/metaphysics is an accurate terminology, while in non-dual formulations all would be

"physics." This is not a trivial observation in as much as contemporary quantum mechanics is

usually called "physics" not "metaphysics" and thus for any worldview there are subtleties of

language about which the reader should remain aware. Since the Indic religious traditions

conceived that action by humans could affect the "laws" of the universe, then one must imagine

that they perceived their ritual and meditation sciences to be based on physics in the case of non-

dual systems of theory and practice.

6 The Vedas claim more ancient pedigree, and the later Bhagavad Gītā (BG) may be the most

widely read text in India today, but the corpus of Upaniṣads has commanded almost equal

authority for over two thousand years. Originally the specialized texts of Vedic schools, these texts

were given renewed authority by a series of philosophers, from Bādarāyaṇa (second century CE) to

Śaṅkara (ninth century CE) to Vivekānanda and other interpreters in the twentieth century.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

"connections" are hidden, a derivative meaning of upaniṣad is "secret." In this valence, upaniṣad especially connotes secret teaching, secret knowledge, or secret doctrine. The Upaniṣads overflow with examples of the connections between different aspects of reality: between the Self and ultimate reality, between words and truth, between sounds and creation, and between breaths and sunbeams, to name but a few. These connections or correspondences connect entities, things, and practices to structures and functions inside the human being: for example, the interiorization of stars, moon, and sun within the body, or the internalization of the sacrifice in asceticism and meditation. Thus, the semantic field of the term upaniṣad contains the denotation of esoteric connections and the connotation of secrecy. Therefore, a text called Upaniṣad is a secret teaching revealing the hidden connections that energize the matrix of reality.

Correspondence theory and interiorization theory and practice tend to develop together. Correspondence theory is fundamental to Vedic and Brahmanical thought; Indic religious thought explores and develops correspondence theory continuously within its esoteric traditions.⁸

Correspondence also provides the logic that underlies the development of

⁷ For definitions: Olivelle (1996), p. lii; Renou (1946); Falk (1986); Deussen (2000). These equivalencies (such as ātman = Brahman) might be called allofoms, after Bruce Lincoln's usage (1986).

⁸ Brian K. Smith (1994). Smith's study in Vedic classification and hierarchy examines this aspect of Indian worldview and ethos in detail.

43

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

interiorization theory and in particular the bodily internalization of ritual models.9

In brief, the (meta)physics of the Indic explorations into the nature of reality

employ the concept of multiple cosmic correspondences between the microcosm

(human), the mesocosm (mediating structure), and the macrocosm (divine).10

Exploration of the nature of these correspondences is the science of upaniṣad

(connection). Inherent in this ideology of multi-cosmic interrelation is the concept

of hierarchy. "The Upanisadic connections are hierarchically arranged, and the

quest is to discover the reality that stands at the summit of this hierarchically

interconnected universe."11

Correspondence is the theory; interiorization is one of its practical methods.

The body (ātman)12 and essential Self (ātman) are connected to the cosmos. The

ātman is ultimately identified with the absolute of reality, Brahman. This is not

simply theory; it leads to method for the sages. Knowledge of the connections is a

program for practical experimentation for activating (real-izing) the connections

within the consciousness of individuals. By the logic of hierarchical multi-cosms,

9 For example, the body and om as the two parts of the fire drill, or the oblation of breaths.

Śvetāśvatara Upaniṣad 1.14; Chāndogya Upaniṣad 5.19-24. For interiorization see Eliade (1973); pp.

111ff; Whicher (1998), pp. 11-21;

10 David White (1996), p. 15, § 2, p. 360. White's summary (pp. 15-47) of categories and

numbers in Indian thought presents a series of interconnected ideas that are employed across

Indian religious traditions. The Upaniṣads extensively draw on these numbers systems and

categories.

11 Olivelle (1996), p. lii-liii.

44

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

actions in the microcosm affects changes in the mesocosm or macrocosm, and vice

versa. Since the homologies are hierarchical, it follows that some connections are

stronger, more potent, and more concentrated than others are. Thus, the

Upanisadic traditions seek for the most potent homologies or connections. They

seek these connections intellectually but also through asceticism and meditation.

The strongest connections provide the most potent realizations of truth and

reality. These most potent realizations of reality lead directly to liberation and

unconditional immortality. Pure ignorance will lead to nothing but pain and

suffering. A lesser secret truth will lead to profitable rebirth. A greater secret

truth will lead the sage to liberated immortality.

From the Sanskrit roots upa + ni + sad, the terms together were formerly

misunderstood to mean "to sit close beside."13 Deussen and others understood

this definition as a title describing the educational setting of the texts. This

characterization has been so widely disseminated since Deussen (1890s) that one

will find it in almost every introduction to the texts, both Western and Indian. It is

12 The use of ātman for body demonstrates that the terminology of later Brahmanical

Hinduism had not yet been systematized in the early Upaniṣads. The term śarīra is the more

commonly used term for the body in later Sanskrit literature, and appears in the early texts as well.

13 See Monier-Williams (1988), p. 201. On this issue, Deussen (2000) was simply

overworking the etymology. Homologies are things that "cling," "sit close together," or "are those

things that are side by side." He overlooks his own evidence in seeking a literal "origin" for the

term.

14 Deussen introduces the subject succinctly by saying that Upaniṣad occurs with three

distinct meanings in the earliest texts: "secret word," "secret text," "secret import." See Deussen,

(2000) p. 16. Other terms of secrecy are used throughout the texts: guhya’ ādeśaḥ (Chāndogya

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

a misleading although largely harmless interpretation. It is not an altogether

useless definition since these texts were originally, and for a long time, oral-aural

teachings that were learned by students (śiṣyas) at the feet of their teacher (guru).

These "secret" (rahāsya) teachings were passed down by word of mouth to the

followers of a particular teacher.14 In the words of the Śvetāśvatara Upaniṣad 6.22:

"This supreme secret was proclaimed during a former age in the Vedānta. One

should never disclose it to a person who is not of a tranquil disposition, or who is

not one's son or pupil."15 Muṇḍaka Upaniṣad 3.2.11 states: "Here is the truth that

the seer Aṅgiras proclaimed of old. A man who has not performed the head-vow

may not learn it."16

A student perfects his esoteric intuition through learning and meditation

practice in order to perceive and recognize the hidden connections (upaniṣad)

between different philosophical and mystical concepts and between the different

levels of reality, microcosm, mesocosm, and macrocosm. The Vedic traditions

Upaniṣad 3.5.2), paramam guhyam (Kaṭha Upaniṣad 3.17; Śvetāśvatara Upaniṣad 6.22) vedaguya-

upanisatsu gūḍham (Śvetāśvatara Upaniṣad 5.6), guhyatamam (Maitrāyaṇīya Upaniṣad 6.29).

15 Translations of the classical Upaniṣads are adapted, sometimes with minor alterations

from Olivelle's (1996) translations unless otherwise noted. Alterations in translations are based on

my examination of the Sanskrit, and are generally employed to provide rendering of terms

consistent with the discussion and arguments of this study.

16 The exact meaning of "head-vow" referred to in this verse is unclear. Most scholars

argue that it is a vow of asceticism and connect it with the meaning of the title name, "Muṇḍaka,"

which literally means "shaven" or "shaven headed." Regardless of the exact meaning of the term,

this verse demonstrates the importance of secrecy.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

long employed the terms nidāna and bandhu to denote such connections,

analogues, and homologies.17 This analogical thinking provides much of the basis

for seeking knowledge of truth in the Indian mystical traditions. Operating in

such systems is the recognition that by knowing the systems of connections, one

could perceptually grasp the nature of reality and ultimate truth by a kind of

intuitional fusion. Connection, homology, and substitution provide a mystical

algebraic vocabulary for knowing the fundamental structures, nature, and

possibilities for reality, the cosmos, and the Self. As a technology of hidden

connections the Upaniṣads also afford almost limitless opportunities for

developing original methods for discovering the "reality of reality" (satyasya

satya). These speculations and technologies provide the fundamental metaphysics

of all of the elite traditions of Hinduism.

Brahmanical Hindus consider the Upaniṣads, like all Veda texts, to be direct

cognition of absolute reality (śruti) in contrast to "remembered" texts, or

traditional wisdom (smṛti).18 Thus the "classical" Upaniṣads (800-200 BCE), like the

Saṃhitās (1500-800 BCE), Brāhmaṇas (900-650 BCE), and Āraṇyakas (ca. 800 BCE)

are revelations of eternal knowledge (veda), eternal Word (veda, brahman, śabda),

17 See Brian K. Smith (1989 and 1994) for an in-depth analysis of these concepts in Vedic

literature. The term "upanisad" replaces these other terms in the Vedānta.

18 Śruti is literally "that which is heard." The primordial seers (ṛsis) "heard" and "saw"

transcendent truth in visionary trance and later communicated in words their direct experiences of

Veda (knowledge, truth, Word). The rise in importance of the Upaniṣads was an historical process,

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

and the limited manifestation of the unchanging, unlimited matrix of reality

(Brahman).19 The tradition divides these Vedic corpora into two further

classifications, the karma kāṇḍa and the jñāna kāṇḍa. Karma kāṇḍa—Saṃhitās,

Brāhmaṇas, and Āraṇyakas—is the division (kāṇḍa) pertaining to ritual procedures

and actions (karman). The jñāna kāṇḍa—Upaniṣads—includes the texts that deal

with liberating knowledge or transformative wisdom (jñāna).20

Before the common era, these texts were all learned orally together as a

body of literature that taught the twice-born, especially those of the priestly

classes, the ritual procedures and metaphysical speculations of their religious

traditions. These textual corpora developed over several centuries in North

India.21 They passed via the educational systems from father to son, or teacher to

student, growing, developing, and changing due to internal evolutions,

speculative experimentation, and social or material challenges and influences.

There were many other cultural and religious influences and developments that

similar to that of the Atharvaveda as well as other texts. Their acceptance and ascendancy was a

continuing process up to the medieval period.

19 The Vedic literature can be generally classified in four parts corresponding to the priestly

offices of the Soma rites: Ṛg, Yajur, Sāma, and Atharva. Each of these in turn comprises a Saṃhitā, a

Brāhmaṇa, and a Sūtra. The Brāhmaṇa(s) of each are further divided into three parts: Vidhi,

Arthavāda, and Vedānta (Upaniṣad). This system of classification evolved over time as the traditions

developed and grew. See Charles Malamoud (1996), pp. 169ff for discussion of vidhi and arthavāda;

and section 2.3 here. See also Deussen (2000), pp. 1ff. See also Barbara Holdrege (1996), pp. 29ff,

for a comprehensive discussion of these text genres.

20 Jñāna is cognate to the Greek word "gnosis" and to English "knowledge." In the Vedic

context its meaning is closer to gnosis than it is to English "knowledge," as the texts are explicit that

the knowledge gained is experiential and transformative, not simple intellectual cleverness.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

led to the later minor Upaniṣads, such as those of the class Yoga Upaniṣads.

Understanding the tradition of the older classical Upaniṣads is crucial for

understanding the meaning and authority of the medieval texts that are classified

as minor Upaniṣads.22

In the early Brahmanical context, the classical Upaniṣads can be understood

in two additional ways. (1) They are the texts containing the esoteric meaning of

the Veda. (2) They can be understood as texts embodying an evolution in Vedic

thought or even as a revolutionary shift of focus away from the rituals and

practices of the earlier Vedic traditions. First—and this is especially true of the

earliest Upaniṣads—these texts present themselves as carrying the real or secret

meaning of the Vedic traditions, texts, and rituals. In other words, when the

priests (hotṛ, udgātṛ, advaryu, and brahman) and patron (yajamāna) of the Vedic fire

rituals (yajña) recite, chant, or mutter and enact their rituals then there is a hidden

meaning and power in these acts. These secrets are the subject of the Upaniṣads.

The tradition does not see these hidden meanings as symbolic or metaphorical,

but rather they are the real purpose and power of Vedic practice.23 The Sanskrit

21 See Olivelle and his sources for the North Indian geographic origin of the classical texts.

Olivelle (1996), pp. xxvii-xl.

22 Although exact locations are not known in many cases, unlike the classical texts that

were produced within the Gangetic heartland of Brahmanism, the minor titles originate from

location all over the subcontinent.

23 In this way the Upaniṣads are not even novel, as the Brahmanas begin this process

commenting and explaining esoteric identification (nidāna) between victims and sacrificer, the

power inherent in meter and mantra, the path of a sacrificial knife equated with the path/bridge to

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

texts that emerge from this oral wisdom tradition represent the discussions,

opinions, and meditations of various individuals and schools united in their quest

for the "reality of reality."24

The second interpretation of the classical Upaniṣads, which has been

emphasized by most scholars, is that they represent a shift away from the magical

and ritual pursuit of sons, cattle, and heavenly immortality found in the karma

kānḍa.25 Instead, these texts embody the explorations of renunciant sages into the

nature of reality and the pursuit of liberation beyond both earthly and celestial

pains and pleasures. Moreover, they foster a rejection of the social and religious

norms of the ritual tradition.

It is likely that both interpretations of the texts are correct because they

recognize two tendencies in the texts. The former interpretation arises out of the

traditions themselves, yet it is also born out through scholarly investigation. Even

the celestial world, ātman in relivening of a victim, et al. Cf. Charles Malamoud (1996), pp. 169-

24 Lawrence F. Hundersmarck (1995), p. 155. Malamoud, Heesterman, B. K. Smith, B.

Holdrege, A. Parpola, and others have in different ways all presented pieces of a puzzle whose

whole picture shows a sacrificial tradition with centuries of accompanying mysticism, esotericism,

and even meditation. The Upaniṣads were written during the period in which breath, mantra, and

meditation were discovering a textual voice as an independent, self-sufficient system that came to

be called yoga.

25 This is an example of a seminal idea perpetuates itself, not always among specialists in

the Vedas and Brāhmaṇas, but among scholars of the later Indic traditions (such as Hinduism,

Buddhism, or others), in general introductions, encyclopedia articles, and the like. Scholarly

emphasis can also effect such theorizing. For example, where Fritz Staal has striven to accentuate

the importance of syntax and procedure in ritual, he has unfortuntely marred otherwise excellent

analysis with an overly dismissive attitude to meaning, knowlege, and "secret" connections and

interpretations.

50

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

the Ṛgveda contains enigmatic, mystical, and visionary speculations on the

unmanifest Absolute ("that One" [tad ekam]), the cosmic embryo (garbha), and the

cosmic man (Puruṣa). It also includes practical mysticism related to such concepts

as austerities (tapas) and the importance of the Word (speech, vāc).26 These

speculations already represent early embryonic speculative mysticism. Moreover,

with the speculations on cosmogony, cosmology, cognition, the Word, and the

nature of Brahman found in the later Saṃhitās27 and the Brāhmaṇas, it is easy to

understand the Upaniṣads as the extension or end of these speculations, not a

rejection of them. This is especially true of the Bṛhadāraṇyaka, Chāndogya, Taittirīya,

Aitareya, Kauṣītaki, and Kena Upaniṣads.

Chāndogya Upaniṣad 1.1.9-10 states:

(9) . . . the Adhvaryu priest says "Oṃ" before he issues the call; the

Hotṛ says "Oṃ" before he makes an invocation; and the Udgātṛ

says "Oṃ" before he sings the High Chant. They do so to honour

this very syllable, because of its greatness and because it is the

essence. (10) Those who know this and those who do not both

perform these rites using this syllable. But knowledge and

ignorance are two very different things. Only what is performed

with knowledge, with faith, and with an awareness of the hidden

connections (upanisad) [the hidden connection is that the secret

meaning of Udgīta is Oṃ] becomes truly potent.28

26 Holdrege (1996), p. 35.

27 In the usage as "secret text", the Taittirīyaka school ends certain sections of its text with

iti upaniṣad demonstrating that the word "upanisad" was in use even before the formalizations of

the genre.

28 See also Deussen, (2000) p. 16.

History, Text, and Context of the Yoga Upanisads: Chapter 2

The authors of these verses do not reject the sacrificial tradition. Instead,

they insist that the practitioners have an inner awareness of secret powers of the

rites. The earliest Upanisads derive from the oral gnosilogy of the Vedas. When

enacting the rites the priests and patrons were to be in a state of lived wisdom and

knowing contemplation during the rites. This secret experiential gnosis (jñāna) of

the rites energizes and strengthens the efficacy of the rites.29

Seeing the Upanisads as a revolution in the Brahmanical world—one

looking away from the theurgic efficacy and orthopraxy of Vedism, and looking

forward to the concerns of post-Vedic Hindu traditions—is a position that is also

found in the Upanisads themselves. Kārma, māyā, moksa, and yoga are fundamental

ideas in later post-Vedic traditions. For Vedic India, fundamental concepts

include heat/austerity (tapas), initiation (dīksā), sacrifice (yajña), knowledge (vidyā,

29 The strength of this interpretation lies in a comparison between the earliest Upanisads

and the earlier Samhitās and the Brāhmanas. Thorough explanation is outside the purposes and

arguments of this work. For exploration of the gnostic and mystical trends in the earlier tradition

that anticipate the Upanisads see Jan Gonda (1963), Bronkhorst (1993 and 1998), Walter O. Kaelber

(1989), Holdrege (1996). Ultimately there is not enough evidence to dislodge the “revolutionary”

theory of the Upanisads, yet the combination of the two theories discussed above appears equally

plausible. Opposition to such a theory relies heavily on two notions. The first notion is promoted

by Fritz Staal (and others) who argues that the early Brahmins were radical ritualists, concerned

only with the magical efficacy of the rituals and their strict ritual and mechanical implementation

(the “meaningless ritual” school of thought). The second as argued by Jan Heesterman (along with

a long tradition in European scholarship especially influenced by the early study of Indian

Buddhism) is that the Upanisads must be seen as a radical shift from the entrenched ritualism of the

Brahmins (the “axial age” theory). Staal and Heesterman present these theories throughout their

many works, see especially Staal’s (1993) Rules without Meaning: Ritual, Mantras and the Human

Sciences, and Heesterman’s (1993) The Broken World of Sacrifice: An Essay in Ancient Indian Ritual.

History, Text, and Context of the Yoga Upaniṣad's: Chapter 2

jñāna), and homology (nidāna, bandhu).30 Although the Upaniṣads look forward

more than backward, they are clearly and powerfully the corpus of texts that

represent the doorway open-but-separating these two milieus of Indic ethos and

worldview.

Deussen emphasizes these two tendencies in the texts: the esotericism of the

Vedic rituals and the later rejection of the supremacy of ritual. However, he

reverses the order of influence. He judges the anti-Brahmanical and Śramaníc

trends of the Upaniṣads, especially found among Kṣatriyas as primary.31 According

to his analysis, interpretation of the Upaniṣads as a gnosiology of Vedic thought

and rite was a later phase. Thus, Deussen and other scholars generally interpret

this evidence to conclude that this phase was motivated by the Brahmanical

attempt to co-opt the śramana-kṣatriya forest-born philosophy and mysticism back

30 See Eliade's(1973, p.3) presentation of the fundamentals of Hindu thought. See Kaelber

(1989, p. 1) for discussion of Eliade's ideas and the suggestion of the Vedic alternatives. Eliade

employs the term nirvāna instead of mokṣa. Scholarship often associates the former more with

Buddhist thought and the latter with Hindu thought. The Indic traditions sometimes use both

words interchangeably without the strict sectarian emphases associated with contemporary

scholarship.

31 Śramana is a term often used to refer to the ascetics of the forest traditions that influenced

the Upaniṣads, Buddhism, and the Jains. See Bronkhorst (1993, 1998). The kṣatriyas are the noble

class according to the varṇa social classifications of Sanskritic culture. Within the dialogues of the

Upaniṣads, the doctrines of Brahman and ātman are often favored by kṣatriya kings (for example

king Janaka in BĀU). Deussen argues that it was in the royal intellectual-philosophical circles that

these doctrines were first formed. Subsequently the priestly class adopted these ideas to "stay up

with the times." These arguments are not conclusive, for even in the given example, it is

Yājñavalkhya, a Brāhmaṇa who gives the doctrine that is rejected by Brāhmaṇas although accepted

by Janaka. See Deussen (2000), pp. 8, 17ff. See also Olivelle (1992), pp. 33-38.

53

History, Text, and Context of the Yoga Upanisads: Chapter 2

into their own tradition. In other words, the Brahmins brought these doctrines

into their sphere of power and control.

Ultimate answers concerning the origins of these traditions may be

irrelevant to our primary goal of exploring the medieval heirs of the old Upanisads.

The traditions of the śramanas and ksatriyas were so thoroughly mixed with

Brahmanical thought by the beginning of the common era that the medieval

traditions inherit the tradition as a synthetic whole. This synthetic whole was

never consistent and without contradictions, but its heirs accepted its authority

and unity, attributing contradiction to inability to understand its hidden truth.

According to this understanding, contradictions are only surface appearances to

those who know that upanisad is the secret science of seeing through

contradictions and differences to find the hidden connections.

2.3. Philosophy of the Classical Upanisads

Although varied and often contradictory, the classical Upanisads provide

the foundation for the system of Vedānta that became the most widespread basis

of Indian thought even up to the present day.32 Deussen divides this system of

32 Both insiders and outsiders of the Indic traditions often stress Vedanta principles as

universal for all systems and schools of Hindu religious thought. This point can be overstressed,

as it tends to obscure the influences of non-Hindu traditions (Buddhists, Jains, and others) and to

obscure the importance of the many non-Vedānta systems of the sectarian Āgamas and Tantras. If

these exceptions are not ignored, then it remains accurate to say that in pervasiveness of influence

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Vedānta into four primary divisions: (1) theology, (2) cosmology, (3) psychology,

and (4) soteriology and ethics.33 The doctrine that Brahman is original, eternal,

and unchanging is the pinnacle of the theology. Brahman is the transcendental

and immanent absolute source and basis of reality.34 This worldview

distinguishes manifestations of reality in terms of volatility verses permanency.

Change, fluctuation, and transformation characterize phenomenal reality.

Stability, eternality, and unity characterize the ultimate source of all reality. The

Upanisadic worldview turns upon the distinction between things that are ever

changing and in flux and the underlying unchanging reality of Brahman. The

tradition considers the visible fluctuating manifestations of reality as insecure,

undependable, and undesirable. With this rejection of the instability of change,

the tradition seeks knowledge and practical methods for escaping the "spinning

chariot wheel" of changing reality. The hopes and the goals of the sages are to

discover and realize the unchanging realm of Brahman.

both in time and across traditions Vedānta did become the primary foundation for Indic faith,

knowledge, and practice.

33 Deussen posits and analyses these divisions, (1906), p. 526f. The summary above is

essentially Deussen's, although I have revised the language because in some valences Deussen's

readings are overly Kantian (or even overtly Christian). Where Deussen reads eschatology, I have

added soteriology. Deussen's "ethics" is an acceptable term, but it should be read in the more

comprehensive sense as employed by Kant. The common-sense usage of the term ethics may be

too restrictive to include its widest range of philosophical meanings.

34 Since Brahman is ultimately "everything" it is not surprising that these heterogeneous

texts describe Brahman in numerous ways: "a formulation of truth," the Veda, the life breath,

speech, and the "origin" are but a few.

History, Text, and Context of the Yoga Upanisads: Chapter 2

Upanisadic cosmology describes and examines the doctrine that Brahman

evolves to manifest the universe. Structurally, this is an emanational cosmogony.

The universe evolves (or devolves) from the "One," Brahman, into the multiplicity

of the phenomenal world.35 The doctrine of the existence of Brahman as the

essential Self, or ātman, in all living creatures in the universe, evolved from

Brahman, provides the basis for Upanisadic psychology. Soteriology includes the

doctrines of the nature of ātman relative to birth, life, death, and rebirth. Ethics or

practice includes the manner of life and type of behavior required to realize the

nature of Self and the nature of the cosmos to awaken to the reality of Brahman.

The thought and practices that embodied these divisions of the Vedānta

developed and explored a series of important concepts. These explorations

include the identity of Brahman and ātman, the theory of reincarnation, the

necessity of liberation from repeated births, and the practices and lifestyle that

lead to this liberation—meditation and renunciation.36 Beyond these concepts, the

Upaniṣads contain a range of other truths, ideals, and goals. These selected few

35 This evolution is likewise not systematic across the texts. Some of these cosmogonic

systems include: ritual cosmogony, meditation cosmogony, creation from speech or mantras, proto-

Sāṃkhya physics, and theistic cosmogony like that of the Bhagavad Gītā.

36 A thorough analysis of the broad system of Vedānta would include karma. One finds

karmic concepts, but the system and mechanisms of karma are systemized later than the earliest

Upaniṣads. They are well developed in the early centuries of the common era among the Jains and

Buddhists and in the Hindu books of dharma and in the Bhagavad Gītā. The later Yoga Upaniṣads

offer extensive presentation of an already well developed view of karma.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

become the focus of the later Vedānta that the Yoga Upaniṣads and other minor

Upaniṣads explore.

First among these concepts, the sages declare that the absolute reality,

Brahman is identical with the essential Self, ātman, in each individual.

In the beginning this world was only Brahman, and it knew only itself (ātman), thinking "I am Brahman" (Aham brahmāsmi). . . . If a man

knows "I am Brahman" in this way, he becomes this whole world. Not

even the gods are able to prevent it, for he becomes their very self (ātman). So when a man venerates another deity, thinking, "He is one,

and I am another," he does not understand.57

The complex Brahmanical notions of cosmology and hierarchy pervade the

Upaniṣads. For all their divergent subjects and goals, the Upaniṣads hold that the

pinnacle of the cosmic hierarchy is the ultimate and basic essence of all of reality,

Brahman. The characterizations of Brahman are diverse, even including that

Brahman is beyond classification. A fundamental speculation of the Upanisadic

sages is that Brahman always retains its oral-aural character: Brahman is

śabdabraḥman, the "sound" of truth or reality. The individual sage must appreciate

this notion and seek the experiential knowledge of it in order to know the "reality

of reality." Thus, this theme emphasizes both the model of reality and the model

for knowledge and action.

57 BĀU 1.4.10. See also Müller (1962), p. 88.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Second, this ātman, being eternal, and embodied again and again through

repeated births, punarjanman.38 The earliest Upaniṣads refer to this process by the

negative assessment "repeated death," punarmṛtyu.

It is like this. As a caterpillar, when it comes to the tip of a blade of

grass, reaches out to a new foothold and draws itself onto it, so the self

(ātman), after it has knocked down this body and rendered it

unconscious, reaches out to a new foothold and draws itself onto it. . . .

After it has knocked down this body and rendered it unconscious,

makes for himself a different figure that is newer and more

attractive—the figure of a forefather, or of a Gandharva, or of a god, or

of Prajāpati, or of Brahman, or else the figure of some other being.39

Third, the Upaniṣads maintain that the natural course of endless rebirth is

negative. The sage who knows the secrets should escape this cycle of rebirth

through realizing the true nature of his Self, ātman, and realizing the identity of

ātman with Brahman. The concepts of rebirth and its driving power, karma,

emphasize different values but operate according to the same logic as Brahmanical

ritual practice. In the Brahmanical system

rites achieve their results by their own autonomous power and

according to a ritual law of cause and effect; ritual success does not

depend on the will of a god. The moral law that governs the rebirth

process operates in a similar manner; those who perform good actions

are rebom in good situations, while those who do the opposite proceed

to evil births.40

38 ChU 5.10.7. See also BĀU 4.4.5-6; 6.2.2,9-16; ChU 4.15.5;5.3.2; 5.10.5; 8.15.1; AU 2.4; KSU

1.2; KāṭhU 1.6;3.7-8; 5.7; ŚvetU 6.16; MunḍU 1.2.7-10; 3.2.2; PraśU 1.9; 5.3-4.

39 BĀU 4.4.3-4.

40 Olivelle (1996), p. xlvii. See also Deussen (2000), pp. 313ff.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

The Upaniṣads, just as they seek the eternal, unchanging Brahman, seek a permanent solution to the vicissitudes of karma and retribution. Although proper ritual action (also denoted by the term karma) can win a good rebirth after a sojourn in the realm of the moon, the Upaniṣadic sages seek freedom from all rebirths, whether good or evil. The liberating knowledge of the Upaniṣads is that there is an eternally blissful realm beyond the vault of the heavens through which one must pass to attain liberation. Accordingly, “The sun is viewed as a lid that covers the only opening in the vault of heaven, the only door to freedom; the sun permits the liberated individuals to pass through that opening and escape to the immortal condition outside the universe.”41

The liberation that the sage of the Upaniṣads seeks can be experienced at death if the sage has properly prepared through meditation and asceticism. Liberation can even be accelerated by these same means. These practices are diverse but have the following general pattern. The sage must go beyond plurality and flux and realize directly the unchanging nature of reality. One practical means to liberation is removal of all desires and attachments. There are internal means and external means. Internal means involve meditation practices through which the sage withdraws the organs, vital functions, and breaths from the objects of the senses and concentrates them within the body-self. Through this practice the

41 Olivelle (1996), p. xlviii. See also Deussen (2000), p. 218. See Iśa Upaniṣad for a concise

59

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

individual seeks to concentrate all of his powers within the essential Self (ātman)

and to literally shake himself free from the phenomenal world of reality by

latching on to the hidden truth of eternal Self. Renunciation and asceticism

provide external and social means to escape. By rejecting home, possessions,

family, social identity, and even the body itself, the sage cuts the external ties that

bind him to the phenomenal cycles.

2.4. Key Concepts in the Classical Upaniṣads

There are several concepts found within the classical Upaniṣads that the

minor Upaniṣads further develop. Of these many different ideas, a few are of

particular importance for the development of the yoga traditions in general and

specifically for the Yoga Upaniṣads. Concepts concerning Upanisadic physiology

and psychology, different stages of consciousness beyond waking and sleeping,

speculations on the mantra om, and meditation practices (yoga, tapas, dhyāna, etc.)

are the most relevant mystical embryos that mature in the later textual traditions.

The focus of the Upaniṣads is the human person: the construction of the

body, its vital powers and faculties, the cognitive processes, and the essential core

presentation of this doctrine and its imagery.

60

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

of the human being.42 From the earliest records of the Brahmanical tradition, in

Rgveda 10.90, creation, the sacrifice, and the body of the cosmic person, Puruṣa, tie

views of the cosmos to the form of the human body. With the interiorization

found in the Upaniṣads, the human body is the laboratory of the sages of the

Upaniṣads.

The most significant terms concerning embodiment in the Upaniṣads are

ātman and prāṇa.43 The term ātman, as discussed above, refers to the essential Self

or innermost essence of the human being. In addition, in the Upaniṣads the most

common term for the "living, breathing body" is also the term ātman.44 Ātman can

thus refer to both the human body and the eternal, inner Self. Prāṇa means

"breath" or "vital energy." This word also has a plurality of related meanings.

The term connotes the multiple "breaths" or the five essential agents of life force.

It also means the vital functions or vital powers of the body—breathing, thinking,

seeing, hearing, speaking (also movement, the senses, and other powers).45

Physical respiration was the most important of all of these animating functions

42 Olivelle (1996), p. xlix. See ‘Human Physiology and Psychology,’ pp. xlix-li for Olivelle’s

brief but insightful discussion. For in-depth analysis of these topics, see Deussen (2000), pp. 256-

43 Both terms are derived from the verb root an, which means "to breath," or "to live."

These are two important selected terms. Sarira, "body," is also an important term in the Upaniṣads,

and as terminology becomes systematized ultimately replaces the use of ātman as an term for body.

There are many additional concepts of embodiment in the Upaniṣads. For one example, these texts

also develop the concept of manas (mind).

44 Olivelle (1996), p. xlix.

45 BĀU 1.5.21. See Olivelle (1996), p. l, and Deussen (2000), pp. 271-274.

and powers. The texts equate breath with life and even with the essential Self,

ätman.

Fundamental to the later yoga traditions, the classical Upanisads distinguish

several forms of breath in the body. For this early formulation, there are five

different breaths: pràna, apäna, udäna, vyäna, samäna.46 For these texts, the human

body is alive by virtue of a multitude of fluidic winds that move through arterial

passageways. Thus, circulation, breathing, and digestion are “breaths.”

Important references describe the primary life-breath-energy of the body as either

“the breath within the mouth” (BĀU 1.3.7-27) or “the central breath” (BĀU 1.5.21-

22).47 Without this primary breath, the body dies.

In these early conceptualizations of the body, the heart is the organ of

extensive speculation on the part of the sages. The sages examine the nature and

roles of the seven openings of the body, especially the eyes, and the head, yet the

heart is the most consequential organ.48 For this worldview, the heart is the

location of the mind (manas). The heart is the locus of the five pränas (both in the

sense of five vital functions and five breaths). The texts call the heart the “cave” or

46 Olivelle (1996), p. I. The words always imply both substances and actions. The

meanings are not fixed but generally these breaths are: out breath, in breath, breath that moves up,

the breathing that traverses, and the linking or equalizing breath. These conceptions elaborated in

yoga, Indian medicine (Āyurveda), and tantra developed elaborate systems of breath/vital energy

that were pervasive wherever Indic traditions spread. In the body laboratory of yoga

understanding and manipulating pràna is the fundamental theory and practice. See also Deussen

(2000), pp. 274-280.

47 See Olivelle (1996), p. li.

62

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

"cavity" (guhā) in which ātman resides as the hidden Brahman or as the minute

immortal golden person, Puruṣa.49 Therefore, the heart is the center, anchor, or

foundation of the living being. The heart is the seat of the senses, mind, breath,

and essential Self (indriya, manas, prāṇa, and ātman).50

Bṛhadāraṇyaka Upaniṣad 2.1.19 presents the first articulation of the

importance of multiple pathways of vital energy (prāṇa) within the body: "There

are seventy-two thousand veins named Hitā that run from the heart to the

pericardium. He (Puruṣa) slips out of the heart through these veins and rests

within the pericardium."51 The five colored fluids that fill these pathways are

counterparts to the five rays of sunlight.52 Through this connection between veins

and light, the heart of the human being connects to the liberating sun. At death

the sage who "knows" this connection gains liberation. It is along these pathways

of color and light that the essential Self travels to the sun, the path of liberation.

The texts themselves do not provide a systematic explanation of the

mechanisms operating in the bodily microcosm and solar macrocosm. I interpret

48 Deussen (2000), pp. 286ff. Olivelle (1996), p. li.

49 BĀU 2.1.17; TaitU 1.6.1. Deussen (2000), p. 288.

50 See Gonda's (1963), pp. 276-288, for his discussion of the nature and function of the heart

in Vedic texts.

51 Based on commentaries, Müller reads "body" instead of pericardium. This is likely a

later extrapolation. The pericardium is the fluid filled membrane that surrounds the heart. These

veins called Hitā are mentioned in several places: BĀU 4.2.3; 4.3.20; 2.1.19; ChU 8.6.1, 3; KāthU, and

KSU 4.19. This number of 72,000 pathways is the foundation of the later developments of subtle

body (especially concerning the nāḍīs) theory in the medieval yoga traditions.

52 See BĀU 4.3.2 for the five colored fluids.

these many different passages as presupposing the following processes. The

essential Self is transformed into heat or light (taijasa/tejas: radiance) and then it

escapes the body at death via the colored pathways surrounding the heart. In the

Chāndogya Upaniṣad, the process of involution proceeds from speech to mind,

mind to breath, breath to radiance, radiance to essential Self (ChU 6.8.6).

Following the metaphor (in ChU 8.6.1-5): the sage (one who knows and practices

the secrets), whose nature is essentially breath, slips at death into the heart-veins

that are connected to the sun. In the form of radiance, he escapes the body with

the sound "om." He then becomes one with the sun and is not reborn again. In

practice, the sage focuses his speech and mind with the mantra om. Next, the om-

breath becomes hot and radiates along the hitā-veins, as the heat-self, escaping the

body and mingling with the heat rays of the sun. The heat-self follows these rays

to the realm of the immortal sun. Here we see the depth of the term upaniṣad in its

sense as "connection." The connections between the veins or pathways and the

rays of the sun are not just intellectual correspondences but instead are essential

connections that lead to liberation.

The early descriptive metaphors for the process of liberation are not

systematically fixed. Further in the same passage, Chāndogya Upaniṣad 8.6.6

presents an alternative view, quoting from another, probably oral, source:

One hundred and one, the veins of the heart.

One of them runs up to the crown of the head.

64

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Going up by it, he reaches the immortal.

The rest, in their ascent, spread out in all directions.53

Accordingly, we find two similar theories, both of them posited for explaining the bodily mechanics of liberation. One system describes 72,000 veins (finer than hairs) running from the heart to the pericardium, in which the essential Self rests in deep sleep. These veins connect directly to the sun via its five rays and thus lead to liberation. The om mantra is either the sound that the Self makes as it departs the body or (more likely) the secret trigger that produces heat and thus activates the connection with the rays of the sun. Unelaborated but for the single verse, the second theory describes 101 veins. One of these leads to the crown of the head (it does not say where it originates). Following this vein, the self wins immortality.

This emphasis on the head connects with two ideas from the Bṛhadāraṇyaka Upaniṣad and the Taittirīya Upaniṣad. At death, the essential Self, together with the breaths, departs from the heart along a channel that exits the body. Either the ātman exits through the crown of the head (TaitU 1.6), or it exits through the eye, head, or other part of the body (BĀU 4.4.2).54 I would suggest that these two verses connect the speculations on the heart and veins in the following way. The Bṛhadāraṇyaka Upaniṣad 4.4.2 relates to the isolated 101 vein verse from Chāndogya

53This quote recurs in the KāṭhU 6.16.

54Olivelle (1996), p. li.

Upaniṣad 8.6.6. Chāndogya Upaniṣad 8.6.6 states, "The top of the heart lights up,

and with that light the self exits through the eye or the head or some other part of

the body." This verse does not describe which exit is superior but simply states

that there are various exits. The quote in Chāndogya Upaniṣad 8.6.6 suggests that

there are multiple exits ("The rest [of the veins], in their ascent, spread out in all

directions"), yet it privileges the head-vein as the path to immortality. The

Taittirīya Upaniṣad verse (1.6) states:

In this space here within the heart lies the immortal and golden person

(puruṣa) consisting of the mind (manas). And this thing [the uvula] that

hangs like a nipple between the two palates, it is Indra's passage.

Bursting through the two halves of the skull at the point where the

hairs part, he establishes himself in the fire by making the call bhūr, in

the wind by making the call bhuvas, in the sun by making the call suvar,

and in Brahman by making the call mahas.

Here the text does not mention death, the sun, and the veins, yet many of

the other elements are present and consistent. The ascent of the Self (Puruṣa, in

this verse) is again associated with the secret knowledge and practices of mantras.

Less secure but suggestive, the larger contents of the Taittirīya Upaniṣad imply that

"Indra's passage" might be read as "the starting point or passage of the lord, om."

Thus the heart, the uvula, and the cranial suture connect via the passage that leads

the Self, the ātman, to liberation. A sage who knows the secret mantras may open

this passageway. However, those who do not possess the secret knowledge of the

hidden connections cannot escape.

History, Text, and Context of the Yoga Upanisads: Chapter 2

These early formative notions—the importance of the heart, special internal

pathways, and the use of mantras to activate hidden powers—provide the

physiology, physics, and metaphysics on which all of the later yoga traditions

build. Locating the cave of the heart as the locus of the prāṇas and ātman relates to

another set of important ideas, the four states of consciousness: waking

(vyutthāna), dreaming (svapna), deep sleep (nidrā, suṣupta), and the “fourth state”

(caturtha turya, turīya).35 The waking state is the normative mode of human

awareness: speech, thought, action, and ego-awareness. During waking

consciousness, the vital functions and breath spread throughout the body. Unlike

the experiences of waking, dreaming (svapna) is the time when the vital functions

retreat to the heart. The vital functions draw inward and rest in the heart. Speech,

action, and the other activities of waking occur but as an inner life, and in this

sense dream life is of the same quality as waking life. Deep sleep (suṣupta) is a

state of perfect rest in which the mind and senses are silent. The web of

manifestation (the evolved cosmos) entangles the individual when he experiences

35 The earliest Upaniṣads show evidence of only three states of consciousness. In general,

that unnamed state beyond the three is the liberation attained when the Self leaves the body at

death and escapes the material universe to the undifferentiated absolute Brahman. With the

development of certain yogic practices, realization of liberation becomes possible in life through

pure contemplation or deep meditation experiences. The oldest passages to consider this fourth

distinct state are the relatively late Māṇḍūkya Upaniṣad and Maitrāyaṇīya Upaniṣad 6.19 and 7.11.

See Deussen (2000), pp. 309-312. This also participates in the Vedāntin tendancy to expand the

older groups of three correspondences with taxonomies of four correspondences. One finds this

with the patterns of classification in general, such as in the case of the Atharvaveda being added to

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

waking and dreaming. The manifestations of the manifold universe disappear in

the bliss of deep sleep. This bliss is dependent on the identity of ātman with

Brahman, undistracted by the chatter of the senses. Generally speaking, a lack of

self-awareness characterizes the bliss of deep sleep. The "fourth state," caturtha

turya, turīya, is beyond the first three but combines the characteristic of self-

awareness with the bliss of deep sleep. Thus, in the fourth state the sage

experiences unity with Brahman but in a state of full awareness.

It appears that in the early Upaniṣads that the fourth state is death. This

identification of turīya with death changes over time as such concepts develop as

the person who is "liberated while alive" (jīvan mukti) or alive but in a liberated

state of pure contemplation (samādhi in classical yoga). Olivelle summarizes:

In sleep, the cognitive powers distributed throughout the body during

waking hours are gathered together in the cavity of the heart. . . .

During deep and dreamless sleep, the self slips out of that cardiac

space and enters the veins going from the heart to the pericardium;

there it remains oblivious to everything. At death the self, together

with the vital powers, departs from the heart along a channel and exits

through either the crown of the head or the eye.56

These concepts are unsystematic descriptive biophysics. They describe the

natural but secret structures and laws of liberation. The texts repeatedly indicate

that many of these functions operate automatically, such as in dreaming or at

the first three Vedas, or the om mantra composed of three phones, then later four, and in the minor

Upaniṣads even further divisions.

56 Olivelle (1996), p. li. See also Deussen (2000), pp. 296-312.

death. The Upanisadic sages' knowledge (jñāna) of the hidden connections

(upanisad) gives to them special powers. With this knowledge and empowered by

mantras, at death the sage may escape suffering and rebirth by following the path

to the sun, the realm of Brahman beyond the cosmos. What is mostly description

in these early texts becomes the basis for the later medieval traditions'

technologies of liberation. In other words, post-Vedic developments in asceticism

and yoga are active attempts to take control of such natural laws, structures, and

functions and to harness them for liberation.

The importance of mantras in the classical Upanisads likewise foreshadows

the developments in the later medieval traditions. The sages draw their

speculations on the mantras in the Upanisads directly from the oldest speculations

of the Vedas and Brāhmaṇas.57 The "bull," the "Indra" of Vedic mantras, is the

syllable "om." Similar to the other concepts previously discussed, most of the

early presentations of the om mantra are declarative and descriptive. Following the

logic of miniaturization and concentration of power (as in a seed), om is the sound

equivalent (alloform) of Brahman-ātman.58 The Praśna Upaniṣad (5.1-5) provides

the following formulation:

57 See Padoux (1992), pp. 1-29 for discussion of the importance of speculation on sound,

word, and mantras from the very beginning of the Brahmanical tradition.

58 A few of the multiple references to this mantra are BĀU 5.15.1; ChU 1.1.1-10; 1.4.1-5; 1.5.1-

5; 1.12.5; 2.23.3; 8.6.5; TaitU 1.1.1; 1.2.1; 1.8.1; 1.12.1; KathU 2.15; IU 17; ŚvetU 1.13-14; MunḍU 2.2.4,6;

PraśU 5.1-7; MaṇḍU 1-2, 8-12.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Then Śaibya Satyakāma asked him: "Lord, if some man were to

meditate on the syllable OM [AUM] until his death, what is the world

that he would win through that meditation?"

"If a man meditates on its first phoneme [A], he gets his knowledge

just from that; so he comes back to earth very quickly and is led to the

human world by the Ṛg verses. . . ."

"If, on the other hand, a man becomes mentally absorbed in the first

two phonemes (AU), he reaches the intermediate region and is led up

to the lunar world by the Yajus formulas. . . ."

"A man who meditates on the highest person by means of this very

syllable OM with all three of its phonemes (AUM), on the contrary,

enters the effulgence in the sun. He becomes released from evil . . ."

The Māṇḍūkya Upaniṣad deals almost exclusively with the mantra om. "OM—this

whole world is that syllable," the texts begins. It then goes on to identify om with

Brahman and ātman. Māṇḍūkya Upaniṣad also discusses the individual phones

(members of the phoneme om), equating them with waking, dreaming, and deep

sleep. Regarding the fourth state (turīya), it says, "The fourth, on the other hand,

is without constituent phonemes; beyond the reach or ordinary transaction; the

cessation of the visible world; auspicious; and unique."59

The psychophysiology and mantra recitations are employed in practice to

win liberation for the knowing sage of the Upaniṣads. The classical texts are

generally impressionistic and vague as to the details of actual meditation

practices. The following section examines the "preclassical yoga" of these texts.

59 MaU 12.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

2.5. Yoga in the classical Upaniṣads

The yoga of the classical texts is impressionistic; it is conveyed in the form of

hints and allusions. In this section, I will briefly outline the yoga found in the

classical texts which provides the foundation for the yoga traditions of the minor

Upaniṣads. The earliest of the Upaniṣads contain various elements that provide the

foundation for later developments.60 The Bṛhadāraṇyaka, Chāndogya, and Taittirīya

Upaniṣads contain many suggestive ideas that are cultivated in the later systems of

yoga. The Kaṭha Upaniṣad contains several of the fundamental principles and

concepts of yoga that differentiate it as a true preclassical yoga text. The

Śvetāśvatara and the Maitrāyaṇīya Upaniṣads employ yoga as a technical term and

describe meditation and other activities that demonstrate that these texts reflect

preclassical traditions of yoga. Both of these texts demonstrate early Śaiva

sectarian inclinations that are characteristic of many of the later yoga traditions of

Hinduism.

Bṛhadaraṇyaka Upaniṣad 1.5.23 previews the role of the control of the breath,

combined with a mentally repeated phrase, that will become the sine qua non of all

later yoga practice.

60 Numerous scholars have explored this issue. Cf. Deussen (2000), Eliade (1973), Varenne

(1971, 1976), Whicher (1998), Feuerstein (1998), passim. Most of the classical Upaniṣads have some

elements of interest to the student of yoga. The examples discussed herein are selected important

themes or practices that are developed by or explored in the later traditions.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Therefore, a man should undertake a single observance—he should breathe in and breathe out with the thought “May evil death not capture me.” And if someone undertakes it, let him resolve to pursue it to the end. By doing that he will win union with and the same world as this deity.61

Bṛhadāraṇyaka Upaniṣad 3.1.10 provides another example that demonstrates the already developed notions of different breaths in this earliest Upaniṣad.

Explaining the esoteric connection (upaniṣad) between the chants of the udgātr priest and the body (another use of the term ātman), the text states: “The hymn recited before the sacrifice is just the out-breath, the hymn that accompanies the sacrifice is the in-breath, and the hymn of praise is the inter-breath.”

The Chāndogya Upaniṣad begins with an elaborate discussion of the sacred mantra oṃ. Speculation about and practices centered upon recitation of this mantra (also called the praṇava, the “droner” or “humming”) continue throughout the development of the Vedānta traditions and throughout all later systems of Hindu yoga and tantra. These speculations are especially refined in the Yoga Upaniṣads texts with the “bindu” titles, but are also of interest to most of the northern recensions.62 Chāndogya Upaniṣad 8.15 hints at what will be later elaborated in the yoga traditions as the withdrawal of the senses, “pratyāhāra.” It states that drawing the sense organs into the Self combined with other behavioral observances leads to

61 Bṛhadāraṇyaka Upaniṣad 1.5.23. The deity referred to in the text is Vāyu, or rather prāṇa, the vital functions, by analogy.

72

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

escape from rebirth through attaining the world of Brahman. Taittirīya Upaniṣad

2.4.1 records the term yoga ātman, which the translators interpret as an early

technical usage similar to what was later systematized by Patañjali for the

conscious control of the sense faculties (indriya).

Bṛhadāraṇyaka Upaniṣad, Chāndogya Upaniṣad, and Taittirīya Upaniṣad

provide suggestive passages, yet the Kāṭha Upaniṣad (fifth century BCE) is the first

text to deal explicitly with yoga.63 Kāṭha Upaniṣad 1.2.12 employs a technical term,

adhyātman yoga (yoga of the inner self) for a practice whose goal is realization of the

god (deva) hidden in the “cave” (of the heart). The Kāṭha Upaniṣad combines the

Vedāntin notion of ātman with various elements later systematized in Sāṃkhya,

such as the seven levels in the hierarchy of existence (indriya, viṣaya, manas, buddhi,

mahātman/mahat, avyakta, prakṛti, puruṣa).64 Kāṭha Upaniṣad 2.3.11 provides a

definition of yoga as the “steady holding of the senses” that leads to attentiveness

(aparmatta), the condition of inner equilibrium.65

The Śvetāśvatara Upaniṣad (ca. third century BCE) promotes a yoga of dhyāna

(meditation) that leads to immortality. Sāṃkhya, yoga, and theism (of Rudra) all

62 See chapters 3 and 4, here for description and explanation of mantra yoga in the Yoga

Upaniṣads.

63 For discussion of the KāṭhU, see Feuerstein (1998), pp. 180-183; Whicher (1998), pp. 18-20.

64 See Whicher (1998), p. 19. Whicher suggests that the text is combining elements of

Vedānta with Sāṃkhya not as metaphysical speculation but instead as contemplative directives for

the process of yogic interiorization involving the expansion of consciousness to subtler levels of

identity. The goal state is for the yogin to identify himself with the transcendental self, puruṣa.

65 Whicher (1998), p. 19.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

influence this text. The text is clearly aware of practical yoga and dhyāna, and it

foreshadows several themes that later traditions will return to repeatedly.

Śvetāśvatara Upaniṣad 1.3 states: “Those who follow the discipline of meditation

have seen God, the Self, and the power, all hidden by their own qualities. One

alone is he who governs those causes, from “time” to “Self.” The metaphor of the

gander (hamsa) is used to symbolize the ātman (1.6 and 3.18). By twirling the om

mantra within the body through meditation, like a fire drill, the yogin will see God.

The Śvetāśvatara provides a detailed statement about the environment for

practicing yoga that is reminiscent of the later yoga manuals:

When he keeps his body straight, with the three sections erect, and

draws the senses together with the mind into his heart. . . .

Compressing his breaths in here and curbing his movements, a man

should exhale through one nostril when his breath is exhausted. A

wise man should keep his mind vigilantly under control. . . . Level and

clean; free from gravel, fire, and sand; near noiseless running waters

and the like; pleasing to the mind but not offensive to the eye;

provided with a cave or a nook sheltered from the wind—in such a

spot should one engage in yogic practice. . . . Equipped with the

attribute of yoga, that man, obtaining a body tempered by the fire of

yoga, will no longer experience sickness, old age, or suffer death.

Lightness, health, the absence of greed, a bright complexion, a pleasant

voice, a sweet smell, and very little faeces and urine—that, they say, is

the first working of yogic practice. . . . Once he has perceived the true

nature of the self, becomes solitary, his goal attained, and free from

sorrow. . . . [He] sees here the true nature of Brahman, he is freed from

all fetters, because he has known God, unborn, unchanging, and

unsullied by all objects.66

66 Śvetāśvatara Upaniṣad 1.8-15.

74

History, Text, and Context of the Yoga Upanisads: Chapter 2

Śvetāśvatara Upaniṣad 2.11 is especially instructive for the later Yoga

Upaniṣads. It describes visions experienced during yogic practice: mist, smoke,

wind, fireflies, lightning, crystal, and moon are all visions that a yogin might see

that prepare him for the full realization of Brahman. The Advayatāraka Upaniṣad

and the Maṇḍalabrāhmaṇa Upaniṣad contain similar information about visionary

experience in meditation. The importance of the heart as a center for focused

meditation and as the home of the Self as miniature "person" also appears

throughout the Śvetāśvatara. This Upaniṣad provides no explicit system of yoga but

includes so many details that it betrays a vast store of theory and practice. The

text reflects authors who belonged to a community that practiced a form of

preclassical yoga and who venerated Rudra (especially in his "glowing" [hara] and

"benign" [śiva] forms). The text introduces its sectarian emphasis (3.5): "That form

of yours, O Rudra, which is benign and not terrifying, which is not sinister-

looking — with that most auspicious form of yours, O Mountain-dweller, look

upon us."67

Yoga as a system is first delineated in Maitrāyaṇīya Upaniṣad 6.18 (ca. second

century B.C.E):

67 In especially Śvetāśvatara Upaniṣad 3 and 4, Vedic quotations abound that praise Rudra

but also recognize him via epithets and adjectives that later will become proper names (Śiva, Hara).

However, here the verses are not just songs of praise but enmeshed in the "secret connections" of

the genre. 4.15 states: "Who is finer than the finest, in the midst of disorder; who is the creator of

the universe displaying various forms; who, alone, encompasses the universe — when someone

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

This is the rule for achieving it (viz. concentration on the mind on the object of meditation): restraint of the breath (prāṇāyāma), restraint of the senses (pratyāhāra), meditation (dhyāna), fixed attention (dhāraṇā), investigation (tarka), absorption (samādhi), these are called the sixfold Yoga. When beholding by this Yoga, he beholds the gold-coloured maker, the lord, the person, Brahman, the cause, then the of sense, body, etc.) to be one in the Highest Indestructible (in the pratyagātman or Brahman).68

The Maitrāyaṇiya Upaniṣad in this and other passages refers to a systematic six-limbed system of yoga. With the exception of tarka (investigation or reflection), five of these limbs are the same as the classical formulation found in Patañjali’s eight-limbed system. Beginning with practices that anticipate yoga and ending with a six-limbed system, the classical Upaniṣads provide rich ground for the growth of the later medieval traditions. The Yoga Upaniṣads elaborate on many of these themes.

2.6. Classifications: “Classical” and “Minor” Upaniṣads

The extant texts of the Upaniṣads include hundreds of texts. Authors of this genre still compose texts of this class today. There are two important historical distinctions among the texts. The classical Upaniṣads are ancient Indian texts composed before the beginning of the common era: “principal” or “major”

recognizes him as the Benign One, he attains unending peace.” See Olivelle (1996), pp. 257-260 and notes. See also Deussen’ introduction to the text (1997), pp. 301-304, and throughout his notes.

68 Müller (1962), pp. 318-319.

76

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Upaniṣads are alternate terms used for these texts. The minor Upaniṣads—also

called Atharva or Atharvanic Upaniṣads—date from the common era. These texts

do not always possess close ties with authoritative Vedic schools or textual

traditions. As suggested by the name Atharva, many of these texts were attached

(after-the-fact) to the Atharvaveda.69 In most cases, the minor Upaniṣads are

medieval texts; many of them date from between the ninth and fifteenth centuries.

The classical Upaniṣads are the esoteric teachings of North Indian Vedism, whereas

the minor Upaniṣads are esoteric teachings drawn from the many traditions of

Brahmanical Hinduism.70

The term “classical” is both an emic and etic form of reference for these

texts. In the first meaning, these texts are classical because they were deemed

most important by the Indic commentarial traditions. These texts are interpreted

and cited by Bādarāyaṇa (second century CE) in his Brahma Sūtras, in Śaṅkara’s

(ninth century CE) commentary on the Brahma Sūtras, and in other Upaniṣad

commentaries attributed to Śaṅkara or his school. Thus, the classical texts were

those texts that Brahmanical communities considered the fundamental texts of

Vedānta, since no later than the ninth centuries. The designation “classical” is also

69 Contrary to Deussen’s assertions, there were long standing schools of the Atharvaveda.

Nine schools are mentioned in the Atharvaveda-Pariśiṣṭa, in Sāyaṇa, and in the Caraṇavyūha of the

White Yajurveda: Paippalāda, Sauda/Tauda, Mauda, Śaunakīya, Jājala, Jalada, Brahmavada,

Devadarśa, and Cāraṇavaidya. More to the point, the ties between these schools and most texts

classed as Atharva Upaniṣads is nebulous. See also T. Goudriaan & J. A. Schoterman (1994), pp. 4-5.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

a contemporary classification concerning the dating and historical context of the

texts. The accepted number of classical texts varies. The category contains as few

as eleven to as many as fourteen texts. The classical texts date from before the

common era, and were appended to various Brāhmaṇas and Āraṇyakas of the Vedic

schools.

The earliest texts are in prose with some verse (metrical quotes). These

include, in historical order, the Brhadāraṇyaka, Chāndogya, Taittirīya, Aitareya,

Kauṣitaki and Kena. These texts retain close ties with the Vedic śākhās (branches or

schools), forming a textual continuation of their Brāhmaṇas and Āraṇyakas. These

texts supply mystical, esoteric, and allegorical interpretations of the Vedic ritual

traditions.

The metrical Upaniṣads include in historical order: Kaṭha, Iśā, Śvetāśvatara,

and Muṇḍaka. Later prose Upaniṣads are the Praśna and Māṇḍūkya. In these latter

two categories, the connections with ritual traditions or Vedic śākhās is “sometimes

doubtful, sometimes artificial, and in any case is loose.”71 Deussen includes the

Mahanarayana as the latest of the early metrical texts and the Maitrayanīya in

between the Praśna and Māṇḍūkya. The Mahānārāyaṇa and Maitrāyaṇīya are

difficult to date. The extant recensions of the Mahānārāyaṇa contain medieval

70 Even the Sākta and Tantric Upaniṣads are cast in the mode of Brahmanical texts. On this

topic see T. Goudriaan & J. A. Schoterman (1994) and Brooks (1999).

71 Deussen (2000), p. 25.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

materials but also contain sections that might belong to the ancient period. The

Maitrāyaṇīya Upaniṣad is ancient, if slightly later than the other classical texts,

dating to no later than the second century of the common era.72

The classical Upaniṣads were composed as part of Vedic priestly schools, the

śākhās, especially those associated with the canons of the first three Vedas. Not all

of the Upaniṣads associated with the Atharvaveda developed in longstanding Vedic

schools. The ties between Upaniṣads and particular Vedic schools are often

arbitrary, except for the earliest texts, sometimes being derived from the

introductory blessings (śānti verses), and attributions can differ from one list to

another.73

Figure 2.1. Vedic Śakhā-Upaniṣad associations.74

Rgveda Kṛṣna Yajurveda Śukla Yajurveda Sāmaveda Atharvaveda

Aitareya Taittirīya Bṛhadāraṇyaka Chāndogya Muṇdaka

Kauṣītaki Śvetāśvatara Īśa Kena Praśna

Kaṭha Māṇḍūkya

In north Indian systems, the latest classical texts and all minor Upaniṣads are

linked with the Atharvaveda. Thus, some scholars refer to these texts as Atharva,

or Atharvanic Upaniṣads. Contemporary authorities on Indian literary history

72 [n my opinion, this text dates to the second century BCE, although it may date to as late

as the second century CE.

73 T. Goudriaan & J. A. Schoterman (1994), p. 1-2. This critierion is often unreliable.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

have overwhelmingly favored the north Indian tradition of Vedic attributions.75

Scholars generally argue that the Atharvaveda did not or could not protect its

literary heritage from the addition of new texts. Thus, over time the Atharvanic

corpus grew to sixty-eight established texts and dozens more that are not as well

attested.76

It is generally true that the relation between the Atharvanic Upaniṣads

and the Atharvaveda Saṃhitā is purely theoretical (except perhaps for

the relatively frequent mentioning of typically Atharvanic ṛṣis), with

the exception of the Cūlikā Upaniṣad, as noted by Bloomfield (1899,

19).77

The many Atharvanic Upaniṣads spread across India wherever Brahmins had

schools of thought, literate culture, and patronage. It is also the case that Upaniṣad

became a genre type, or even just a label associated with history and authority.

In South India, the original schools of Brahmanism that developed often

adopted and adapted older texts as well as authoring original works. In the late

medieval to early modern period (after 1300), southern redactors gathered and

edited collections of Upaniṣads. These collections did not have strong ties with the

old northern schools of Brahmanism, and often their compilers developed novel

categorizations for the classical and minor Upaniṣads.

74 See Olivelle (1996), p. xxxi. Simplified version of his chart included here as Figure 2.1.

75 T. Goudriaan & J. A. Schoterman (1994). See also Weber (1876), Deussen (1997 and 2000),

Winternitz (1908), and Shende (1952).

76 Goudriaan & Schoterman (1994), p. 4.

80

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Southern collections of Upaniṣads divide classical and minor texts among

each of the four Vedas.78 The Telugu characters edition (1883) of the Muktikā

Upaniṣad lists 108 Upaniṣads by name. It then proceeds to attribute these texts to

all four Vedas. It associates the classical and the medieval texts variously with the

Rgveda (10 texts including the Nādabindu Upaniṣad); Śukla(White) Yajur Veda (19,

including Triśikhibrahmaṇa, Mandalabrāhmaṇa, and Advayatāraka Upaniṣads); Kṛṣṇa

(Black) Yajurveda (32, including Amṛtabindu, Amṛtanāda, Tejasbindu, Dhyānabindu,

Brahmavidya, Yogatattva, Yogaśikha, Yogakundali, and Varāha Upaniṣads). The Cūlikā

Upaniṣad, under the title Mantrikā, is also associated with the Kṛṣṇa Yajurveda.

Sāmaveda (16, including Yogacūd̄āmaṇi); and Atharvaveda (31 texts, including

Śāṇḍilya, Pāśupatabrahmā, Mahāvākya Upaniṣads).79 These assignments to the

different Vedas appear to be arbitrary and, in Deussen’s assessment, “[are] without

any semblance of justifiability.”80 This list includes classical texts but in some

77 Goudriaan & Schoterman (1994), p. 1. See also Goudriaan & Schoterman discussion

(pages 1-5, and appendix III) especially the chart of Atharvanic Upaniṣad divisions numbered 1-8.

Bloomfield (1899), p. 21 § 47 gives a survey of AV quotations in several Upaniṣads.

78 In north India, late Upaniṣads have always been attached to the AV. This practice

changed among the south Indian Telugu Brahmins, who had weak relations or even no ties to the

traditional north Indian śākhās. Therefore, they developed their own traditions of association,

attributing minor texts to all four Vedas to fit their own systems of classification. Bouy even states

that the recensions were so different that the later compilers who added many quotations from

other yoga texts did not utilize the Atharvanic (northern) documents, but instead some southern

documents that transmitted a specific version distinct from the Atharvanic version: Bouy (1994), p.

79 The title abbreviations are all from the canon of Yoga Upaniṣads.

80 Deussen (1997), p. 558. Likewise, Tukārām Tātia’s 108 Upaniṣads (1895-96) Bombay

edition does not maintain the Telugu attributions. It follows the traditional attributions, placing all

minor texts with the AV.

History, Text, and Context of the Yoga Upanisads: Chapter 2

cases alters their traditional attributions (for example, it attributes the Íśa to the

Rgveda). This list attributes northern Atharvanic Upaniṣads to other Vedas. From

the examples above, the Amṛtanāda is linked to the Kṛṣṇa Yajurveda.

Scholars classify all Upaniṣads produced by the later post-Vedic

traditions—whether attributed to the Atharvaveda or not—as minor Upaniṣads.81

The minor Upaniṣads are artificially grouped as the Vedānta Upaniṣads,82 Sannyāsa

Upaniṣads,83 Yoga Upaniṣads,84 and sectarian texts, including the Viṣṇu Upaniṣads,

Śākta Upaniṣads, Śiva Upaniṣads, and a limited number of other sectarian

Upaniṣads.85 Adyar Library (AL) follows this scheme in their publications of texts.

There are also other schemes such as seen in the Kaula Upaniṣads edited by Sitārām

Shāstri.86 Many of these texts continue the themes of the earlier texts but did not

81 Deussen (1905) expanded the original classification from Weber (1887). The Adyar

Library adopted these classifications with the addition of the Śākta Upaniṣads in the early twentieth

century. Under the direction of F. Otto Schrader Adyar Library began to publish collections of

these texts in Sanskrit and in translation. Of the minor Upaniṣads, the Sannyāsa Upaniṣads are the

only collection critically edited. All other collections were published without critical apparatus

and reproduced the versions commented on by Upaniṣad Brahmayogin. Many of these texts have

multiple manuscripts differing in length, content, and even titles. The classification of texts as

Yoga, Vedanta, and the rest is a convention and therefore there are exceptions and variations as to

which title is attributed. For example Aiyar, in 1914, classified some of the Yoga

titles as Vedanta texts. See also Unpublished Upaniṣads, edited by the Pandits of the Adyar Library

under the supervision of Dr. C. Kunhan Raja, Adyar Theosophical Society, 1933.

82 English translation by A. G. Krishna Warrier (1991).

83 English translation by Olivelle (1992) from Schrader’s critically edited edition.

84 Yoga Upaniṣads have been translated partially or in full in English and French. See

chapter 1, 1.4 and chapter 3, 3.1, here, for discussion of the various translations.

85 English translation of Śākta Upaniṣads by A. G. Krishna Warrier (1999). In addition to the

Śiva, Śākta, and Viṣṇu texts, there are texts devoted to Skanda, Gaṇeśa, Sūrya, and even Allah. See

Feuerstein (1998), p. 171.

86 Sitārām Shāstri, Kaula and other Upaniṣads, published in the Tantrik Texts Series as Vol. XI

in 1922. Shāstri also addresses the relationship of minor Upaniṣads to the AV, by commenting on

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

attain their final recensions until the medieval period. Some contents of minor

Upaniṣads may reach backward to the latest of the classical texts. However, their

contemporary forms are medieval to modern.

The composition dates of the minor Upaniṣads are uncertain in virtually all

cases. The texts and manuscript traditions do not include any information on their

authors, dates, or geographic origins.⁸⁷ Many of the texts are not original

compositions but compilations. The authors of the texts drew from Brahmanical

sources, from Tantras and Āgamas, and from Nāth Siddha texts, and from other

sectarian and vernacular textual traditions.⁸⁸ The texts’ derivative nature

occasionally supplies data for dating and source analysis. In some cases, the

Upaniṣads quote from other datable texts. Date ranges of one to two centuries can

be suggested based on these types of citations. Some minor Upaniṣads may

contain materials from the early centuries CE. Many minor Upaniṣads date from

the thirteenth to fifteenth centuries.

Neither Bādarāyaṇa (second century) nor Śaṅkara (ninth century)

commented on or cited any of the Yoga Upaniṣads.⁸⁹ Therefore, it is likely that no

the unexplained “special affiliation” between the literature of Śākta Tantrism and the

Saubhāgyakhaṇḍa of the AV (unknown to the old Atharvan tradition). See Goudriaan & Schoterman

(1994), p. 2.

⁸⁷ Olivelle (1992), pp. 8-11. Olivelle includes a similar assessment of the problems with

dating the Samnyāsa Upaniṣads, although his generalizations are true for all of the minor Upaniṣads.

⁸⁸ See Bouy (1990a, 1990b, 1994, 1995); Goudriaan & Schoterman (1994).

⁸⁹ Bādarāyaṇa’s traditional dates are the fifth century BCE, but it is more likely that his

work dates from around the second century CE. Śaṅkara’s dates are still debated but most scholars

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

one considered them to be canonical—or that the texts did not exist—until after

the beginning of the ninth century.90 Eleven titles of the Yoga Upaniṣads are

attested by Nārāyaṇa (1500-1700) and ten titles appear in the Persian translation,

Oupanekhat or Oupnek’hat (commissioned 1656).91 All of the titles of the Yoga

Upaniṣads appear as part of the 108 Upaniṣads in the Muktikā Upaniṣad. The date of

the Muktikā is also in question, but some recension of it existed before the

fourteenth century (and probably after the ninth century). The Jīvanmukti Viveka

of Viyāraṇya (literary activity 1330-1385) may quote from the Muktikā, and it

clearly cites certain Yoga Upaniṣads (Amṛtabindu and Amṛtanādabindu). Therefore, it

is possible that the compiler of the Muktikā Upaniṣad wrote before the middle of

the fourteenth century. The widely available southern recensions of the Muktikā

Upaniṣad date to the middle of the eighteenth century, so it is not firmly

established that the list of 108 Upaniṣads appeared in the same form (if at all) in

believe that he lived between 722 and 820 CE Hajime Nakamura (1951) argues for earlier dates of

700-750 CE, pp. 63ff. On the dating of the Yoga Upaniṣads, see Bouy (1994), passim; Feuerstein

(1998), p. 415.

90 This is instructive, if frustrating. The tradition shows a long history of incorporating and

classifying texts slowly. The AV is not a Veda in the earliest textual references. Thus, the argument

that no minor Upaniṣads existed at the time of Śaṅkara because he did not comment on them is

circular and insecure. It does provide the basis for hypothesis but not proof. One must look to

other forms of supporting evidence to date the minor texts. Too often scholars have

conventionalized the judgement that lack of proof constitutes disproof—clearly an anti-intellectual

approach. Careful judgement on these issues likewise should not fall into the opposite

traditionalist trap of pushing the dates backward in time because they could be as old as the

insiders claim.

91 The Nārāyaṇa and Oupnek’hat lists are identical except for the absence of the Nādabindu

Upaniṣad from the latter. See below for summary of the four main collections of Upaniṣads.

84

History, Text, and Context of the Yoga Upanisads: Chapter 2

earlier recensions.92 In other words, the Muktikā Upanisad list of 108 specific

Upanisad titles is not attested before the first half of the eighteenth century.

The living Brahmanical traditions of India generally push the dates of all

Upanisads backward in time. This is a standard feature of the insiders'

constructions of legitimacy and authority. The tradition considers these texts to be

timeless, unauthored revelation. Many Western scholars have accepted

uncritically various dating schemes concerning these materials. As for the

classical Upanisads, Feuerstein places the earliest of these texts as composed

around the middle of the second millennium BCE.53 This dating scheme relies on

the revision of the earliest literature dates and a reevaluation of the relationship

between Vedic culture and the Indus Valley civilization.94 This revision of dates

does not withstand close scholarly scrutiny. Other historians of ancient Indian

civilizations, such as Romila Thapar and Ram Sharan Sharma, have conclusively

demonstrated within several of their works that the early dating of the Indian

92 See Bouy (1994), pp. 29ff. In several cases Bouy demonstrates that the southern

recensions of the minor Upanisads are late compilations that expand core north Indian Upanisads

with extensive quotations and insertions. Thus, although the Muktikā Upanisad list that includes all 21 Yoga Upanisads cannot be attested before 1751 (the

date that Upanisad Brahmayogin finished his commentaries on all 108 southern recensions).

However, it is also obvious that the texts must predate their commentaries.

53 Feuerstein (1997), pp. 315; (1998), p. 165ff. I have included references to these dating

schemes because Feuerstein writes extensively on the Yoga traditions of India. Although his work

is often of good quality on many aspects of yoga, he consistently presents ancient dates according

to traditionalist arguments. These arguments do not withstand scholarly criticism.

94 Feuerstein, Subhash Kak, and David Frawley (1995).

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

texts is not provable and is actually implausible. Olivelle has also argued for

later dates for most of the Indian textual traditions. He suggests that most of the

classical philosophical Sūtras, Upaniṣads, and dharma texts are from the second half

of the first millennium BCE through the first half of the first millennium of the

common era.

2.7. Collections of Upaniṣads

Both classical and minor Upaniṣads appear in various lists or collections.

There are thousands of edited and unedited Upaniṣad manuscripts in collections

around the world. Four primary listings of canonicity have established the list of

Upaniṣads, although in every case there are additional recognized Upaniṣads that

See R. S. Sharma (2001a, see also 1995 and 2001b). Although the topics are beyond the

scope of this work there are three powerful motivations that drive the historically nonviable early

dating schemes. First, the tradition itself equates sacred authority with antiquity and uses

descriptive terms such as "time immemorial" and "hoary antiquity" for any otherwise undated

text. Some aspects of the sociology and historiography of these characterizations are modern and

parallel the uses of history by modernist fundamentalist sects in the Middle East and the West.

There are two additional political motivations for the suggestions of unrealistically early dates by

some contemporary scholars. These are variations on two post-colonial theories. One of these is

the Indigenous Aryan theory that seeks to identify the Indus Valley civilization (or some other

early north Indian communities) as identical with the Vedic Aryans. Then borrowing dates from

Indus Valley archeologists, these scholars contend that the date of entire Vedic Corpora must be

pushed backward one thousand or more years. Another politically motivated contemporary

theory seeks to locate earliest Indian culture in the now dry Saraswati Valley instead of that of the

Indus. The simplest motivation of these arguments is that if they were true, then the locus of

Indian historical civilization would be in northern modern India, instead of along the Indus in

modern Pakistan. The totalities of these arguments are rooted in contemporary political

quandaries and have no intellectual value for understanding Indic history. Summarily, there are

some modern Indian scholars who neither want their traditions tied to (or originating from)

"foreign" Indo-Iranian nomads nor with any culture associated with the geography of modern

Pakistan.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

do not appear on the lists. The lists also contain titles that have no known

manuscripts.

As mentioned earlier, Śaṅkara’s commentaries established the canon of the

classical texts. The nature and history of the formation of the later canon is

obscure and can only be reconstructed through conjecture.96 Many of the minor

texts appear between the ninth and fifteenth centuries, some as late as the

eighteenth century. Upaniṣads also began to be recognized as a kind of text

independent from the extensive canons of Veda. Well known religious reformers

such as Śaṅkara (ninth century) and Rāmānuja (eleventh century) wrote

commentaries and added medieval devotional interpretations to the old texts,

more often praising and echoing the Upaniṣads than the other Vedic texts. It was

in this milieu that Upaniṣads were collected, studied, and many new texts were

written. After the fourteenth century, Upaniṣads were read separately from other

genres of Vedic texts, such as the Brāhmaṇas or Sūtras, and sometimes even lost

their traditional attestations. For example, as mentioned earlier, the classical text,

Iśa, was associated in South India with the Rgveda instead of its school of

origination, the Śukla Yajurveda. Upaniṣads that are always associated in North

India with the Atharvaveda, might be associated with other Vedas. During this

time, Brahmanical communities collected these texts together in lists of Upaniṣads.

96 Deussen (2000), p. 33.

History, Text, and Context of the Yoga Upanisads: Chapter 2

The existing collections of Upanisad titles are of two classes: complete lists

and lists that include only Upanisads of the Atharvaveda. The collections in the

Muktikā Upanisad and the Oupanekhat (Oupnek’hat) offer complete lists, although

they leave out some of the classical titles.⁹⁷ The collections of Nārāyana and Henry

Thomas Colebrooke include only Upanisads of the Atharvaveda (although these

include titles attached to other Vedas by the Muktikā Upanisad). All of the Yoga

Upanisads titles appear in the Muktikā Upanisad list, whereas only ten or eleven of

the titles appear in the other lists.⁹⁸

These lists of Upanisads also fall into another dual classification: lists with

forty-eight to fifty-two titles are usually associated with North India, and lists of

108 titles are associated with South India. Many of the North Indian collections

are associated with the brahmins of Benares, whereas the southern editions are

associated with Tamil Nadu and the Telugu brahmins. All of these lists differ

slightly as to order and text names, and some of them include sections or verses

from one known Upanisad within the body of another title. Not all of the known

texts from larger collections are included in the lists. Scholarly commentators

⁹⁷ Weber used the Oupanekhat and Sanskrit manuscripts to determine his list of canonical

Upanisads. He offered these for the first time to Western readers in Indische Studien Vol. I, II, and

IX. See also Deussen (1997), p. 559.

⁹⁸ See below for the chart of Yoga Upanisads titles and where they appear in the collections.

ATU, DŪ, MBU, MVU, PBU, ŚU, TŚBU, VŪ, YCU, YKU only appear in the Muktikā Upanisad list.

History, Text, and Context of the Yoga Upanisads: Chapter 2

consider the title order in each collection to be arbitrary, yet the ordering of texts

predates the manuscripts in which they first appear.

The Oupanekhat (Oupnek'hat) is a Persian translation of a collection of fifty

Upanisads. Sultan Muhammed Dara Schakoh commissioned these translations in

1656 in Delhi. The Oupanekhat is a North Indian collection; pandits from Benares

provided the manuscripts for this translation of Upanisads. Like the Muktikā

Upanisad, this collection includes texts from both the classical and minor titles and

therefore includes texts from all four Vedas. The Oupanekhat includes translations

of ten of the titles designated as Yoga Upanisads.

The Muktikā Upanisad includes a list of 108 Upanisads. As discussed

earlier, some scholars argue that the Muktikā Upanisad dates to as early as the

fourteenth century, whereas others argue that there is no definitive evidence for

the text before the first half of the eighteenth century. This South Indian text offers

original associations between the texts and the four Vedas. The Muktikā Upanisad

includes all of the generally designated Yoga Upanisads titles. Although the text

99 Anquetil Duperron translated the Oupanekhat into Latin (two volumes, Agentorati 1801-

1802). This literal Latin translation from a rather free Persian translation of the original Sanskrit

was the first available translation of Upanisads in Europe, and was likely the edition that was of

interest to such thinkers as Schopenhauer. For discussion and general introduction to this

collection, see Deussen (1997), pp. 558-561.

100 The dates of the Muktikā Upanisad are not sure. The 108 Upanisads were first made

available in a poorly edited printed edition, Sanskrit in Telugu characters, in 1883. For discussion

of the Muktikā Upanisad see Deussen (1997), pp. 556-558. For Sanskrit with English translation of

the Muktikā Upanisad see Aiyar (1997), pp. 1-9. For general comments, see also Bouy (1994), p. 27,

passim.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

attributions are not generally valid, they do demonstrate that the southern Telugu

Brahmins had their own tradition, independent from the northern Vedic schools.101

The Muktikā Upaniṣad (verse 44) suggests that it drew its list of 108 from an even

larger body of material. Even if the text dates to as early as the fourteenth century,

no one has demonstrated through manuscript evidence that the recension of the

Muktikā Upaniṣad that presents 108 titles is the same recension as the earliest

versions of this text.102 In addition, the list includes titles only, not quotations from

the minor Upaniṣads. Therefore, it establishes that the titles of the twenty-one Yoga

Upaniṣads were known by 1751. How long before this date the previously

unattested ten texts (of the total of twenty-one Yoga Upaniṣads) were written is still

open to conjecture.103

Nārāyaṇa (also called Nārāyaṇabhaṭṭa), the son of Bhaṭṭa Ratnākara, was a

commentator on classical and minor Upaniṣads who refers to an already

101 Bouy (1994), pp. 32, passim; Deussen (1997), pp. 556-558.

102 A conclusive statement is not yet possible. Bouy (1994), does demonstrate the late dates

of YCU, YKU, and the MBU and establishes that the southern recensions of the older Yoga

Upaniṣads were revised and expanded after the fifteenth century. Based on his arguments it seems

likely that none of the additional ten titles is of great antiquity. Although they do not present

systematic evidence, earlier scholars (early twentieth century) such as Deussen ignored these

additional texts as of “doubtful origin” (meaning by this, late medieval to early modern in

composition).

103 Bouy (1994) comments on some of these titles. There is no evidence for these text before

the first half of the eighteenth century. Anything more definitive will have to wait until there is

more manuscript research focused on Tamil Nadu and other southern Upaniṣad manuscript

collections.

90

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

established collection (list) of texts.104 Nārāyaṇa's readings generally agree with

those found in the northern recensions. Similarities to the other collections and to

the readings suggest that Nārāyaṇa was a North Indian commentator who wrote

between 1500 and 1700. He does not list the texts from this collection, but there is

enough evidence to reconstruct the list from his work.105 Thus, as with the other

collections, he appears to be working from an assumption of canonicity and the

list he comments on was not his creation. Nārāyaṇa cites both Śaṅkara (ninth

century) and Śaṅkarānanda (ca. 1300), the disciple of Ānandātman, and therefore

wrote his commentaries (Dipikās) after their times. Although there are no exact

dates for Nārāyaṇa's writings, Deussen concludes that he was dependent on the

same canonical list and similar manuscripts to those of Colebrooke.106 Nārāyaṇa's

collection includes eleven Yoga Upaniṣads.107

104 See Deussen (1997), pp. 562-565, for a general introduction. Several of Nārāyaṇabhaṭṭa's

commentaries (Dīpikās) were published along with their Upaniṣads by Rāmamaya Tarkaratna in the

Bibliotheca Indica (Calcutta).

105 Unlike the Muktikā Upaniṣad and the Oupnekhat, Narayaṇa does not provide a list of

titles. The list of 52 titles associated with his name was a reconstruction in 1897 by Deussen

(Deussen (1997), pp. 562). Lacking a definitive list, there is some minor confusion over the

numbering of the texts.

106 Numbers 1-34 agree with Colebrooke's 1-40 (with six omissions) in names and serial

order. Numbers 35-45 agree with Colebrooke's 41-52 in names and serial order. Afterward,

Nārāyaṇa further enumerates a series of titles concerned with Krṣṇa and Gaṇeśa that do not

appear in Colebrooke's collection. See Deussen (1997), p. 564.

107 See Bouy (1994), p. 30. For exhaustive study see P. K. Gode, (1938).

91

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Henry Thomas Colebrooke (nineteenth century) gathered his list of fifty-

two Upaniṣads from North Indian sources.108 According to Deussen, Colebrooke

and Weber both considered this list to represent a canonical list of Atharvaveda

(minor) Upaniṣads that were widely disseminated and generally recognized.

Although shorter than the southern collection, the list is canonical in that

Colebrooke (with Aufrecht, Weber, and Deussen) considered titles widely

disseminated if they appear on all lists. Therefore, the shorter northern list

becomes the de facto measure of canonicity, since it includes fewer titles and the

southern lists contain all of these northern titles. Generally, the southern texts

include significantly large amounts of interpolated medieval materials. Because of

this, scholars argue that these long texts were expansions of the shorter ones.109

Eleven Yoga Upaniṣads texts appear in Colebrooke’s list (identical to Nārāyaṇa’s).

108 For general summary, see Deussen (1997), pp. 561-562. I have not been able to establish

the origin and location of all of Colebrooke’s texts. Deussen includes several references to

Colebrooke obtaining manuscripts in Banares.

109 Bouy (1994), passim.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Figure 2.2. Yoga Upaniṣads Crossed Referenced by Upaniṣad Collection List

Based on conclusions and implications discussed throughout Deussen's

works, there are good reasons to view the short list of minor Upaniṣads as the

definitive list of broadly accepted canonicity. The majority of scholars believe that

110 The numbers designate placement in each list. Deussen and Weber did not consider

these numbering schemes to have any relevance for dating. See Bouy (1990a, 1994) concerning the

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

many of the minor Upaniṣads were originally composed in North India.

Brahmanical traditions carried the texts south, to which there were added several

southern texts. It is not certain that the extant northern recensions are the same

texts that northern Brahmanical traditions carried to South Indian. The northern

recensions collected in the Ānandāśrama Sanskrit Series and the Bibliotheca Indica

differ significantly from the southern Adyar Library Series in more ways than just

the insertion of extra materials.111

There was a southern Telugu Brahmin tradition of Vedānta and there were

multiple northern Vedānta traditions (Poona, Calcutta, Bombay, but especially

Benares). The Telugu tradition significantly expanded several of the texts to

include various additional materials inserted into the older texts and several new

titles. In the particular case of the Yoga Upaniṣads, the compilers drew these

materials from the traditions and texts of haṭha yoga and tantric kuṇḍalini yoga.112

Caraṇavyūha lists that Weber and others used to compile the contemporary collections.

111 ĀSS was published in the late nineteenth century, Poona, largely from south Indian

manuscripts. BI was published in the middle to late nineteenth century in Calcutta. The ALS was

published in Madras in the early twentieth century. For manuscript variations see Bouy (1994),

passim.

112 Bouy (1994), pp. 29-49; 68-72; 82-110.

94

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

2.8. Dating: Analytical Confusion Regarding the minor Upaniṣads

Scholars have made various claims regarding the dates of the minor

Upaniṣads. Dating assumptions influence conclusions about the development and

history of textual traditions and intellectual movements. Deussen's work still

asserts considerable influence over scholars' judgements concerning the dating,

arrangement, philosophy, and social context of the Upaniṣads. His work is of high

quality in most ways, yet it is a century old. Thus, referring to his definitive work

without some caution might cause the reader to fail in grasping the nature of

Upanisadic scholarship since his time. For the purposes of this study, it is not

necessary to provide an exhaustive analysis of Deussen's work. However, some

comments regarding his ideas about the dating of the texts is helpful for the

arguments of this work.

Deussen consulted the southern (Telugu, 1883) recensions of the Yoga

Upaniṣads but relied on the shorter northern manuscripts to make his translations.

He did not provide definitive discussion of his sources, but from reading his notes

and translations, it is clear that he worked primarily from the Bibliotheca Indica

texts. It is also clear that he consulted several other texts in the case of corrupt or

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

confused readings.113 He does not suggest dating for all of the texts but provides

the following comments.

Deussen states that Sāṃkhya of the Cūlikā Upaniṣad is "of an older form as

appears, e.g. in the Maitr 5," not the classical form of the Sāṃkhya Kārikā and the

Sūtras.114 However, he does not argue that the Cūlikā Upaniṣad actually predates

the Kārikā. Deussen comments that the Nādabindu Upaniṣad appeals to Atharvaveda

13.3.14 and several of the classical Upaniṣads, but he makes no guess as to its age.

He considers the Amṛtabindu (also titled Brahmabindu) to be quite old: arguing that

Śaṅkara probably quoted an earlier recension of the text. Deussen considers the

quotes common with the Maitrāyaṇīya Upaniṣad to be woven more naturally into

the Amṛtabindu. He states: "It is therefore quite probable that the Brahmabindu

Upaniṣad [= Amṛtabindu] is cited in the Maitrāyaṇīya Upaniṣad."115 This dating

estimate (like his others) relies exclusively on the content of the texts. Deussen

does not present his estimates of dates as incontrovertible; in all cases, he only

presents the estimates with terms such as "probable." Bouy's (1994) work

suggests that the Amṛtabindu (or Brahmabindu) dates from no earlier than the

twelfth to thirteenth centuries, and his estimate relies on much more extensive

113 For the minor Upaniṣads, he particularly utilized Ātharvaṇa Upaniṣads with the

commentary of Nārāyaṇa, ed. by Rāmanaya Tarkaratna, 5 fasc., Calcutta, 1872-74 (Bibliotheca Indica,

76).

114 Deussen (1997), p. 676. Italics added.

115 Deussen (1997), p. 687. Italics added.

96

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

research.116 He states that the Amṛtanāda was cited by Sāyaṇa (fourteenth century)

and Śaṃkarānanda (ca. 1300), establishing its date as before these

commentators.117 Not one of these statements is conclusive, and Deussen makes

no further hypotheses about dating the Yoga Upaniṣads.

Mircea Eliade suggests that the Cūlikā Upaniṣad was probably written at the

same period as the Maitrāyaṇīya Upaniṣad. This would make this Yoga Upaniṣads

very early, dating from no later than the second century CE. Eliade cites Deussen

regarding this date.118 Although Deussen does associate the contents of the Cūlikā

with those of the Maitrāyaṇīya Upaniṣad, he does not suggest that they date from

the same period. Eliade places the remainder of the Yoga Upaniṣads and Sāṃnyāsa

Upaniṣads as contemporaneous with the didactic portions of the Mahābhārata.119

This analysis does not provide exact dating, but Eliade then proceeds to suggest

that the Yoga Upaniṣads date from just before the Vedānta Sūtras and the Yoga

Sūtras. He does not cite the titles or specific Sūtras to which he is referring.120

Eliade is not completely clear as to how he dates the materials, but the general

116 Bouy (1994), p. 48, 118, passim. Some materials in the ABU may predate the twelfth

century but it is not attested with certainty until that time.

117 Deussen (1997), p. 691.

118 It is unclear to whom Eliade attributes this date because this citation includes Deussen’s

Sechzig Upanishad’s des Vedas, p. 637; E. W. Hopkin (text not named but probably) The Great Epic of

India, pp. 100, 110; and Hauer’s Der Yoga als Heilweg, p. 34.

119 Eliade (1973), pp. 127-135. The didactic portions of the MBh likely date from between

the third and fifth centuries CE

120 Eliade’s reference is not clear. He may be referring to the Sūtras of Bādarāyaṇa and

Patañjali whose dates are likewise insecure but likely date from the second (to fifth) centuries CE If

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

trend of his arguments is that the Saṃnyāsa Upaniṣads and the Yoga Upaniṣads were

written during the first five centuries of the common era.121 This dating is

conceivable but unlikely. There are too many medieval characteristics to the texts

to make these early dates probable.

In general, Eliade relies heavily on Deussen’s dating estimates, but he also

relies on the texts of the Adyar Library Series of both the Saṃnyāsa Upaniṣads and

Yoga Upaniṣads. This approach allows a glaring error to enter the calculations.

Deussen does suggest relatively early dates for several of the Yoga Upaniṣads, but

his northern manuscripts are of ten to twenty Sanskrit verses, whereas Eliade

comments on the long southern versions of the texts (over 100 verses). Deussen

engages in a different type of discussion as to these early dates. He worked from

multiple manuscripts and employed a form of textual archaeology by attempting

to isolate the root texts within compilation texts. Deussen clearly envisions the

long versions of many of the texts as late medieval to modern, while maintaining

the possibility that the short manuscripts might represent early medieval Ur-texts

of the later compilations.122

this is correct, Eliade’s dates for the Saṃnyāsa Upaniṣads and Yoga Upaniṣads signal a span from the

first to fifth centuries CE

121 Eliade divides the Yoga Upaniṣads into two groups. He states that the first group date as

described above and that the second group was written later. In the first group, he places the ABU,

ANU, BVU, DhBU, K.U, NBU, TBU, YSU, YTU. With the exception of these nine titles, all other

designated Yoga Upaniṣads fit into the later grouping.

122 As stated throughout, Bouy (1994) provides substantial support for this hypothesis.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

While it is possible that the short versions of the texts are older texts, the

longer versions clearly date from the medieval period, as they quote from other

medieval texts. Because he does not understand Deussen’s analysis, Eliade draws

some mistaken conclusions concerning the historical developments of certain yoga

materials. Having examined the contents of the long version of the Yogatattva

Upaniṣad, Eliade concludes that the lists of āsanas and mudrās recur in the

Haṭhayogapradīpikā. He also suggests that various elements of the Yogatattva

Upaniṣad are those “which tantrism will develop to the utmost.”123 Bouy

demonstrates that many of the southern recension Yoga Upaniṣads quote the

Haṭhayogapradīpikā (as well as numerous other works on yoga), not the reverse.124

Bouy does not address the Yogatattva Upaniṣad specifically, but the northern

recensions of the text do not include haṭha yoga materials, which suggests the

southern Yogatattva Upaniṣad has a similar history to the other Telugu expanded

texts. Available evidence suggests that the Yoga Upaniṣads borrowed most of its

yoga from the Tantras and Nāth Siddha texts.

123 Eliade (1973), p. 132, italics added. Eliade’s overall discussion of the YTU primarily

focuses on characteristics that demonstrate its content as heavily tantric and full of haṭha yoga.

124 Bouy (1994), pp. 80ff, passim. Bouy does not generalize about the Yoga Upaniṣads,

instead he examines specific titles from the minor Upaniṣads. He examines both northern and

southern recensions. In every case the southern recensions are late (post-fifteenth century) and

they quote extensively from Nāth and other yogic literatures. Bouy does not analyze the YTU in

his work on the Nātha Siddha influence in the Yoga Upaniṣads. Although he does not comment on

textual influences, he does provide recension analysis on the YTU elsewhere (1990a, p. 97ff).

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Eliade makes the historical argument that the yogas of the tāntrikas and

Nāths were directly or indirectly derivative of the traditions of the Yoga Upaniṣads.

This assertion is not credible.125 Even if Deussen’s early dates are accurate for the

northern recensions of the texts, these short versions do not contain significant

references to the haṭha yoga materials that Eliade finds so instructive. The northern

short versions focus upon mantra yoga, especially speculations on om recitation.

They do not include haṭha yoga or significant tantric materials. Both northern and

sourthern versions will be discussed throughout the remaining chapters of this

study.

Guy Beck draws conclusions similar to those of Eliade in his discussion of

the Yoga Upaniṣads. Beck begins his discussion with a confused reading of Jean

Varenne. For his study, Beck employed the southern recensions of the texts, in

both Sanskrit and English translation. Like Eliade and Varenne, Beck consulted

the late Telugu tradition of the expanded Yoga Upaniṣads texts. He did not consult

the earlier northern short texts. Beck begins his discussion with the following

observation:

The twenty-one available Yoga-Upaniṣads, though often understood to

be of late origin, are attested to reflect practices of a much earlier

period: [Quoting Varenne] “Although the first Tantras and Yoga-

Upaniṣads date from the 8th century after Christ, we may and should

assume as a certainty that their contents were formulated at least ten

centuries earlier, then continuously elaborated upon and renewed

125 The entire study of Bouy (1994) is devoted to this issue.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

before being finally fixed in the texts we possess." This statement is

more realistic than the estimation of Eliade, who says that these texts

were contemporary with the didactic portions of the Mahābhārata and

probably very little earlier than the Vedānta-Sūtras and the Yoga-

Sūtras.126

Beck concludes that the Yoga Upaniṣads contents date from the last centuries BCE

and that the dates of Eliade are too late. As discussed earlier it is probable that

even Eliade's dates are too early and accordingly Beck's estimation of the dates of

the texts even further misses the mark.

Furthermore, Beck misunderstands the context of Varenne's remarks.

Varenne does not clarify his statement as to what details of the Yoga Upaniṣads and

Tantras date from the last two centuries BCE. Varenne follows this statement with

a presentation of general elements of Hindu thought that were in long evolution in

the schools of tantra and yoga.127 Varenne concludes these general comments with

the statement that

[The teachings of the modern masters of yoga] are always strongly

influenced by Tantrism, even when they claim to have derived them

from other currents of classical Hinduism (usually the Vedanta School,

126 Beck (1995), p. 92. Quoting Varenne (1976), p. 182, and Eliade (1958), p. 127. In listing

the Yoga Upaniṣads, Beck mistakenly includes the titles Brahmabindu, Amṛtabindu, and Amṛtanāda

(BBU, ANU, and ABU) as three separate texts. These three titles represent only two different texts.

See chapter 3, here for discussion of these (and other) title changes. As is typical of authors who

consult only the southern tradition, Beck does not include the Cūlikā (CLU) as a Yoga Upaniṣads.

127 Varenne is vague on these points. If he is envisioning six or eight limbed systems of

yoga, the importance of breath, basic meditation techniques, and some rudimentary notion of

breaths moving through body pathways (a proto-subtle body) then he would be well supported in

this notion.

History, Text, and Context of the Yoga Upanisads: Chapter 2

which over the past two or three centuries has become the "accepted

Philosophy" of Indian intellectuals).128

Varenne actually avoids providing conclusive statements as to the dating of

the Yoga Upanisads and, unlike Eliade, does not argue that the Yoga Upanisads

influenced the yoga of the Nāths or tāntrikas.129 Although he is not definitive in his

statements, Varenne appears to see the direction of influence as coming from tantra

into the Yoga Upanisads texts.

In Beck's defense, his concerns are not primarily regarding the dating of the

texts, but he is concerned rather with the importance of mantra yoga and nāda yoga

within the corpus of Yoga Upanisads. Unfortunately, he does not focus solely on

these aspects of sound in the texts. Instead, he draws historical conclusions about

them. In discussing the Tantras and Nāth Siddhas, Beck states that the

Kaulajñānanirnaya describes a practice of yoga that appears to be a further

development of that contained in the Yoga Upanisads. These practices concern the

interior sounds heard during meditative trance (KJN 14.85-86). Beck concludes:

"These five sounds appear to reduplicated from the 'earlier' ten or eleven sounds

128 Varenne (1976), p. 183. In this statement, Varenne is talking about modern practitioners,

not directly about the Yoga Upanisads.

129 Regarding the dating and character of the texts, Varenne states: "...hérétique par

essence, il se devait, s'il voulait s'intégrer dans le brahmanisme, forger au plus vite un ensemble de

textes soi-disant védiques (car toutes les Upanisad prétendent se rattacher au Veda) qui

canstitueraient ses garants traditionnels. Ainsi furent composées, à une date absolument

indéterminable, plus de vingt Upanisads relevant du Yoga. Varenne (1973), p. 17.

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

found in the Yoga-Upaniṣads.¹³⁰ The Kaulajñānanirnaya predates the southern

recensions of the Yoga Upaniṣads by several centuries. Although there are

Upanisadic models for sound manifestations, the sounds of the Kaulajñānanirnaya

predate the sounds expounded in the Yoga Upaniṣads.¹³¹

Further details that support the late dating of all minor Upaniṣads will be

discussed throughout chapter 3. The preceding cases provide a few examples of

confusion regarding the dating and historical importance of the Yoga Upaniṣads. I

have provided these examples to demonstrate that it is not only a point of

historical correctness (description) that is at stake. Misunderstanding the dates of

the texts leads to errors in analysis and in explanation of the historical

development of sectarian movements, texts, and intellectual complexes. In

addition, these kinds of arguments lead to a perpetuation of the claims of the

"accepted philosophy" of modern Indian intellectuals (Vedānta) as described by

Varenne. One primary theme of the arguments explored throughout the

remaining chapters of this study is how it was that the Tantras, and later the haṭha

yoga texts of the Nāths, served to shape the practices described in the Yoga

¹³⁰ Beck (1995), p. 99.

¹³¹ The Hamsa Upaniṣad (HU) does have a northern recension that includes meditation

"sounds," that dates form the thirteenth century or shortly afterward. The KJN dates from the

ninth to tenth centuries. Thus, even the earliest attested HU recension does not allow for Beck's

assessment. See White (1996), p.73, and throughout for dating comments on the KJN (or KjnN) and

its tradition. There are several Tantras (Śārada-Tilaka and others) that describe and comment on

sounds heard during meditation. See Padoux (1992), throughout for "sound" meditation

manifestations and the Tantras.

103

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

Upaniṣads. Certainly there is a large body of shared meditation and yoga practices

among all the traditions of India over the last two millennia. However, sharing

always represents networks of interconnection over time: that is, these traditions

have histories. Even though these histories are not fully recoverable the Yoga

Upaniṣads demonstrate a parentage of at least two traditions: the classical

Upaniṣads and the Tantras. Scholars have often praised the classical parent while

ignoring the tantric one.

2.9. Concluding Remarks

The Yoga Upaniṣads build on all of the concepts, elements, and details of the

classical traditions of the Upaniṣads and Vedānta. Although separated by centuries

and recipients of additional influences—Patañjali and his commentators, Tantras,

Āgamas, Nāth Siddhas, and others—the authors of the Yoga Upaniṣads wanted

more than anything else for their texts be viewed as heirs to the classical traditions

and texts of Vedānta. In order to do this the authors of the Yoga Upaniṣads drew on

many of the concepts discussed in this chapter: ātman, Brahman, manas, prāṇa, the

heart and its pathways, the importance of mantras (especially om), as well as other

Upanisadic concepts, metaphors, and views. However, the Yoga Upaniṣads were

not developed within the old Vedic śākhās (schools), although their character is

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

overwhelmingly Brahmanical. They were born out of the milieu of the Tantras and

the Purāṇas; as such, they are truly syncretistic Hindu documents.

Collections of minor Upaniṣads (including some Yoga Upaniṣads) began

appearing from the ninth to eleventh centuries. They were widely shared among

several Brahmanical Hindu traditions by the fifteenth century. One may conclude,

on the basis of the translation of the Benares tradition into the Persian Oupanekhat

(1653), that Upanisadic mantra yoga was the subject of ten of the Yoga Upaniṣads. In

South India, the Upaniṣad tradition of the Telugu Brahmins developed and

maintained its own traditions of the texts and expanded them greatly with

materials from the Nāth Siddhas and other South Indian yoga traditions.

Sometime before the middle of the eighteenth century, the southern tradition

added several texts, raising the number of Yoga Upaniṣads to twenty-one texts. By

the late nineteenth century manuscripts were collected and printed in many of the

great urban centers of India. The northern traditions recognized a canon of eleven

Upaniṣads while the texts of the southern traditions totaled twenty-one. With

modern printing, all of India (and even the world) recognized a canon of twenty or

twenty-one texts called the Yoga Upaniṣads that contained the yoga traditions of

classical Upanisadic mantra yoga, the late South Indian Nāth-Vedānta synthesis, as

well as materials drawn from the Āgamas and Tantras. Building on the foundation

of the classical Upaniṣads, discussed in this chapter, the remaining chapters of this

History, Text, and Context of the Yoga Upaniṣads: Chapter 2

study will trace and examine the other materials and influences that shaped the

texts in the so-called canon of Yoga Upaniṣads.

106

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Chapter 3

The Yoga Upaniṣads: Textual Synopses

The colors of cows are many

Yet, the color of milk is one

Let the essence be seen as knowledge

Like the milk of the cows is seen

Just as butter is hidden in milk

Discernment dwells inside every being

Constantly twirling them in the mind

By means of fire-drill of being

Joined with the drill-string of knowledge

Mind should carry up to the highest

Like fire is drawn out from the drill

"I am that Brahman, undivided, still, and peaceful."

Thus, it is remembered

—Amṛtabindu Upaniṣad 1.19-twenty-one1

3.1. General Introduction to the Canons of Yoga Upaniṣads

The texts of the Yoga Upaniṣads draw their ideas and images from the

heritage of classical Upanisadic Vedānta. They are medieval texts, but their

imagery plays with the metaphors, symbols, and objects of Vedic fire rites,

Vedānta of the classical Upaniṣads, the Yoga Darśana, Puranic mythology, Āgamas

and Tantras, and the "forceful yoga" of the Nāth Siddhas, as well as other medieval

1gavāmanekavarṇānāṁ kṣirasyaāpyekavarṇatā | kṣīravat paśyate jñānaṁ linginastu gavāṁ yathā

1 119 ghrtamiva payasi nigūdham bhūte ca vasati vijñānam | satatāṁ manasī manthayitavyāṁ

manomanthānabhūtena 1 120 jñānetraṁ samādhāya coddhrretovahnivat param | niṣkalam niścalam

śāntam tad brahmāham iti smṛtam | |twenty-one. These Amṛtabindu verses come from the Adyar

Library Series recension (southern).

History, Text, and Context of the Yoga Upanisads: Chapter 3

yoga textual traditions. Their goals are moksa and Brahman. The practices of their

northern tradition are mantra yoga. Later, in South India they draw much from the

practices and ideas of tantra and hatha yoga. Meditation on the most sacred

Brahmanical mantra om is the cornerstone of the yoga of these texts, while the

hamsa mantra and its lore complete this yoga's foundation.

The texts gathered together as the Yoga Upanisads include the following

twenty-one titles:2 (1) Advayatāraka, (2) Amṛtabindu/Brah mabindu, (3) Amṛtanāda/

Amṛtanādabindu, (4) Brahmavidyā, (21) Cūlikā, (5) Darśana/ Yogadarśana,

(6) Dhyānabindu, (7) Ham sa / Hamsanāda, (8) Kṣurikā, (9) Mahāvākya, (10)

Maṇḍalabrāhmaṇa, (11) Nādabindu, (12) Pāśupatabrahm a, (13) Śāṇḍilya, (14) Tejobindu,

(15) Triśikhibrahm aṇa, (16) Varāha, (17) Yogacūḍāmaṇi, (18) Yogakuṇḍali, (19)

Yogaśikhā, and (20) Yogatattva.3 Although it is not included in the definitive Adyar

Library Series edition of the Yoga Upanisads , Deussen includes the Cūlikā Upaniṣad

as a yoga text (called the Māntrika Upaniṣad in the southern collection of 108

Upaniṣads).4 Because of Deussen's inclusion of the text in his list of yoga texts and

because of its content, I include a summary of the Cūlikā below (# 21).

The lists of Yoga Upaniṣads generally include more than twenty-one texts,

but this is due to confusion over the names of some of the texts. It is common for

2 Numbered to correspond with section numbers below. (1) = 3.2.1, (2) = 3.2.2, and so on.

3 Text abbreviations: ATU, ABU, ANU, BVU, CU, DU, DhBU, HU, KU, MVU, MBU, NBU,

PBU, ŚU, TBU, TSBU, VU, YCU, YKU, YŚU, and YTU.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

different Sanskrit texts to share the same names or for one Sanskrit work to

possess multiple titles for its various manuscripts and recensions. Throughout

this study, I have dealt with the titles in the following way. The four titles

Brahmabindu Upaniṣad, Amṛtabindu Upaniṣad, Amṛtanāda Upaniṣad, and

Amṛtanādabindu Upaniṣad refer to only two different texts. Brahmabindu Upaniṣad is

the original title for a text of twenty-two ślokas. A different text of thirty-eight

ślokas was originally titled as the Amṛtabindu Upaniṣad. Deussen suggests that the

more well attested name for the text of twenty-two ślokas was Brahmabindu, yet by

the fourteenth century this title was dropped due to some oversight. The title

Brahmabindu was replaced with Amṛtabindu, a title that also belonged to

Amṛtabindu Upaniṣad of thirty-eight ślokas. Due to this shifting of names, it became

necessary to choose a new name for the text of thirty-eight ślokas. The text that

was originally called Amṛtabindu was then called the Amṛtanāda Upaniṣad, a title

used by both Śaṅkarānanda (ca. 1300) and Sāyaṇa (fourteenth century).5

Feuerstein employs both Amṛtanāda and Amṛtanādabindu for the text of thirty-eight

ślokas.6 Generally, I have referred to the texts as the Amṛtabindu (twenty-two

ślokas) and Amṛtanāda (thirty-eight ślokas) following the titles of the texts in the

Adyar Library Series edition. Although conventional, this practice does cause a

4 I have not reviewed a Sanskrit edition of the CU. I have relied on Deussen’s translation

for analysis and summary of the text.

5 Deussen (1997), p. 691.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

minor deviation from the the traditional names ascribed to the group of five

"bindu" texts, which are designated in their original titles Brahmabindu,

Amṛtabindu, Dhyānabindu, Nādabindu, and Tejobindu. The whole canon of Yoga

Upaniṣads does not together represent any intentional community of Brahmanical

or Vedāntin yogins, yet the bindu texts are attested at roughly the same time and

do present a doctrine of oṃ meditation.⁷ Of all the twenty-one Yoga Upaniṣads,

these five texts fit together in their content and their vocabulary and in the style

used by their authors.⁸

The Yogadarśana Upaniṣad and the Darśana Upaniṣad are titles of a single text

of 224 stanzas (# 5 below). These titles refer to only one text, and the Yoga portion

of the title is sometimes added by convention since the text's subject is yoga and

since it is classified as a Yoga Upaniṣad.⁹ The Cūlikā Upaniṣad also bears the title

Māntrika Upaniṣad in its eighteenth-century South Indian recension, which could

lead to further confusion over the number of texts in this category. Haṃsa

Upaniṣad also appears under the title Hamsanāda Upaniṣad (Persian Hensnad) in the

Persian translation of Benares manuscripts, the Oupanekhat (1656).

⁶ ANU: see Feuerstein (1997), p. twenty-one. For ANBU, see Feuerstein (1998), p. 415.

⁷ The TBU is an exception, as but for its first verse it does not refer to mantra yoga.

⁸ I intend by this statement the northern recensions (called A below) of the bindu texts.

⁹ It is also a logical title as the text summarizes the whole yoga tradition and darśana means

"philosophy" or "viewpoint." Thus, The secret teaching of the Viewpoint of Yoga accurately

summarizes the contents of the texts. See the summary below for additional analysis of the title.

110

History, Text, and Context of the Yoga Upanisads: Chapter 3

Throughout this study I have employed the titles as listed by the Adyar

Library Series collection of the texts, noting the variations within the textual

synopses and elsewhere where appropriate. The texts appear in Roman

alphabetical order for the readers' convenience, with one exception. The Cūlikā

Upaniṣad appears below as number twenty-one because it forms a special case in

that contemporary scholars have generally not included it as a yoga text.

Accordingly, it does not appear in the Sanskrit Adyar Library Series edition nor

the Ayyangār translation of Yoga Upaniṣads. The Adyar Library Series presents the

texts in Sanskrit alphabetical order. The traditional numbering of each text differs

from one Upaniṣad collection list to another in the traditional lists of forty-eight to

108 titles. The southern list is somewhat arbitrary in its ordering and differs from

traditional Vedic schools as to which Veda it assigns each text. The northern

collections attach all of these texts to the Atharvaveda. The order of the titles in the

North Indian lists may relate to their relative dating, but this assessment is

uncertain. Thus, I do not employ the alternate classifications systems for ordering

the textual synopses below.10

10 These classification systems are discussed in chapter 2 and presented here in each

summary.

111

History, Text, and Context of the Yoga Upanisads: Chapter 3

There are many recensions of these texts divided into North Indian and

South Indian collections.11 The northern collections (Ānandāśrama Sanskrit Series

and Bibliotheca Indica)12 contain eleven of the texts in short versions (forty ślokas

or less): Amrtabindu, Amrtanāda, Brahmavidya, Dhyānabindu, Hamsa, Kṣurika,

Nadabindu, Tejobindu, Yogaśikha, Yogatatva, and Cūlikā Upaniṣads. Northern

recensions are referred to herein as (A) or recension (A). The southern collections

contain all twenty-one designated Yoga Upaniṣads, including the eleven texts found

in the northern collections (several of the titles contain significantly expanded

texts) and ten additional texts not attested in the northern tradition. I designate

the longer southern recensions as (B) or recension (B).13

A list of fifteen Upaniṣads of the Atharvaveda, that dates from the ninth to

eleventh centuries, includes the titles of the Brahmavidya, Kṣurika, and the Cūlikā

Upaniṣads. All eleven of the northern titles are attested in a redaction of twenty-

11 The classification of recensions into northern and southern divisions is an idealization of

the manuscript traditions. There are many variant manuscripts scattered across the subcontinent.

In general, the northern recensions are the short versions; the expanded texts are the southern

recensions. These designations signal the origination of the manuscript traditions rather then their

geographic locations, as short recensions are also found in south India. The primary contemporary

collections are the Adyar Library Series = Adyar Library Series, Madras; ĀSS = Ānandāśrama

Sanskrit Series 29, Poona; BI = Bibliotheca Indica 76, Calcutta. There are many additional private

and published collections as well as isolated manuscripts. Several scholars note their examination

of Banares manuscripts ((A) recension texts), as well as others versions. It may be that the Poona

ĀSS collection was constructed primarily from south Indian manuscripts. This collection deserves

further research concerning its exact disposition. Bouy (see Bibliography) examines and comments

on several additional manuscript traditions.

12 ĀSS = Ānandāśrama Sanskrit Series, 29, Poona; BI = Bibliotheca Indica, 76, Calcutta see

previous note.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

eight Upaniṣads of the Atharvaveda, between the twelfth and thirteenth centuries.

This collection lists the Yoga Upaniṣads in the following order: Brahmavidyā,

Kṣurika, Cūlikā, Nādabindu, Brahmabindu, Amṛtabindu, Dhyānabindu, Tejobindu,

Yogaśikha, Yogattatva, and Hamsa Upaniṣads.14 With certain reservations, one may

surmise that the order of these lists closely approximates the sequence of

authorship. It is likely that the Brahmavidyā, Kṣurika, and Cūlikā Upaniṣads are the

earliest Yoga Upaniṣads, followed by the bindu texts and yoga texts (Yogaśikha and

Yogattatva). The Hamsa Upaniṣad was probably the latest of the northern Yoga

Upaniṣads. Since serial ordering is not certain, we may at least assert that the texts

were written in these three phases.15 Narāyāṇa's dīpikā attests the (A) recension of

the texts, between 1500 and 1700, in the same order as the Caraṇavyūha list of

twenty-eight texts. This at least demonstrates that the eleven northern Yoga

Upaniṣads (in the above order) were considered authoritative and canonical texts

of the Atharvanic tradition by around the fifteenth century. Moreover, they were

already attested before that date.

13 Referring to the short northern texts as (A) and the southern texts as (B) is a modified use

of the terminology of Bouy (1994). As already stated, this practice is an idealization, as (A) or (B)

may represent various manuscripts originating in different locations at different times.

14 For the lists of 15 and 28 see the comments of Bouy (1994) on the Caraṇavyūha.

15 For example, the ABU (also called the BBU) is attested before any of the other bindu texts

but appears after the NBU in the list of 28. It is most likely that the bindu texts (A) were all written

in relatively the same time period. From different perspectives, both Deussen (1997) and Bouy

(1994) would agree on a series of three phases. See Bouy (1994), pp. 117–118, and throughout for

his discussion of the Caraṇavyūha collection of the Upaniṣads of the Brahmaveda (the Atharvaveda).

See also Bouy (1990a), throughout.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The expanded (B) recensions are first attested in Upaniṣad Brahmayogin's

vivaraṇa (1751) and the bhāṣya of Appaya Dīkṣita (late nineteenth century).16

The printed edition of the Aṣṭottaraśatopaniṣadah (Sanskrit written in Telugu characters)

was printed in 1883 and saw relatively wide circulation until the Adyar Library

Series (Sanskrit written in Devanagari characters) edition was published in the

early twentieth century (1920). The Telugu characters edition was published

without word separations or critical apparatus, and it contains numerous mistakes

and corrupt passages. It is likely that several of the (B) recensions were expanded

by the redactor(s) of the collection of 108 Upaniṣads probably in the late

seventeenth or early eighteenth century.17 Thus, the (B) tradition represents the

editorial transformation of the canon by an individual or particular school of

thought active in Tamil Nadu before or around 1700.18

As discussed in chapter 1, all of Yoga Upaniṣads, with the exception of the

Cūlikā Upaniṣad, were collectively translated by T. R. Śrīnīvāsa Ayyangār in a text

edited by Pandit A. Mahadeva Sastri (Madras: Adyar Library, 1938). Ayyangār

prepared his translation from the Adyar Library Series Sanskrit edition (1920)

commented on by Upaniṣad Brahmayogin (commentary 1751). Ayyangār relies

heavily on Upaniṣad Brahmayogin's commentary in his translation, often

16 For discussion of the recensions and attestation, see Bouy (1994), p. 30, and passim.

17 See Bouy's (1994) concluding remarks relative to several of the southern expanded texts.

Even the yoga texts that he does not assess appear to have ta similar developmental history.

114

History, Text, and Context of the Yoga Upanisads: Chapter 3

interpolating material into the translation from the commentary.19 Moreover, the

Adyar Library Sanskrit edition, upon which Ayyangār’s translation is based, is not

a critical edition. It is unedited and without critical apparatus and includes

several corrupt and confused readings.

Additional translations by various scholars are noted below with each

synopsis. Aiyar’s translation was made from expanded southern recensions of the

texts, but his translation is based on manuscripts that differ in some details from

the Adyar Library edition. The 1997 revised edition of Aiyar’s translations

includes the Sanskrit text in Devanagari characters, which differs from the Adyar

Library edition in several details. It is not clear from his translation whether it was

these Sanskrit texts were the texts that Aiyar used, as they were not included in

the 1914 publication. In preparing his translations, Deussen consulted the

Ānandāśrama Sanskrit Series and Bibliotheca Indica collections as well as the

southern 1883 printed edition, the Aṣṭottaraśatopanisadāḥ (written in Telugu

characters). Deussen’s translations always represent the northern tradition of

unexpanded texts (usually the Bibliotheca Indica), although he did consult the

18This is an hypothesis, not a proven fact. To date, it best fits with what we know of the

southern tradition. This issue is examined in detail in this work’s conclusion.

19Although it is often necessary for translators to rely on commentaries, the reader is

reminded that the commentaries interpolate and extrapolate and are often from different times and

places than the texts themselves. This process tends to conventionalize the texts in terms of the

cultural mileiu of the commentators. Varenne and Feurstein also rely heavily on Upaniṣad

Brahmayogins comments. Deussen tends to rely on Nārāyaṇa and Colebrooke, and even

Śaṅkarānada for his interpretation of the northern recensions (A).

115

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Telugu edition concerning corrupted and unreadable passages. Varenne prepared

his French translations by consulting southern recensions, probably the Adyar

Library Series. Feuerstein's translations derive from the Adyar Library Series

edition, as he refers to the commentary of Upaniṣad Brahmayogin concerning

difficult readings.20

In the following section, I summarize the twenty-one Yoga Upaniṣads, both

northern and southern recensions. With modern publication of the Adyar Library

Series edition, the southern recensions gained prominence as the definitive Yoga

Upaniṣads. Most contemporary scholars have relied on the Adyar Library Series

(or Ayyangār's translation from the Adyar Library Series) for their analysis and

comments on these texts. The Adyar Library Series is randomly organized

(Sanskrit alphabetical) and thus does not provide a logical or topical introduction

to the texts. Feuerstein (1998) has attempted to give the texts a logical order, and

his grouping of the Adyar Library Series recensions is useful for a thematic

introduction to the southern tradition.21

The traditional southern ordering, which provides numerical sequence of

108 texts, appears to be arbitrary. This is somewhat less so with the numerical

20 Neither Varenne nor Feuerstein cite their Sanskrit sources. From their versification,

length, and translations it is clear that they employed southern recensions. Both authors appear to

rely on Upaniṣad Brahmayogin's commentary (1751).

21 Even so, his is a brief and general textual summary. Feuerstein does not address history

and dating, recensions, location, or other more detailed issues. Feuerstein is unaware of the two

different manuscript traditions or the sources of the expanded texts.

116

listings of the northern collections. Reading the (A) recensions in serial order does

provide some consistency to the body of eleven northern texts. Should the reader

wish to assess groups of texts in their likely historical order, the following

numbered entries should be read according to the timeline chart that follows (see

Figure 3.1). The dating in Figure 3.1 is based on textual attestation through title

lists, quotations, or commentaries. All dates are approximate. Terminus ad quem

dating is inferred from the absence of commentary within known texts of the same

tradition. For example, that Śaṅkara did not comment on any of the Yoga

Upaniṣads, or that the northern commentators did not comment on the expanded

southern texts, suggests that the texts were written after the commentaries. This

form of dating is ultimately undependable but can be relatively accurate when

combined with supporting evidence. Author, authorial date, and geographic

locations are unproven in all cases.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Figure 3.1. Approximate Dates of Attestation for (A) and (B) Recensions of the Yoga Upaniṣads

800-1100

Early attested northern (A) recensions

(A) Brahmavidyā Upaniṣad

(A) Kṣurikā Upaniṣad

(A) Cūlikā Upaniṣad

1100-1300

(A) Nādabindu Upaniṣad

(A) Amrtabindu Upaniṣad

(A) Amṛtanāda Upaniṣad

(A) Dhyānabindu Upaniṣad

(A) Tejobindu Upaniṣad

(A) Yogaśikhā Upaniṣad

(A) Yogatattva Upaniṣad

1200

(A) Haṃsa Upaniṣad

1300

Earliest B recension

(B1) Yogakundali Upaniṣad

1300 (-1700) (?)

(B1) recensions that are similar in length and composition to the (A) recensions

(A/B1) Amṛtabindu Upaniṣad

(A/B1) Amṛtanāda Upaniṣad

(A/B1) Kṣurikā Upaniṣad

(B1) Mantrikā Upaniṣad (southern recension of the (A) Cūlikā Upaniṣad)

(continued)

118

History, Text, and Context of the Yoga Upanisads: Chapter 3

Figure 3.1. (continued)

(1500-1700)

(B) expanded recensions that have previous short (A) or (B1) recensions

(B) Brahnavidyā Upanisad

(B) Nādabindū Upanisad

(B) Dhyānabindū Upanisad

(B) Tejobindu Upanisad

(B) Yogasikha Upanisad

(B) Yogatattva Upanisad

1600-1750

Late texts: unattested before (B) recension

Short text with unknown sources

(B1 ?) Advayatāraka Upanisad (may be a source of Mandalabrāhmana Upanisad)

Long compilations with known and unknown non-Upanisadic sources

(B) Darśana Upanisad

(B) Mahāvākya Upanisad

(B) Mandalabrāhmana Upanisad (may be expansion of Advayatāraka Upanisad)

(B) Pāśupatabrahmā Upanisad

(B) Śāndilya Upanisad

(B) Triśikhibrahmana Upanisad

(B) Varāha Upanisad

(B) Yogacūdāmani Upanisad

(B2) Yogakundali Upanisad

119

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

3.2: The Twenty-one Yoga Upaniṣads22

3.2.1. Advayatāraka Upaniṣad ("Secret Teachings of the Non-Dual Liberator")23

The Advayatāraka includes nineteen numbered sections of varying lengths:

some metered verse, others in prose. The southern Telugu tradition (1883) of

Upaniṣads is the first attestation of the Advayatāraka.24 Dating is uncertain, but it is

probable that the redactor of the 108 Upaniṣads compiled this text between 1600

and 1700. The text takes inspiration from the Siddhasiddhāntapaddhati (1600-1650)

and contains certain passages that are reminiscent of portions of the

Laghuyogavāsiṣṭha (ca. 1350), the Āmanaskayoga, and similar to passages in the

Vijñānabhairavatantra.25 In the latter example, there appears to be no direct

borrowing. This text is brief and "experimental" in its orientation. Although it is

mostly in prose, it is reminiscent of the older aphoristic northern Yoga Upaniṣads.

22 For additional summaries of these texts see the following. Feuerstein provides

summaries for 20 (B) Yoga Upaniṣads (1997: by title, and 1998: pp. 413ff). Eliade summarizes the (B)

DhBU, NBU, and YTU in detail (1973: pp. 128ff). Varenne comments on the contents of several of

the (B) texts (1973 introduction and notes, and 1976: throughout). Banerji (1995: pp. 379-381)

provides thumbnail summaries of (B) recensions, too brief to be of much use to the reader. All of

these summaries analyze the southern recensions (Adyar Library Series edition). Deussen offers

detailed introductions and summaries to his translations of the eleven northern (A) recensions.

23 Additional translations: (B) recension; Feuerstein (1998), pp. 427-431.

24 This text may have north Indian versions. It is not named in the normative lists of

northern (Atharvanic) texts (Caraṇavyūha, Oupnek’hat, Nārāyaṇa, Colebrook, Deussen).

25 Many portions of the ATU contain similar vocabulary to that found in the LYV. I have

not discovered any direct borrowing. There as also a Nāth text from the twelfth to thirteenth

centuries, Amanaska Yoga (1.50-98) of Gorakhnāth that contains section on Tāraka yoga with similar

focus and language.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

T. R. S. Ayyangār introduces the text with the following gloss, derived from

Upaniṣad Brahmayogin's commentary. "This Upaniṣad, which is Fifty-third

among the 108 Upaniṣads and forms part of the Śukla Yajurveda, fixes its goal in the

brahman and brahman only and seeks to explain the essentials of Rāja-yoga." The

latter assessment can only be interpreted to mean that the commentator saw this

as a meditation text instead of a haṭha yoga text because Vivekananda's term, rāja-

yoga, is otherwise not apparent in the text.

The Advayatāraka describes a discipline called tāraka yoga. Taraka ("that

which liberates") has several applications in the larger yogic historical context.

The term appears in yogic usage as early as Patañjali (3.55) in an adjectival sense

meaning "transcendent" or "liberating." In the Advayatāraka, tāraka means "the

liberator" in an objectified sense as an entity or force outside the practitioner as

well as a program of practice, liberator-yoga. This polyvalent usage of the term is

similar to the employment of other terms with expansive semantic fields (such as

bindu, nāda, and Brahman). The text calls visual meditation practices "the

liberator" in a similar way to those in which tāra and the tāraka are employed to

refer to the mantra om. Forms of the word are used as an epithet for om in the

Pāśupatabrahma Upaniṣad and other minor Upaniṣads whose subjects are not yoga:

Atharvaśira Upaniṣad, Nṛsiṃhapūrvatāpanīya Upaniṣad, and Tāraka Upaniṣad.

121

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The author of the Advayatāraka Upaniṣad describes tāraka yoga as involving

two levels of practice: conscientious visual meditations and a higher practice that

is beyond mental exercises or spontaneous. In the text, tāraka often refers to the

various light phenomena that appear to the yogin during meditation. Other yoga

texts and the Tantras describe similar light phenomena. The text describes the

yogin as seeing blue, yellow, and other colored lights—along with other visual

effects, such as lightning and darkness, similar to those described in the

Śvetāśvatara Upaniṣad—both inside and outside the body. Some of these lights are

cultivated as meditation visualizations. Other light manifestations seem to be the

spontaneous result of the meditation practices. This characteristic signals to the

reader that the Advayatāraka is primarily an experimental meditation text. There is

no cosmology or metaphysics in the Advayatāraka. It presents brief descriptive

passages that would accompany meditation practices. The text assumes that the

reader already knows and practices visual meditation. Thus, it provides

meditation "reminder notes" rather than a comprehensive or systematic

exposition of the tāraka yoga.

The text describes the various light effects as different kinds of "signs"

(lakṣya), inner, outer, and intermediate.26 Practice of this yoga results in special

26 The Śāndilya Upaniṣad (1.7.15) defines khecarī mudrā as the breath and mind resting on the

"inner sign" with the same phrasiology as the ATU. These two may share sources. See Śāndilya

Upaniṣad summary.

History, Text, and Context of the Yoga Upanisads: Chapter 3

psychophysical powers and liberation. Word play and puns concerning the word

tāra (pupil of the eye) and tāraka (the liberator) tie the discourse of visual

meditation and the goal of liberation together. The text includes specific

references to sounds heard in the ears and different actions of the eyes and

eyebrows. The liberator (tāraka) is presented in the text as a being or state to be

realized and as the transformed identity of one who practices this yoga. Technical

terms from the kundalini yoga of the Tantras (Kashmiri Śaiva texts, Krama Tantras,

and the Nāth literature) appear throughout, although the text does not give a

systematic presentation. Explicit references to haṭha yoga breathing techniques and

postures are conspicuous in their absence.

This text is likely the predecessor to the longer Maṇdalabrāhmaṇa Upaniṣad,

as the latter contains many similar and identical verses scattered throughout the

text (especially in the first two chapters). The Maṇdalabrāhmaṇa Upaniṣad includes

Vedānta speculations and embeds the visualization yoga of the Advayatāraka in a

discussion of haṭha yoga. Prose verse 7 of the Advayatāraka, like chapter 4 of the

Maṇdalabrāhmaṇa Upaniṣad appears to be inspired directly from the

Siddhasiddhāntapaddhati (ca. 1600-1650) attributed to Gorakhnāth. As such, the

Maṇḍalabrāhmaṇa Upaniṣad may be an expanded compilation based partially on the

Advayatāraka. This is consistent with the style of compilation of all the late

southern Upaniṣads. The exact character of the Advayatāraka is yet unknown since

123

History, Text, and Context of the Yoga Upanisads: Chapter 3

it is a late attestation just as the Mandalabrāhmaṇa Upaniṣad is. It seems possible

that the Advayatāraka is a summary of the Mandalabrāhmaṇa Upaniṣad: in which

case, the more abbreviated text would be derivative of the longer compilation

text.27 The final five verses are praise of the teacher (guru) and offer one Vaiṣṇava

sectarian reference. These verses do not relate to the former verses and are

probably interpolations, like the closing verse of the Amṛtabindu.

3.2.2 Amṛtabindu Upaniṣad ("Secret teachings of the immortal point")28

The Amṛtabindu also appears in some collections under the older title

"Brahmabindu Upaniṣad" ("Secret Teachings of the Point that Signifies Brahman").

It is probable that the Amṛtabindu was originally written between the eighth and

thirteenth centuries under the title Brahmabindu Upaniṣad. According to Deussen:

The Upaniṣad rather treats of the dissolution of the Manas (verses 1-10),

the relationship of the Brahman with the world of appearances (verses

11-17), the worthlessness of external perception, bookish learning etc.

(verses 18-22) in a way which makes it appear like a link between the

27 That the ATU might be a summary of the MBU seems unlikely, but possible since the

character of the text is unknown. The ATU employs several technical terms that suggest that it is

heir to a technical textual tradition. So far, I have not been able to find any more direct borrowings

than is found in verse 7. The most promising approach to further elucidation of the ATU would be

discovery of additional texts that expound on "the three lakṣya" or contain similar usage of the

term "tāraka." These may be expounded in Gorakhnāth's Amanaska Yoga; further research is

necessary on this Nath text. All other details and technical terms are relatively wide spread in their

usage in the Tantras and in the yoga literature. Determining the evolution of the ATU will be

difficult without further discovery of direct quotes or citations. Nothing is certain with this text. It

has the tone of being a genuine experimental text more than a derivative textual compilation.

28 Additional translations: (A) recension, Deussen (in English by Bedekar and Palsule),

under the title BBU, (1997), pp. 687–690; (B) recension, Aiyar (1997), pp. 27–28; (B) recension,

Varenne (in French) (1971), pp. 123–127; (B) recension, Feuerstein (1998), pp. 45–47.

124

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

older Upaniṣads and Śaṃkara, whose favourite illustration of the space

in the jar and the universal space already appears here (verses 13-14),

as also the verse 12 which is indeed cited by Śaṃkara under

Braḥmasūtra p. 810,1, possibly from our Upaniṣad.29

Although Deussen’s analyses are in many ways dated, it is certain that the

Amṛtabindu (Brah mabindu Upaniṣad) is one of the earliest texts classed as a Yoga

Upaniṣad. Thus, the Amṛtabindu may date from around the time of Śaṅkara and

offers its reader a glimpse into the mantra yoga of the medieval Vedānta tradition.

The Jīvanmuktīviveka of Vidyāraṇya (literary activity 1330-1385 CE) quotes the

Amṛtabindu. This is a text of twenty to twenty-four ślokas similar to the Amṛtanada

Upaniṣad (a different text, but also called Amṛtabindu in some recensions). It

focuses on Vedāntin interpretations of renunciation and the recitation of the

syllable oṃ. The (A) and (B) recensions of this text are nearly identical except for

minor textual variations. Text (A) is numbered eighteenth in the Atharvanic

collections. The Adyar Library Series (B recension) begins with the following

gloss: “This Upaniṣad, which is Twentieth among the 108 Upaniṣads and forms part

of the Kṛṣṇa Yajurveda, expatiates on the mind as the cause of bondage as well as

liberation of man and shows how, by the knowledge of Brahman only, the final

goal is attainable.”30

29 Deussen (1997), p. 687, italics added.

30 Ayyangār (1952), p. 17, drawn from Upaniṣad Brahmayogin’s commentary.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

This text shares common verses with the Maitrāyaṇīya Upaniṣad. Deussen

argues that the verses make sense and flow better in the Amṛtabindu than in the

Maitrāyaṇīya Upaniṣad, and that the Maitrāyaṇīya Upaniṣad was therefore possibly

quoting the Amṛtabindu.31 However, the Amṛtabindu is not attested in the written

tradition until much later (at least 500 years) than the Maitrāyaṇīya Upaniṣad. The

verses are not original to the Maitrāyaṇīya Upaniṣad, and therefore it appears that

either both texts are drawing on a third (lost) text or sayings-source or the

Maitrāyaṇīya Upaniṣad is quoting from a very early recession of the Amṛtabindu.

Such an early recession is not cited or known from any other sources.

The Amṛtabindu begins its discussion by describing the manas (mind) as the

source of both bondage and emancipation. When attached to objects, the mind is a

source of bondage. When unattached and free of desire, the mind can work for

liberation. The yoga described in the Amṛtabindu focuses solely on recitation of,

and meditation on, the sacred syllable (mantra) oṃ. Pronouncing and repeating

the mantra oṃ is the preliminary practice. Once the mind rests in the heart

through use of the oṃ mantra, then the practitioner must progress to the soundless

recitation. Tonal or sounded (svara) meditation is considered inferior to non-tonal

or soundless (asvara) meditation.

31 Deussen (1997), p. 687.

History, Text, and Context of the Yoga Upanisads: Chapter 3

Several Upanisadic metaphors describe the nature and state of the seeker.

The soundless practice results in knowing the aksara ("imperishable," a term used

to refer to om and to Brahman) and knowing the highest absolute reality

(Parabrahman) and attaining peace in the essential Self (ātman).

The closing two verses associate absolute reality (Brahman) with the name

Vāsudeva, an epithet of the god Viṣṇu. The meter of these verses fails to match

that of the previous verses, and therefore these latter verses are probably later

additions by an editor who was a Vaiṣṇava sectarian.

3.2.3. Amṛtanāda Upaniṣad ("Secret Teachings of the Immortal Reverberation")32

As discussed earlier, this text was originally entitled Amṛtabindu ("Secret

Teaching of the Immortal Point"). It sometimes bears the alternative title

Amṛtanādabindu Upaniṣad ("Secret Teaching of the Immortal Sound Point"). The

Amṛtanāda was probably composed between the twelfth and thirteenth centuries.

It was in existence before the Jīvanmuktiviveka of Vidyāraṇya, who quotes both it

and the Amṛtabindu. Sāyaṇa refers to this text as the Amṛtanāda. He cites its tenth

verse under Taittirīya Āraṇyaka 10.27.33 The northern (A) and southern (B)

32 Additional translations: (A) recension, Deussen (in English by Bedekar and Palsule),

under the title Amṛtabindu, (1997), pp. 691-698; (B) recension, Aiyar (1997), pp. 164-166; (B) recension, Varenne (in French), (1971), pp. 116-122; (B) recension, Feuerstein, under the title of ANBLI, (1998), pp. 416-418.

33 Deussen (1997), p. 691.

History, Text, and Context of the Yoga Upanisads: Chapter 3

recensions of this text are nearly identical. (The (A) text is number nineteen on the

Atharvanic lists. The (B) text has the following gloss:

This Upanisad, which is the Twenty-first amoung [sic] the 108 Upanisads

and forms part of the Krsna Yajurveda, demonstrates that the pure-

minded attain the end and aim of existence, by adopting the

expedients of Śravana-study, and Manana-reflection, etc., while those

with an impure mind attain their enfranchisement from worldly

existence by having recourse to meditation of the brahman and

practice of Sad-anga-yoga.34

This text has thirty-eight to thirty-nine stanzas discoursing on a six-limb yoga

system that is identical to the six-limb system of the Maitrāyaniya Upanisad but

with a different order.

Instead of the later title Amrtanada, the name Amrtabindu Upanisad is better

attested and more appropriate to the subject matter of this esoteric teaching (see

discussion of name changes above, 3.1). This title suggests the subject matter of

the text: "The esoteric doctrine of the nasalization (anusvāra) point (bindu) that

grants immortality." Amrtanada is the title used in the Adyar Library Series and

employed by this study. Without the title change, the text is more easily identified

with the group of bindu texts that, as mentioned earlier, more than any other

subset of the Yoga Upanisads actually coheres as a system of texts.

34 Ayyangār (1920), p. 9, translating Upanisad Brahmayogin's commentary. Unfortunately,

Brahmayogin's summary is nonesensical since the text explicitly rejects bookish learning in favor

of yoga (another example of the prejudices that can appear in a commentary). It is correct as to the

number and Vedic attribution as well as acknowledging a six-limbed yoga.

History, Text, and Context of the Yoga Upanisads: Chapter 3

Like the other bindu texts, the Amrtanāda is a mantra yoga text. The text

begins with a characterization of om as "the chariot." Just as a chariot is left

behind when traveling on high ground, the sounded om is left behind by means of

the unsounded m (= the bindu that signals the nasalization of the m). The

Amrtanāda is more detailed than the Amrtabindu, describing a yoga system of six

limbs: sense-withdrawal (prayāhāra), meditation (dhyāna), breath-control

(prānāyāma), concentration (dhāranā), reflection (tarka), and enstatic absorption

(samādhi). These are all body-meditation practices. The text does not include

moral and ritual obligations such as those found in the eight-limbed systems of

Patañjali or the Buddha, although verse 27 does state that the yogin should avoid

fear, sloth, excessive sleep, excessive wakefulness, excessive eating, and non-

eating. The text describes pratyāhāra as withdrawing the manas and indriyas from

sense objects. Dhyāna, is listed but not explained. Breath-control is described as

being able to recite three times in one breath the gāyatri mantra, vyāhrtis, and

pranava (om apo jyoti raso'mṛtam brahma, bhūr, bhuvaḥ, svaṛ, om).35 Several verses are

devoted to breath-control, the practices of exhalation, inhalation, and holding the

breath (recaka, pūraka, kumbhaka). Each of the three controls is described physically

35 Deussen (1997), p. 693, § 1. Deussen states that this verse is the same as Viṣṇusmṛti 55.9.

Sāyaṇa includes the formula stated above: saying the "head of the Gāyatrī" is the formula that

crowns it (follows it). Olivelle (1992), p. 283, § 14, gives the same formula. Feuerstein's comment is

somewhat unclear on the mantra, probably "Bhuh, bhuvaḥ, svaḥ, om tat savitur varenyaṃ bhargo

devasya dhīmahi dhiyo yo naḥ pracodayāt, paro rajase'savād om." Cf. Maitrāyaṇīya Upaniṣad 6.3-7. See

also Padoux (1992), p.27.

129

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

(verses 11-13).36 Dhāraṇā is concentrating the mind (manas) as intention (saṅkalpa)

in the Self (ātman). Tarka is inference (ūhana) that is consistent with tradition

(āgama). Samādhi is obtained (labdh) when thinking is even (sama).37

The text follows with elements of a practical program for meditation: level

ground, protecting the mind, repeating oṃ (the round of the chariot mantra: ratha-

maṇḍala),38 sitting postures,39 and alternate nostril breathing. Thus prepared, after

drawing in the breath, the yogin may hold steady in himself the fire (agni) and may

meditate on the sound [of oṃ].40 Through repetitive practice of breath-control and

mantra recitation, the yogin moves the breath up through the body through the

gate (door, aperture) of the heart, through the gate of the throat, and then through

the gate of the head. These texts use the term "gate"41 (dvāra) for esoteric centers

of the body, similar to the cakra (center) and padma (lotus) in the Tantras. Simple

36 Many manuscripts confuse verses 12 and 13. Verse 13 (pūraka) should precede verse 12

(kumbhaka). Instead of reordering the verse, some copiest transposed the terms, placing each with

the wrong definition. This led Narayana mistakenly to state that the text referred to a special form

of kumbhaka. See Deussen (1997), pp. 693-694, especially § 9. It seems natural that the Upaniṣad is

not developing these ideas but putting them forward as conventional notions of breath control.

37 The wording of this verse varies considerably across manuscripts. Some do not include

the term "samam."

38 The yogin is instructed to repeat the ratha-maṇḍala. Deussen does not understand this

verse. Feuerstein says that the chariot and the wheels are oṃ and the formulas mentioned in verse

10. This is likely correct as the begining verses set up the metaphor of oṃ as ratha, but should

probably not read chariot and wheels, but a round of the chariot-[mantra] = verse 10: saying the

formula thrice in one breath (a cycle or series of repetitions).

39 padmāsanam, svāstikam, bhadrāsanam.

41 ākṛsya dhārayedagniṃ śabdameva vicintayet (verse 19b). "Agni" may be this texts way of

refering to what the Tantras call kuṇḍalīṃ. Otherwise, it may mean the heat of asceticism (tapas) or

the fire of yoga (yogāgni), ideas that are similar to kuṇḍalīṃ but generally less systematic. As Eliade

and others have remarked, ecstatic practices that hold, cultivate, or raise internal heat in the body

are widespread among Shamanic cultures and throughout the elite esoteric systems across Asia.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

classifications of breaths—name, location, and color—are included. Verse 26 (27

in some texts) contains an interesting list of the gates or doors through which the

breath proceeds. Translators and commentators have interpreted the verse

variously as listing three doors, four doors, or as a list of seven doors or centers. I

will analyze this issue in chapter 4.42 When, through practice, the breath pierces

through the circle (mandala) at the peak of the head (probably the peak of the skull,

the fontanelle) the yogin attains liberation and is never reborn.43 Understanding

verses 26 as well as verse 38 correctly are crucial for evaluating the nature of the

subtle body in the Amṛtanāda Upaniṣad (as will be discussed further in chapter 4).

Here we find a systematic yoga text that, even if impressionistic, contains many of

the developed concepts of medieval yoga.

41 dvīra, door, gate, passage, opening, aperature (especially of the human body).

42 Ayyangār and Varenne interprets seven “doors” and seems to equate them to the

standard seven (six plus one system) cakras (probably because the verse is explained thus in

Upaniṣad Brahmayogin’s commentary). Whether the text actually refers to three, four, or seven

gates, it does not equate to the standard system of the GŚ or Haṭhayogapradīpikā or even other

systems of cakras (such as the five cakras of the Kubjikā Tantras). See chapter 4 of this study for

consideration of this verse.

43 The exact nature of this verse is also unclear or often misinterpreted. Most translators

read to verse as saying that after the breath peirces the circle (mandala) it goes into the head. It

should be read as “escapes at the top of the head” (yati mūrdhan: use of locative as “at” instead of

“into” since mūrdhan means roof, top, peak, uppermost as well as just head or skull). This issue is

also further discussed in chapter 4.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

3.2.4. Brahmavidyā Upaniṣad ("Secret Teachings of the Arcane Science of Brahman")44

Like some of the other texts in this collection, the Brahmavidyā Upaniṣad exists in a short northern version (A) of fourteen verses and a much longer southern version (B). The northern recension of the text is first attested between the ninth to eleventh centuries. It is numbered third on the Atharvanic lists, preceded only by the Muṇḍaka and Praśna Upaniṣads. The shorter northern text is exclusively devoted to analogies and allusions concerning the sacred mantra, om.

The classical Upaniṣads had already broken the sacred mantra into its constituent phones or morae (mātrās) (a + u + m).45 After equating the phones with three of the "four states"—waking, dreaming, and deep sleep—the Māṇḍūkya Upaniṣad suggests that there is a fourth state that is without a constituent phone.

Following Olivelle's translation of Māṇḍūkya Upaniṣad 11, it is "the cessation of the visible world; auspicious; and unique." It is here that the Brahmavidyā takes up and extends the discussion. It adds a third and one half phone: "The buzzing reverberation (nāda) of m, which, along with this latter, was denoted by a point (bindu) of anusvāra placed above the syllable."46

44 Additional translations: (A) recension, Deussen (1997), pp. 667-670.

45 Phone and mora are two terms used to translate mātrā the smallest unit of sound. See also Padoux (1992), pp. 14ff, and throughout for discussion of mantras and om.

46 Deussen (1997), p. 667. To my knowledge, this is the oldest Yoga Upaniṣads, and is the first Upaniṣad to distinguish the third and one-half mora of the om. All the other Yoga Upaniṣads take this characterization of the mantra as given. I do not agree with Feuerstein (1997: "om" entry)

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The Brahmavidyā Upaniṣad begins with statement that its "Arcane Science of

Brahman" is the origination and end of Brahman, Viṣṇu, and Maheśvara. Then it

summarizes that the arcane science consists of four characterizations of om: body,

location, time, and vanishing (śarira, sthāna, kāla, and laya). The presentation of the

"body" of om provides characterizations similar to those found in Praśna Upaniṣad

5.6-7 (metered quotes from unknown source) and Maitrāyaṇīya Upaniṣad 6.5. In

the text, the sound a is equated with the Rgveda, the gārhapatya fire, the earth, and

the deity Brahma (verse 5). The sound u is equated with the Yajurveda, dakṣiṇa fire,

the mid-regions, and the deity Viṣṇu (verse 6). The sound m is equated with the

Sāmaveda, āhavanīya fire, heaven, and the deity Īśvara (verse 7). The half phone is

mentioned in verse 4 but not elaborated. This is because it is considered

transcendent and therefore it possesses no homologues, being identified with the

unqualified Brahman.

The text (verses 8-10) next explains the location of om. The a is in the sun-

circle (sūrya-mandala), being in the middle of the "conch" (śaṅkha47). The u, of the

that the "bindu" is alluded to in the Maṇḍūkya Upaniṣad, as the Maṇḍūkya Upaniṣad describes the fourth state as mora-less

(amātra), instead of as the half-mora (ardha-mātrā) of the bindu-anusvāra. There may be sources

outside the Upaniṣadic tradition that first make this distinction. It is possible that this division

appears earlier in the Grammarians or in the Tantras, but I have not been able to discover an

example of this.

47 Location of the "conch" is not clear from the text. Gorakhnāths Amarauaghaśāsana locates

a "conch" (f. śaṅkhini at the root of the body = kundalini). Other Tantras use the feminine term

śaṅkhini for a nāḍī that opens in the cranial vault. Deussen interprets the conch seat of om as being

in the head = the brain case. Although it is not clearly explained, the following verses show the

pointed flame piercing the central channel. If this piercing is at the lower opening of the central

channel, it is possible that the 'conch' of om is located in the region of the navel. It is sensible that

133

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

appearance of the moon (candra), is stationed within it. The m, has the appearance

of fire (agni), is smoke[less],48 and resembles lightning. Thus, the three mātrās

should be known as the forms of sūrya, soma, and agni. The pointed flame (śikhā)

flickers like in that space above a lamp; the half phone (ardhamātrā) should be

known thus, standing above the praṇava.49

Time or sequence, (kāla) is the next characterization (verses 11-12). The

pointed flame is subtle like a lotus threads, she50 is seen as the highest (parā). She,

the sun-like artery (nāḍi sūryasaṃkaśā), after splitting (bhitvā, "breaking through")

the sun, in the same way [is seen as] the highest (parā). And after splitting the

the mantra could start low in the body and come up through the body as the "pointed flame." For

additional discussion on various "conches" in the body, see White (1996), pp. 254-255, and notes;

and Silburn (1988), pp. 28, 124-131. For further examination herein, see chapter 4, section 4.2.1.

48 Adyar Library Series reads vidhūmo: (?) vapor, smoke. Deussen employing the B1, reads

"smokeless."

49 This verse is of interest as it demonstrates its milieu as one in which Sanskrit was written

(not just oral) and the convention of the candrabindu or bindu, (anusvāra) was well established.

50 Grammatically feminine sā referring to feminine śikhā. It is unknowable whether this

synonym for kuṇḍalini in the Tantras of the Western transmission (paścimāmnāya) (M. Dyczkowski,

personal communication, 2001). What relationship these traditions might have needs further

investigation.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

72,000 arteries, [she, the pointed flame] stands at the top of the head (mūrdhani),51 pervading the space of every and all living beings.52

The laya (verses 13-14), fading away or vanishing, of the sound om is compared to the slow fading of a struck gong. Brahman is the peace, or quiet (śānti), wherein the sound fades away. Here we see (like in the Amṛtanāda Upaniṣad) many of the characteristics of the Tantras but without the language of cakras or mention of kuṇḍalini by name. Even so, it appears that the (A) recension of the Brahmavidyā had some systematic knowledge of the language and practices of the Tantras.

The commentator glosses the southern expanded Brahmawidyā as fortieth among the 108 Upaniṣads and as part of the Kṛṣṇa Yajurveda. He plays on the title by extrapolating from the expanded textual contents. He writes that the text expounds the realization of Brahman through the "Arcane Science of the Honking Gander" (praṇavaḥṃsavidyā).53 The southern text of 111 stanzas (recension B)

51 Like verse 38 of the ANU, mūrdhani grammatically may mean "in the head" or "at the peak of the head." The former is the common English rendering of the locative, the latter seems more logical (also a usage of the locative) both in the larger context of tantric and haṭha yogas as well as making it consistent with the metaphor of the text as the bindu stands above the mantra, as the flame stands above a lamp, so the pointed flame stands at the top (above) the head (not in the head). It may also be read that the 72,000 nāḍīs are split or burst "in the head," but this looses the continuity of the metaphor.

52 Translators read the last foot as "And remains the universe," (Deussen: A); and "Pervading all, as if she is the giver of boons to all" (Ayyangar: B) for varadā sarvabūtānāṃ sarvaṃ vajpayeza tiṣṭhati.

53 See discussion below for the complex imagery of the "Gander." Praṇava (from pra-ṇu) means "humming," or "droning" and is used to mean the mantra om, in the Vedāntin literature. Various Tantras privilege other fundamental mantras as their "praṇava."

135

History, Text, and Context of the Yoga Upanisads: Chapter 3

elaborates mantra yoga and nāda yoga. It includes speculations on om, kundalini-

śakti, hatha yoga practice, five-fold breath control, and renunciation. Two-thirds of

the expanded verses (verses 14-80) reproduce the third chapter of the

Upāsanāsārasaṃgraha (or the Hamsavidyopāsanā).54 The expanded southern text was

probably composed in the seventeenth or early eighteenth century. Like several of

the other (B) texts, it may have been expanded by the redactor of the 108 Upanisads

(who probably redacted some or all of the unattested southern texts and expanded

the already extant Yoga Upanisads).

The extended portions of the text focus on elucidation of the concept of

breath-power and body-mind control via the metaphor of the hamsa. Hamsa is the

masculine form of the word for the migratory South Asian goose (sometimes

translated as swan). Thus, hamsa, is the gander. The gander is a metaphor for the

essential Self (ātman) and for the vital breath (prāna). Midway through the

expanded text (verse 53), the teaching is re-framed in terms of a dialogue between

a sage and his student, Gautama (similar to the frame dialogue of the Hamsa

Upanisad). The mantra hamsa is important in many Indian esoteric traditions, from

the Vedas to the Tantras. Many texts associate the in-breath and the out-breath with

the mantras ham (I) and saḥ (that). Combined together these mantras form hamsa.

54 Bouy (1994), pp. 89-91. The Upāsanāsārasaṃgraha is an anthology in 24 chapters that

dates to between the sixteenth to seventeenth centuries. Consistent with the topic, chapter three of

the Upāsanāsārasaṃgraha also exists as an independent text called the Hamsavidyopāsanā. Other

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Repeated in meditation they take on the esoteric interpretation of their reversal, so 'ham.

In all of these traditions, breathing is tantamount to identifying the individual soul with the absolute: hamso 'ham is a palindrome that can be read either as "the goose! the goose!" or "I am That," i.e., "I, ātman, am That, brahman." The cosmic goose, honking in the void, thus becomes a metaphor for the resorption—of individual breath, sound, and soul—into the Absolute.55

The hamsa discourse is elaborated in most of the expanded southern texts. Hamsa is also used to mean the essential Self and the breath in some of the northern recensions. In the expanded texts of the south, it takes on all the significations found in the Tantras and medieval yoga literature. The expanded Brahmavidyā adopts another tantric notion from its sources, stating that its "Arcane science of the Brahman" is good for people of "all walks of life." Although the text thus includes (via its sources) significant tantric materials, it is conservative and Vedāntin in its overall interpretation and tone.

3.2.5. Darśana Upaniṣad ("Secret Teachings of Internal Vision [of yoga]")56

This text, which also appears under the title Yoga Darśana Upaniṣad, is a south Indian text of 224 stanzas in ten sections. The Darśana Upaniṣad is written in

chapters of this text provide the source of some of the additional verses in the DhBU, YŚU as well as the late southern Yoga Upaniṣads, the VUI.

55 White (1996), pp. 211-212. See also Padoux (1992); Eliade (1974); Feuestein (1997, 1998).

56 Additional translations: (B) recension, Varenne (in English by Derek Coltman), under the title YDU, (1976), pp. 201-222]

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

the style of a Yoga Sūtra exposition, yet it is based on Vedānta metaphysics. Its

framework is similar to the Yogayājñavalkya.57 It is only attested in the southern

collection of 108 Upaniṣads and thus may have been written as late as the early

eighteenth century. The gloss explains that the text is part of the Sāmaveda and

ninetieth in the list of 108 being an exposition of eight-limbed yoga, knowledge of

the supreme Brahman, and a description of the Absolute Brahman. The text is a

late compilation text. Its exact sources are not certain, but it borrows the details

and sequence of details from both the Haṭhayogapradīpikā and the Yogayājñavalkya.58

Its terminology is conventional, but it does show originality and style in its

composition.

The text uses the internalization of ritual pilgrimage as one of its dominant

metaphors. Meditation and breathing exercises are depicted as paths and stops on

a religious pilgrimage. Darśana is one of the many Sanskrit words that imply

worship of gods associated with temples and pilgrimages. It has the dual

meaning of "viewing" as worship, and "viewpoint" as in philosophy. As worship

viewing is the interaction of the worshiper seeing the god (often embodied in an

57 There are a number of Yājñavalkya text titles: Yoga- and Yogiyājñavalkyasṃrti,

Yoagyājñavalkyagītā and Yoagyajñavalkyagītopanisad, Yājñavalkyagītopanisad, and others. There are

also various texts that do not have the same content, but possess similar titles. For in-depth study

of the YYV, see Geenens (2000). The YYV dates from between the tenth and early fourteenth

centuries and was written in south India with Vaiṣṇava sectarian orientation.

58 Further research is necessary to clarify the exact disposition of this text. The Yogasāstra of

Dattātreya might also be a source for this text. Additionally, it is not clear whether the text

138

History, Text, and Context of the Yoga Upanisads: Chapter 3

image) and the god seeing the worshipper (like an audience between royalty and

the ruled). The title of this text can be read as "the secret teaching of the internal

vision-pilgrimage [of yoga]." This reading is as likely as the philosophical reading,

"the viewpoint of yoga" since yoga is not included in many of the manuscript titles

and since the text makes much of the metaphors of worship and pilgrimage. It is

also possible that both meanings are simultaneously intended, as this kind of

word-play and double meaning is common for esoteric texts.

The frame dialogue of the text has the sage Sāṃkrti asking his guru

Dattātreya to explain the eight limbs of yoga so that he might become a jīvanmukta

(one who is liberated while living).59 This long text includes a systematic

presentation of the classical eight-limbed yoga system. Among Yoga Upaniṣads, it

provides one of the most systematic and thorough treatments. The Darśana

Upaniṣad explains details about yoga postures, breath control, the internal centers

(cakras) and pathways (nāḍis) of the subtle body, and kuṇḍalinī meditation. This

text is completely consistent with the conventional language of the haṭha yoga

manuals of the Nāths, yoga texts associated with Dattātreya, and many late

Tantras. All of these elements together suggest that this text is a derivative

summary of the yoga traditions. As such, it is a text of high quality, bringing

employs the Haṭhayogapradīpikā or draws on its sources, such as the Gorakṣaśataka or the Yogabīja,

texts attributed to Gorakhnāth.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

together elements from across traditions. Its comprehensiveness provides a

helpful single-source summary of how yoga was understood at the beginning of

the modern period—a synthesis that has become the widely shared “universal”

system of modern yoga (with respect to limbs, definitions, locations of subtle body

structures, and so on)

3.2.6. Dhyānabindu Upaniṣad (“Secret Teachings of the Meditation Point”)^{60}

This text has northern (A) and southern (B) versions. The northern (A)

recensions contain twenty-one to twenty-three ślokas in two sections. The

southern text contains approximately106 to 107 ślokas (with mixed prose).^{61}

Traditionally, the (A) recension is number twenty in the collections of Upaniṣads of

the Atharvaveda. In the southern collection of 108 Upaniṣads, the (B) recension is

associated with the Kṛṣṇa Yajurveda and is listed thirty-ninth.

The northern recension was known by the twelfth to thirteenth centuries.

Bouy suggests that the northern (Atharvanic) redaction of this text was

contemporaneous with or after Śaṅkara (eighth or early ninth century) and before

^{59}See § 90 for association of Dattātreya with eight-limbed yoga. Just so, this text might

have some relationship to the ŚU and its sources, especially the Haṭhayogapradīpikā and the

Yogayājñavalkyam as it shows systematic similarities with both texts

^{60}Additional translation: (A) recension, Deussen (in English by Bedekar and Palsule)

(1997), pp. 699-703; (B2) recension, Aiyar (1997), pp. 154-160; recension (B2), Varenne (in French)

(1971), pp. 71-92.

^{61}For a detailed analysis of the textual sources and manuscript variants of the DhBU, see

Bouy (1994), pp. 32-34, 86-92.

140

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Śaṅkarānanda (ca. 1300). Therefore, he places the northern recension between the

eighth and thirteenth century.62 There is some variation between different (A)

traditions. In particular, the Calcutta edition includes as the two introductory

verses the first two śloka of the Yogatattva Upaniṣad. Narāyaṇa's commentary

considered these two verses as incorrectly added to the Dhyānabindu Upaniṣad, a

position that Deussen supports based on logical continuity and textual

comparison.63

The northern recensions (A) of twenty-one to twenty-three ślokas in two

sections, begins with praise of dhyāna yoga for the removal of all sins.64 Next the

text expounds (verses 4-6) the hierarchy of oṃ recitation. First are the phones

(akṣara) a + u + m. Higher still are the ānusvāra (bindu) and its resonance (nāda).

Higher than the reverberation or resonance (nāda), the syllable vanishes as the

sound (śabda) fades. Silence is the highest place. The yogin is instructed to

meditate on the silence after the sound has faded. Next (verses 7–10) follow a

series of similes—fragrance of a flower, butter in milk, and so on—illustrating the

62 Bouy (1994), p. 31.

63 Bouy (1994), pp. 32-33. Deussen (1997), pp. 699-700. Even so, Deussen does include the

verses in his translation of the text in 23 verses.

64 Here I ignore the first two verses of many manuscripts that are borrowed from the YTU

and begin the summary with verse 3. This summary is drawn from Deussen's translation,

compared with the Adyar Library Series. Even when the (A) or (B1) verses are examined in the (B2)

Adyar Library Series, there are several minor differences. Therefore, further analysis of the exact

language of this verse will necessitate future examination of the Calcutta BI collection (as well as

the Poona ASS edition). Bouy's comments suggest that the manuscripts are virtually the same

concerning these verses.

141

History, Text, and Context of the Yoga Upanisads: Chapter 3

subtlety and pervasiveness of ātman. Next, pūraka, kumbhaka, and recaka are associated with body location, flowers, and gods (verses 11-17, cf. ANU 11-13).

With inhalation, the yogin should meditate on four-armed Viṣṇu, like a flax flower in the navel. While retaining the breath, within the lotus-seat in the heart, the yogin should meditate on father Brahmā, red and white (or pale red), with four faces. With exhalation the yogin should meditate on the three-eyed [Śiva] enthroned in the forehead (or between the eyebrows) shining like crystal. "Three-eye's" flower is described as comprising eight petals, blooming downward, with stalk up and calyx down like the plantain flower (verse 14). Manuscripts differ, but at the end of verse 14 Śiva is described as sarvadevamaya or sarvavedamaya, constituting "all the gods" or "all the vedas."65

In the subsequent verses (15-17), another flower is described, and its relationship to the previous verses is unclear. It is a flower that has one hundred petals surrounding its seedpod (or with seedpod fully expanded). There he should meditate on him who is beyond the radiant fire, moon, and sun. Taking hold of the seed (bīja) of the lotus to carry it to the moon, the fire, and the sun, the ātman surely directs the seed. (Directing the seed [bīja] here may mean directing

65 Gods would be the logical reading of the passage, but both are sensible and consistent with the tradition; where I translate "constituting" the translators suggest both "essence" and "form" for this use of -maya. Variants read also sarvavedamañmukham.

the phones of the mantra through concentration).66 He knows the Vedas who

knows the three places, the three paths, and the threefold Brahman as the holy

sound that has three and one-half phones.

The remainder of the text (verses 18-23) contains metaphors for om,

focusing on the silently reverberating tip of om. This tip, the half phone is a rope

that draws the manas out of the well of the heart up the nädi to the location

between the eyebrows (bhruvor madhye) where it melts away. The location

between the eyebrows is also called the great and perennial resting-place.

The southern version of the Dhyānabindu Upaniṣad appears in two

recensions (B1 and B2). (B1) consists of twenty-one ślokas similar to the (A)

recension, yet the text differs enough to be considered a variant transmission from

the northern Atharvanic texts. (B2) is the expanded text of 106 to 107 ślokas,

including the twenty-one ślokas of (A)/(B1). In other words, the (B1) and (B2)

recensions include variants of the (A) editions' twenty-one ślokas, but with enough

differences to suggest that a different manuscript tradition from the extant (A)

recension was transmitted in the south.67

66This usage of bija as seed-syllable is pervasive in the Tantras. It is consistent with the

metaphorical development throughout the text. Therefore, this seems to be the intended reading

of the metaphor in this verse.

67 Bouy (1994), pp. 32, 86-87. See also Bouy's chart on p. 33 for a comparison of the verses

from the three manuscript traditions (eg., (A) BI verses 11–17 = (A) ASS verses 9–16 = (B2) Adyar

Library Series verses 30-36).

History, Text, and Context of the Yoga Upanisads: Chapter 3

The (B2) text of approximately 106 to 107 ślokas (with mixed prose)

examines meditation on the syllable om. All of the supplementary verses and

prose in the southern expanded text come from other known texts. With the

exception of transition phrases and other minor variations, the expanded

Dhyānabindu Upaniṣad (B2) cites the Gītāsāra (ca. 1600-1650), chapter 6 of the

Vivekamārtanda, the Haṭhayogapradīpikā, and the Rājayogaprakarana (fourth) chapter

of the Upāsanāsārasamgraha.68 The (B2) recension probably reached its present

form by shortly after the mid-seventeenth century.

The (B2) text begins with a forty verse discourse on the om mantra,

combining (B1) and the Gītāsāra, with visualization meditations focused in the

heart. This investigation of the mantra om is similar to that of the other bindu

Upaniṣads, but it is expanded by the contents from the Gītāsāra. The second half of

the text is devoted to haṭha yoga and is largely quoted or paraphrased from chapter

6 of the Vivekamārtanda (variant of the Gorakṣaśataka) and the Haṭhayogapradīpikā

(derived from the Nāth Siddha texts). The Dhyānabindu Upaniṣad provides some

details on a six-limbed yoga path (in the style of Gorakhnāth, as derived from the

68 Bouy (1994), pp. 34, 86–92, 135, 142, 143. See Bouy for exact versification and minor

variants. The Gītāsara is a BhG inspired dialogue between Śrī-Bhagavant (Kṛṣṇa) and Arjuna on the

sacred syllable, om. It was probably written between 1600-1650, and exists in variant recensions.

The sixth chapter of the Vivekamārtanda of Viśvarūpadeva is a variant version of the Gorakṣaśataka.

The Haṭhayogapradīpikā was compiled from previous (Nāth) sources by the middle of the fifteenth

century. The Upāsanāsārasamgraha is an anthology of twenty-four chapters compiled in the

sixteenth to seventeenth centuries. Upāsanāsārasamgraha is also utilized in the expanded versions of

the BVU, YŚU, and VUI.

Vivekamārtāṇḍa) and discussion of kuṇḍalini-śakti and khecarī mudrā, with the

standard descriptions of the nāḍis, prāṇas, and six cakras. Unlike some of the

compilation texts, the editor does little with the expanded sources, copying them

often verbatim from the Nāth sources. There are also several verses devoted to

hamssa meditation.

3.2.7. Hamsa Upaniṣad ("Secret Teachings of the [Sounds of the] 'Wild Gander'")69

This Upaniṣad has both northern and southern recensions, mostly in

sections of numbered prose, and is sometimes called the Hamsanāda Upaniṣad. The

northern (A) recension dates from the twelfth to thirteenth centuries. The

southern (B) recension is similar in length and content but with insertions,

reordering of verses, and variants throughout. Recension (B) probably dates from

after the fourteenth century and may be as late as the seventeenth century. Both

recensions show marked tantric influence in their descriptions of practice. The (A)

recension of the Hamsa Upaniṣad is numbered thirty-eight or forty-two in the

Atharvanic collections. The (B) recension is numbered fifteenth of 108 Upaniṣad

and is associated with the Śukla Yajurveda.

Recension (A) introduces the theory of hamsa via a dialogue between

Gautama and Sanatsujāta. Gautama asks questions about the brahmavidyā (the

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Arcane Science of Brahman).70 Sanatsujāta describes the path of the hamsa as it

was given by Śiva to Pārvati (modeled on Āgamas). The text then describes the

ātman as the hamsa, explaining hamsa as the out-breath (ham) and the in-breath (sa).

Through meditation on om, and especially its reverberation (nāda-bindu), the wind

of the body rises through subtle wheels, called cakras. The text (section 3)

describes the six standard cakras and the brahmarandhra (as the seventh location,

above the six cakras).71 Kundalini is not mentioned by name; breath or wind is

what moves through the body centers. 21,606 recitations of hamsa are

recommended, and this "king of mantras" is described as having hamsa as the poet,

pankti as the meter, paramahamsa as the deity, ham as the seed (bija), sa as the power

(śakti), and so'ham as the stem (kalaka). The text describes the seven mantras to be

recited and projected onto the body parts starting with the heart and the others

(head, hair-tuft, armor, three eyes, and weapon), and then "also laid on the

hand."72 Next the ātman is described as a bird (the hamsa), with fire and moon as

69 Translations: (A) recension, Deussen (1997), pp. 717-721; (B) recension, Aiyar (1997), pp.

161-163; (B) recension, Varenne (in French), (1971), pp. 110-115.

70 See also the summary of the BVU above for similarities between these two texts. Some

editions of the HU do not contain the introductory frame dialogue. In (A) summarized here,

Deussen's section numbering is followed. The manuscripts differ as to division and organization,

so Deussen separated the text logically into eleven sections.

71 The HU is the only one of the northern recensions that includeds the six-plus-one

standard system of the cakras: [mūla]-ādhāra, svādhiṣṭhana, and the rest as well as its use of the term

cakra. Without particular understanding of cakras, nor knowlege of the subsequent century of

tantrism scholarship, Deussen judged sections three through five as insertions into the text from a

later period. These verses show the most marked tantric influence.

72 Deussen suggests that these mantras are written on a diagram (yantra ?) to be worn as an

amulet and is also drawn on the hands. See Deussen (1997), pp. 717-719, for the mantras spoken

146

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

the wings and om as the head. See the similar characterization in the Nādabindu

Upaniṣad, discussed below. The text continues with a long description of the heart

lotus (eight petals associated with actions and the states of waking, dreaming, and

so on). Hamsa merged with the reverberation is described as beyond the fourth

state, a characterization that is also found in other Yoga Upaniṣads and the Tantras.

The text discusses hamsa theory relative to the ten levels of the

manifestation of inner sound (nāda): bells, drums, thunder, and so on. The focus

of the text is on the manifestation of inner sounds resulting from meditation. The

yogin is encouraged to focus solely on the tenth sound (thunder) in which he

becomes Brahman. The subsequent metered verses redescribe the sounds as a

series of meditation experiences that lead systematically from one to ten, although

the preceding text privileges the tenth. This sonic yoga has Upanisadic models in

the Chāndogya Upaniṣad (2.22) and Maitrāyaṇīya Upaniṣad (6.22). The examples in

the Hamsa Upaniṣad may derive from an unnamed source cited by Jayaratha and

other tantric sources, possibly through intermediary texts.73 Haṭhayogapradīpikā

4.85ff includes a similar (but not identical) list but does not privilege a thunder

sound. With opposite emphasis, it presents the sounds as a continuum toward

greater subtlety of sound until the sounds are dissolved in the fading nāda. As

onto the body: sun in heart, moon in head, etc. Deussen may be correct in this analysis, however

the text reads as if describing nyāsa the practice of enlivening a yantra, image of a god, or (in this

case) the human body of a yogin. On nyāsa, see Brooks (1990), pp. 59f and notes.

147

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

will be discussed below, these Haṭhayogapradīpikā verses are quoted by the (B)

recension of the Nādabindu Upaniṣad.

Like (A), the southern (B) text has twenty-one prose sections that present a

condensed version of breath control theories associated with hamsa

conceptualizations and practice. The (B) text differs in several details: the

dialogue is between Gautama and Sanatkumāra,74 avyaktā gāyatri meter is

mentioned instead of paṅkti meter, and the southern text (B) emphasizes

Sadāśiva and Śakti instead of Parabrahman. Other than these distinctively tantric

references, the southern edition is essentially the same as (A). For more detail as

to the symbolism of the gander (hamsa), see Brahmavidyā discussed in section 3.2.4

above.

3.2.8. Kṣurikā Upaniṣad (“Secret Teachings of the Razor [of Concentration]”)75

This text has both northern (A) and southern (B) recensions. Both (A) and

(B) are similar in content and length. The (B) text contains elaborations that

incorporate certain details from kuṇḍalini yoga in the text, and it mentions that the

yogin who succeeds in the meditations described herein becomes an adept of

73 In Jayaratha's commentary on the Tantrāloka (TĀ 5.99, comm. vol.3, p. 410) he cites a text

that enumerates ten mūdas, of which the tenth grants liberation. See Padoux (1992), p. 97, § 31.

74 Sanatkumāra is a synonym of Sanatsujāta, so the same son of Brahmā is the likely

referent.

75 Additional translations: (A) recension, Deussen (1997), pp. 671–675; (B) recension,

Varenne (in French) (1971), pp. 106–109; Georg Feuerstein (1998), pp. 434–437.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

aṣṭāṅga yoga (eight-limbed yoga). This text is first attested in the northern tradition

between the ninth and eleventh centuries. Both texts are composed of twenty-four

ślokas, that view the practice of concentration as a razor or knife (kṣura) that severs

the knots of ignorance. Kṣurikā Upaniṣad is numbered fourth on the traditional

Atharvanic lists. The southern (B) text is numbered thirty-first of the 108

Upaniṣads and is attached to the Krṣṇa Yajurveda.

The Kṣurikā Upaniṣad instructs the yogin to use the kṣurikā form of dhāraṇā,

in which concentration acts as a razor. The yogin who practices razor-

concentration is never reborn. The text suggests the standard yogic prerequisite of

silent places and proper postures, which are not elaborated. These verses are

expressed by Svayambhū (Śiva, or a sage of this name). The limbs are drawn in

like a tortoise, and the manas is locked in the heart. Om is pronounced gradually

according to the twelve phones (Cf. Nādabindu Upaniṣad (A) verses 8-11 for the

twelvefold division of recitation of the om mantra).

The yogin next fills the body with breath and closes all of its doors (body

openings). The yogin then fixes the breath (made steady by the phones of om) in

different body parts (toes, ankles, and so on) a prescribed number of times (saying

the mantra while breath is fixed in the particular body part for two to three

rounds). After moving the breath through the body, it enters the navel area,

which is the location of the Suṣumnā. The breath should then slip into this subtle

channel and move upward, after which it moves to the heart and the neck. There,

the razor-manas cuts or shaves off "names," "forms," and all the lower parts of the

body.

The technical language used throughout is that the razor cuts the "joints,"

"vital organs," or "vital points" (marman). This term is used in Āyurvedic

literature for the joints and vital organs of the body. The most important marman

of the medical texts are described as those places where vitality is concentrated.

Vital organs are areas that, if injured, bring pain and often death. The early

recension of this text is relying on medical language instead of the cakra or padma

theories of the Tantras. The image of the text is that the nāḍis are knotted at the

marman areas, and these knots must be cut with the sharp blade of the

concentration razor.

The razor cuts off or splits everything except for the central nāḍi. Within

the text, the suṣumnā is referred to as taittilam, "the support" or "the pillow" of the

nāḍis.76 The metaphor follows that the yogin is to "stuff the pillow" with all states,

good and bad. With the "pillow" stuffed full, the breath cannot fall back into the

body, and the yogin escapes rebirth. The remaining verses contain metaphors of

liberation: a bird freed from its tether, flame burning all of its fuel, and so on. The

76 Deussen reads taittilam as pillow (or "support") and then shows how the metaphor of

stuffing the pillow follows. Feuerstein reads taittilam as "rhinoceros" and states that the meaning

History, Text, and Context of the Yoga Upanisads: Chapter 3

razor is pointed with breath-control, sharpened with the phones of om, and

whetted on renunciation. Thus, the razor cuts the cord and the yogin, like a bird,

flies free.

The terms and symbols for inner meditation practices are simple and

without the poetic elaboration characteristic of kundalini yoga. Here, as in many of

the other earliest Yoga Upanisads, concentration moves prana (via the mantra om)

through the body, then into the primary artery or vein and up through the body.

This is the most detailed and pointed of the early texts as to the actual practices of

meditation. Similar meditations are still taught as part of hatha yoga meditation yet

today. Although the nadis are important to this text, it is not really a kundalini

tract. As such it represents the widespread (and older) notion of subtle arteries

without the elaboration found in the Tantras and hatha yoga texts.

3.29. Mahavakya Upanisad ("Secret Teachings of the Great Utterance")77

This is a South Indian mantra yoga text of twelve slokas expounding om and

hamsa theory. There are no known earlier recensions for this text. The Mahavakya

Upanisad is associated with the Atharvaveda and numbers ninety-two on the list of

and usage in the text is "esoteric" or unclear. Upanisad Brahmayogin's comments indicate that he

did not understand its usage here either.

77 Additional translations: (B) recension, Varenne (in French), (1971), pp. 128-131.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

108 Upaniṣads. The text is attested by the seventeenth or early eighteenth century.

Authorial date is unknown.

The text contains little practical information and is primarily a simplistic

Vedānta tract that praises the hamsavidyā. This "Arcane Science of the Wild

Gander," is the same mantra yoga encountered in several other Yoga Upaniṣads.

Practice focuses on the recitation of hamsa. Specifically, this text states (verses 5-6)

that "hamsa connected with the ajapāgāyatri,78 that 'I am.' That is capable of being

acquired, by prāṇa and apāna flowing inward and outward from opposite

directions in this manner, by uttering the hamsamantra." Unlike many yoga texts,

the Mahāvākya Upaniṣad lauds the condition of aikya (identity, oneness) with

Brahman. It considers siddhis (yogic psychophysical powers) and samādhi (enstatic

absorption) as inferior to oneness and does not use the more common term used to

refer to a state of unity, kaivalya / kevala ("unity, isolation, aloneness").

Verses 1-10 are generic Vedānta, with no sectarian references. Verse 11

contains a meditation saying, "I am Śiva. I am the brilliant white radiance." Verse

12 praises study of the Vedas and concludes the text declaring that attainment of

oneness with Mahāviṣnu is the goal of study. The Mahāvākya Upaniṣad has some

affinity with the expanded (B) text of the Tejobindu Upaniṣad (3.2.14 below).

78 The "unpronounced mantra," that is the "mantra" that consists of inhalations and

exhalations. Ajapā and ajapāgāyatri are used as technical terms in yoga literature to stand for the

hamṣa that is created from breathing.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

3.2.10. Maṇḍalabrāhmaṇa Upaniṣad (“Secret Teachings of the Doctrine-Wheel of [Five] Brāhmaṇas”)

This is a South Indian text of five sections, termed Brāhmaṇas, mostly in prose. Another text of the same title reproduces parts of the Śatapatha Brāhmaṇa.

Since 1300 some commentators have cited Mandalabrāhmaṇa Upaniṣad text derived from the Śatapatha Brāhmaṇa causing confusion regarding the dating of the yogic Maṇḍalabrāhmaṇa Upaniṣad.

There does not appear to be any relationship between the long reproduction-text based on the Śatapatha-Brāhmaṇa and the yogic text of the same name.

The yogic Maṇḍalabrāhmaṇa Upaniṣad of the Yoga Upaniṣads is not attested until the formation of the South Indian corpus of 108 Upaniṣads (seventeenth century).

Thus, the yogic Maṇḍalabrāhmaṇa Upaniṣad is a compilation text, like the other late southern Upaniṣads, that has no measurable claim to antiquity.

In the southern list of 108 Upaniṣads, the Maṇḍalabrāhmaṇa Upaniṣad is associated with the Śukla Yajurveda and is listed as number forty-eight.

The Maṇḍalabrāhmaṇa Upaniṣad appears to be a combination of several different texts (sections 1-2, and sections 3, 4, and 5): among its source texts are the Rājayoga Bhāṣya and the Siddhasiddhāntapaddhati.

Additional translations: (B) recension, Aiyar (1997), pp. 185-193.

Bouy (1994), p. 107. The RYB of Sadanandavadhuta, or his lineage (often wrongly attributed to Śaṅkara), is medieval but of unknown date. The SSP attributed to Gorakṣa was probably written before or around 1600 to 1650. [Although this the RYB is clearly employed in this recension of the MBU, I have not been able to access the text of the RYB to assess its possible relationship to the related text, the ATU.]

153

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

material with another southern Upaniṣad the Advayatāraka, incorporating most of

the content of the Advayatāraka into its first two sections. It is not clear whether the

Advayatāraka Upaniṣad was a source for the Mandalabrāhmaṇa Upaniṣad, whether

they possess a common source, or whether the Advayatāraka is a fragment or

extraction from the Mandalabrāhmaṇa Upaniṣad. It seems most likely that the

Advayatāraka is a source for the Mandalabrāhmaṇa Upaniṣad. Bouy suggests that the

Mandalabrāhmaṇa Upaniṣad was probably written by the redactor of the 108

Upaniṣads.⁸¹ The Rājayoga Bhāṣya is the source of its inspiration but with original

elaboration. The text was probably compiled in the late seventeenth to early

eighteenth century.

Aiyar explains the title as follows: The Puruṣa in the sun (Nārāyaṇa, a name

for Viṣṇu) reveals this text to an Upanisadic sage named Yājñavalkya in the

celestial realm of the sun (ādityaloka). It is as likely that the title derives from

structural layout of the work in five sections, or Brahmaṇas. These five sections are

the spokes of a wheel of doctrine, or maṇḍala. (Maṇḍala means wheel, disk, and

circle, as well as other specialized meanings.) Aiyar’s interpretation is consistent

with the text but probably elaborates on the practical title: “Secret Teaching of the

Wheel of Doctrine with Five Spokes [five sections].”

⁸¹ Bouy (1994), pp. 107.

154

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

In its first two sections, the text includes many of the meditation practices

described in the Advayataraka Upaniṣad (see above, 3.2.1). The Mandalabrahmana

Upaniṣad places this discourse on taraka yoga within the context of an eight-limbed

yoga system thaat is identical with the classical formulation. The goal of practice is

to realize the lights and signs through internal vision. Accomplishment in these

practices leads to oneness with brahman and the experience of being,

consciousness, and bliss (sat, cit, ananda).

The text describes various visions of colors and lights and ties this

discussion to the five states of consciousness. (Classical Upaniṣads, as discussed

earlier, generally include four states.) These states are described as waking,

dreaming, dreamless sleep, the "fourth state" (that which is beyond dreamless

sleep), and a state "that is beyond the fourth state."82 The last three sections

present mystical philosophy, describing different states of "aloneness" (kaivalya)

and transcendence according to Vedantin metaphors and categories.

Overall the text has the character of a compilation of various practices and

doctrines from the Advayataraka, the Rajayoga Bhaṣya, the Siddhasiddhantapaddhati,

and the corpus of older Upaniṣads. The fourth section insists that the yogin must

cognize the nine cakras, the six adhāras (meaning unclear), the three lakṣyas ("signs"

82 "Beyond the fourth state" is not usually found in the Upaniṣads, as the fourth state is

beyond qualifications. This term appears in the texts of Kashmir Śaivism (for example, in the

155

of tāraka yoga) and the five ākāśas (space, void, ether) associated with the highest

states of consciousness. The text discusses lakṣyas and ākāśas throughout. Cakras

and ādhāras are not further explained or discussed. This section is inspired

directly by the Siddhasiddhāntapaddhati (II.30-31), from the passage on

vyomapañcaka, "the fivefold space." The lack of explanation in the text concerning

the cakras and ādhāras appears to be due to the author borrowing this small

fragment without establishing its context or other elaboration. Prose verse 7 of the

Advayatāraka contains a similarly inspired passage (to Brāhmaṇa 4) concerning the

five ākāśas, suggesting that it too was inspired in part by the Siddhasiddhānta-

paddhati.

3.2.11. Nādabindūpaniṣad ("Secret Teachings of the Sound Point")83

The text exists in both northern (A) and southern (B) versions. The

northern recensions (A) contains nineteen to twenty ślokas.84 Bouy suggests that

the northern Atharvanic redaction of this text was contemporaneous or after

Śaṅkara (eighth or early ninth century) and before Śaṅkarānanda (ca. 1300).

Therefore, he places the northern recension as written between the eighth and

Mahārthamañjari, Tantraloka), cf. Mark Dyczkowski (1987), pp. 206-207, 213, 215. For Kṣemarāja's

comments see Dyczkowski (1992), pp. 135, 140. See also Padoux (1992), p. 143, passim.

83 Additional translations: (A) recension, Deussen (1997), pp. 683-686; (B) recension, Aiyar

(1997), pp. 194-198.

84 The two northern recensions ĀSS (Poona) and BI (Calcutta) are virtually identical but

with different versification.

156

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

thirteenth century.85 In the Atharvanic lists, the (A) recension of the Nādabindu Upaniṣad is number seventeen.

The (A) recension of the Nādabindu begins with a characterization of the ātman as the hamsa: the sound a is the right wing, the sound u is the left wing, the sound m is the tail feathers, the half-sound (nādabindu) is its head. The following verses continue such homologies between the parts of the hamsa with the guṇas, dharma, adharma, and various realms—bhūr-loka, bhuvar-loka, and so on—ultimately locating the satyaloka at the forehead between the brows (verses 1-6). This hamsa carries the yogin up to the heavenly realms (invoking Atharvaveda 13.3.14 via partial quote).

Verse 6b starts a new set of equations, associating the a, u, m, and half phone with Agni, Vāyu, Sūrya maṇḍala, and Varuṇa, respectively. The three and one-half phones (described as four phones or morae in verses 8-11) each are said to have a threefold aspect.86 Thus, based on the four parts—a + u + m + nadabindu

85 Bouy (1994), p. 31.

86 The twelve partial moras (mātrās) are described as twelve objects of meditation (dhāraṇās): ghoṣinī, vidyunmālī, paṭaṅgi, vāyuvegini, nāmaḍhyeya, aindrī, vaiṣṇavī, śāṅkari, mahatī, dhruvā, maunī; brahmi. All of these divisions/objects of meditation are expressed as grammatically feminine. There is not enough evidence within this text, but these may be read just as adjectives, "rich in sound" or "wreathed with lightening." Alternatively, these types of terms are constantly used for goddesses in the Śakta Tantras. Goddess as focus of meditation inside the body and as associated with phonemes is well developed in the Tantras, such as the Kubjikāmata Tantra and the Saṭsāhasra Saṃhitā (SSS). For discussion of the feminization of phonemes, the goddesses, and the meditation on both, see Padoux (1992) and Dory Heiligers-Seelen (1994), both throughout. Heiligers-Seelen's KMT employs several of these terms and cognate terms for the goddess in the circle of female messengers (Dūtīcakra), pp. 69ff. There is not much evidence to

157

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

(even if the last is actually a half measure)—being threefold, the mantra om has

twelve parts. (The twelvefold phone or mora division of om is also discussed in

the Amṛtabindu, verse 23).

Verses 12-17 explain the rebirth the yogin will achieve if he dies while

meditating on any of the twelve phones of om. If meditating on the first part, he

will be rebom as an emperor in the land of Bhārata (India). Thereafter the the

yogin’s rebirths are elaborated: birth as a Yakṣa, as a Vidyādhara, and others. The

best meditation is on the twelfth part of the mantra that is beyond sound, in the

eternal light of Brahman (Brahman is described as sadoditam = sakṛdvibhātam; cf.

Chāndogya Upaniṣad 8.4.2).⁷⁷ Verses 18-20 summarize how the mind (manas), free

from the senses and guṇas, dissolves itself—this is yoga. The text emphasizes the

gradual nature of this meditation series, saying: “He is gradually liberated from

the body . . . into the highest bliss.” Moreover, these verses specify that these

meditations affect one’s state at the time of death (cf. Praśna Upaniṣad 5). Thus, the

twelvefold practice of meditation is reminiscent of the classical Upaniṣads in

espousing liberation—or another positive rebirth or result—at the time of death,

not a state attained during practice while alive.

support that this passage of the NBLU is influenced by the Tantras or tantric systems, but the

resemblences are noteworthy.

⁷⁷ See Deussen’s introduction to the text, (1997), pp. 683-684. The meaning of Sadoditam is

somewhat unclear, but it means something like “divided in the place of the sacrifice,” which is

pobably a reference to Brahman as Puruṣa. Sakṛdvibhātam means “appeared at once.”

158

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The (B) recension includes equivalents verses to the twenty ślokas of (A) but

with enough variance to suggest that a different manuscript tradition from the

extant (A) recension was transmitted in South India.88 The southern texts contain

fifty-three to fifty-six stanzas on a Vedānta style of nāda yoga (sound yoga). The

southern Adyar Library Series recension quotes the Aparokṣānubhuti (attributed to

Śaṅkara) and the Haṭhayogapradīpikā. All of the additional materials (except verse

30. derive from these two texts, with most verses (twenty-five verses) drawn from

chapter 4 of South Indian recensions of the Haṭhayogapradīpikā.89 Thus, the

expanded South Indian text dates to after the fifteenth century. It is probable that,

like the other expanded texts, the expansion of the Nādabiṇdu Upaniṣad occurred in

the seventeenth or early eighteenth century. In the list of 108 Upaniṣads, the

Nādabiṇdu Upaniṣad is connected to the Rgveda and is listed as number thirty-eight.

The southern text follows the same pattern as (A), but with considerable

elaboration. The inner sound transports the yogin beyond the realm of ordinary

senses. The mantra om is described and discussed in terms of all of its sound

88 There are various different wordings and word order as well as different technical terms.

For example, the numbers ten and eleven of the mātrās/dhāraṇās are dhṛti and nāri in place of

dhruvā and muni. Other insertions include calling the vidha mukti (disembodied liberated one,

who practices this nāda yoga) the lord of Kamalā, another seemingly Āgamic or Tantric phrase.

That the primitive southern recension differs from (northern A), is also observed by Bouy’s (1994)

in his comments on the NBLU. See Bouy’s comments throughout, for suggestion that there

independent, early southern variants on the northern (A) texts.

89 See Bouy (1994), pp. 92-95. The Aparokṣānubhuti dates from between the eighth to

thirteenth centuries. The Haṭhayogapradīpikā dates to the mid fifteenth century, the southern

manuscripts quoted in the NBLU date to after the mid fifteenth century, since the Haṭhayogapradīpikā

was originally written in north India.

History, Text, and Context of the Yoga Upanisads: Chapter 3

elements: a, u, m, and the nasalized echo of the m, signaled in written Sanskrit by

the bindu or dot. These different sounds are described in terms of mātrās (phones,

morae, or measures). For the southern text, mātrā is invested with the valence of a

measure of time. Mātrās of different lengths are associated with different states of

consciousness and meditation.

The latter portion of the (B) recension reports various inner sound effects,

heard in meditation. These verses, 32 to 38, are quoted directly from the

Haṭhayogapradīpikā (4.83-89). The wording and subjects are similar to comparable

passages in the Hamsa Upanisad (discussed above), although the Hamsa Upanisad

does not list the same sounds. The goal of this nāda yoga is disembodied liberation

(videhamukti), as already established in the (A) version of the text.

3.2.12. Pāśupatabrahm a Upanisad (“Secret teachings of the Brahman, Who is the

Lord of Animals”)

To date, Pāśupatabrahm a Upanisad is only known from the southern

collections. It has seventy-eight verses in two sections—pūrva kānda and uttara

kānda—that discuss mantra yoga as a practice of inner sacrifice. The southern text

appears as number seventy-seven on the list of 108 and is attached to the

Atharvaveda. The two sections read, in terms of subject matter, style, and

vocabulary, as if they were originally two separate texts. Recitation of the hamsa

160

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

mantra is described in Vedic terms as the inner sacrifice. The second section is a

summary of various points of non-dual Vedānta philosophy that suggests that

knowlege (jñāna) of the nature of ātman as haṃsa is all that is necessary for

liberation.

The pūrva kāṇḍa of the Pāśupatabrahmā Upaniṣad begins as a discourse given

by Brahmā (called Svayaṃbhū) to his self-begotten progeny, Kāmeśvara and

Vaiśravaṇa (who is also called the Vālakhīlya). The Svayaṃbhū, who is Brahmā,

states that the vidyā (arcane science) of all the world is the phones of the Sanskrit

sound system, or the letters of the Sanskrit alphabet. Of the phones, or letters, of

the Sanskrit sound system, the oṃkāra of four measures is especially important,

and is called "the deity who is the very life of Brahman."

Next follows a long list of powers under the control of Brahmā: yugas (ages

or time cycles), day, night, sun, moon, stars, and so on. Gods of the Vedic

sacrifices are next enumerated. The haṃsa mantra is described as the inner sacrifice

and the paramātman is equated with the haṃsa. Further sacrificial homologies are

enumerated: yajñasūtra, the sacrificial thread, comprises ninety-six threads, and

thus the esoteric brahmasūtra is the thread of ninety-six tattvas. An initiated

brahmin is the only person allowed to know the secret of the brahmasūtra and thus

the only one who can perform the inner sacrifice. Further connections and

corollaries are given, and the praṇava and haṃsa mantra are praised as the only

161

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

means to Brahman. The first section closes with a summary: with Vaiśravaṇa,

Rudra also understood from the mouth of Svayambhū thus, "The final resort of

the Brahma Upaniṣads, I am that praṇava that is the tāraka (the liberator), the

radiance of hamsa, and the Paśupati."90

The uttara kāṇḍa explains (without narrator or frame dialogue) various

points of Vedānta philosopy: Brahman attained through akhaṇḍa vrtti, the world as

māyā, and so on. The text also stresses silence and disregards the importance of

diet and other restrictions in favor of jñāna. The work employs a variety of Vedic

and Śaiva terms, but its general style is that of post-Śaṅkara Vedānta. Overall, the

text has only few passing references to yogic practice. It praises the praṇava hamsa

practice but does not describe it or give details. It does state (2.36-43) that

knowledge of Brahman causes the "knots (grantha) of the navel, the heart, and so

on, to give way." It generally reads as a pastiche of praise of hamsa practice and

Vedāntin jñāna yoga generalizations that has antecedents in the Āgama texts of

Śaivism.

90 Verse 32. See also Ayyangār's translation, heading "The Radiance of the Tārakahṛṣd"

(1952), p. 189. See entries on the ATU and MBU, here, for other use of the term tāraka.

162

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

3.2.13. Śāṇḍilya Upaniṣad ("Secret Teaching to the Sage Śāṇḍilya")91

This South Indian text in three chapters presents a dialogue between the

sages Śāṇḍilya and Atharvan, in which Śāṇḍilya asks questions and Atharvan

answers. This text follows the model of the conversations between Upanisadic

sages and their students or the dialogues between Śiva and Śakti in the Āgamas

and Tantras. The Śāṇḍilya Upaniṣad is a late compilation text with no known

northern recensions. It is number fifty-eight in the list of 108 Upaniṣads and is

associated with the Atharvaveda.

Chapter 1 presents an eightfold yoga system in eleven sections (kāṇḍas). The

southern compiler of chapter 1 drew most of its contents from earlier non-

Upanisadic sources: the Yogayājñavalkya, the Haṭhayogapradīpikā, the

Laghuyogavāsiṣṭha, the Uttara Gītā, and the classical Yoga Sūtra.92

The compiler does not simply quote blocks of text from previous yoga

literature. He selects materials intentionally, reordering it in many places, to

present his own formulation. He manipulates and alters the source materials to

present his own sectarian position. For example, in Śāṇḍilya Upaniṣad 1.7.14-16,

91Additional translations: (B) recension, Aiyar (1997), pp. 132–146.

92For an in-depth discussion of sources, see Bouy (1994), pp. 95–98. The Yogiyājñavalkya

dates from between the tenth and early fourteenth centuries and was written in south India with

Vaiṣṇava sectarian orientation. The Haṭhayogapradīpikā is a north Indian text that dates from the

fifteenth century and derives much of its content from the works attributed to Gorakhnāth

(Gorakṣa), especially the thirteenth century Gorakṣaśataka and Amarāughaprabodha (date unknown

but before 1450). Feuerstein states that the Kashmiri LYV dates from the tenth century, Bouy

163

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

when quoting from the Haṭhayogapradīpikā (4.36-37), the compiler replaces the

Haṭhayogapradīpikā Śaiva terminology with his own Vaiṣṇava term: in the place of

śāmbhavī mudrā, he inserts vaiṣṇavī mudrā. Elsewhere (1.7.38) he replaces the

recommended liṅga worship (Haṭhayogapradīpikā 4.42) with worship of the image

of Viṣṇu. There are several other occasions when he inserts Vaiṣṇava references

and terminology where the original materials are Śaiva. Otherwise, the text is

technically true to its sources, expounding a composite haṭha yoga for Vaiṣṇava

brahmins. The first chapter has standard haṭha yoga practices, as derived from

these sources, with considerable space devoted to the nāḍīs and prāṇāyāma. There

are detailed descriptions of meditation methods and techniques and a long list of

psychophysical powers (siddhis). The primary technique cultivated and praised in

this text is the khecarī mudrā (the seal of the aviator). This mudrā is described as a

series of methods that result in stilled inner vision fixed behind the eyes.

Chapters 2 and 3 are lengthy discussions of Vedānta metaphysics. These

two sections draw much of their inspiration from the classical Upaniṣads. Drawing

on the aforementioned yoga texts, this text attempts to synthesize Patañjali’s

system (via the Yogayājñavalkya) with later traditions. In chapter 2 Atharvan

instructs Śāṇḍilya that Brahman is satya, vijñāna, and ānanda. He teaches that vāc

and manas are not enough to reach Brahman. Instead, jñāna and yoga are

suggests that the LYV dates from approximately 1350. The date of the UG is unknown. The YS

164

necessary.93 Afterward the chapter includes a long series of synonyms and

metaphors for ātman and brahman and promises that knowledge of ātman will

relieve all pain and suffering.

Chapter 3 continues the dialogue on Brahman. From formless

Parabrahman comes three forms: niṣkala, sakaa, sakala-niṣkala (undivided, divided,

both divided and undivided). Several adjectives of negation—not definable, not

mortal, and so on—characterize niṣkala brahman. Maheśvara and

Mūlaprakṛtimāyā together are the manifestations of sakala-niṣkala brahman.

Dattātreya is the form of sakala brahman. The text offers further characterizations

and etymography94 of Brahman as ātman, Maheśvara, and Dattātreya. Although

chapters 2 and 3 contain no specific references to yoga practice, the metaphysics

described is consistent with the synthesis of chapter 1. The Śiva-Śakti pairing of

Maheśvara-Mūlaprakṛtimāyā is reminiscent of both the tradition of the Yoga Sūtra

and that of the Tantras. Dattātreya is a divinized yogin-teacher, especially

dates from the second century. The LYV is also a source of the YYV.

93 Some manuscripts read only jñāna, leaving out "and yoga." This omission hints that

chapters 2 and 3 might also have been a separate text (or are derived from older texts), to which the

yoga section was appended. To maintain consistancy, a later editor added yoga with jñāna.

94 I adopt this term from linguist, Probal Dasgupta, who characterizes the esoteric science

of "creative" etymology as a conscientious science separate from normative linguistic etymological

analysis. The Indic scholar-pandits of the Sanskrit grammar created a systematic science of

linguistics and were well aware of the normative rules of etymological analysis. The widespread

use of "creative," "false," or "folk" etymology was not for them uncritical thinking or poorly

performed linguistics. It was instead a separate "esoteric" science of creatively breaking apart

words and word-roots for the purpose of novel characterizations and numinous inspirations.

Personal communication (August 2001). Etymography likely played a role in memorization

worshiped in western India and South India.95 Dattātreya is widely portrayed as

an avatār of Viṣṇu, although he is more prominently revered as a yogin-sage.

Śaiva traditions also claim him, especially in this latter character.

The author of the text considers bhakti and pūjā as preliminary to yoga

practice, as evidenced by multiple Vaiṣṇava sectarian references and the image

worship described. Based on the dates of its sources, the Śāndilya was certainly

compiled after the fifteenth century. It is possible that, like the Mandalabrāhmaṇa

Upaniṣad, this text was written by the redactor of the collection of 108 Upaniṣads.

In this case, the text would date from the late seventeenth to early eighteenth

century.96

3.2.14. Tejobindu Upaniṣad ("Secret teachings of the radiant point")97

The Tejobindu Upaniṣad has both northern (A) and southern (B) recensions.

The northern recension of fourteen ślokas is first attested in the twelfth to

thirteenth century. It is twenty-first on the lists of Atharvanic Upaniṣads.

techniques as well. Remembering large blocks of text verses or remembering a particular dogmatic response or argument is often "keyed" by a specialized interpretation of roots or phonemes.

95 Dattātreya is especially worshiped as a god in Maharashtra (incorporating imagery from Brahmā, Śiva, and especially Viṣṇu). He also has extant cult in Kerala and other parts of south India. At least one historical Dattātreya authored works on yoga and tantra during the thirteenth century. The fourteenth century Śāriṅgadhara paddhati classifies the two major forms of yoga as Gorakhnāthi (six-limbed) and 'that of the son of Mṛkaṇḍa (Dattātreya) (eight-limbed). While there were two systems recognized (six- and eight- limbed).

96 See Bouy (1994), pp. 45-46.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Recension (B) is greatly expanded, containing 465 stanzas of mixed prose and

verse in six chapters. The South Indian recension is attached to the Kṛṣṇa

Yajurveda and is counted thirty-seventh in the list of 108 Upaniṣads. The southern

recension, like the other expanded texts probably dates to the seventeenth or early

eighteenth century. The sources of most of the added material are the

Aparokṣānubhūti and the Ṛbhugītā.⁹⁸

The northern (A) recension of the Tejobindu Upaniṣad is similar the other (A)

recension bindu texts, but is less clearly devoted to yoga and meditation.⁹⁹ It may

even take its title from tejobindu being its first word (like Kena and Īśā).¹⁰⁰ It is a

brief text and, with the exception of verses 1a and 4b, presents a series of negations

and opposites that characterize Brahman, which is beyond duality and all

attributes (thinkable/unthinkable, empty / not empty, and so on). The fourteen

verses of the text sequentially deal with the main features of Vedāntin doctrine:

⁹⁷Additional translations: (A) recension, Deussen (in English by Bedekar and Palsule),

(1997), pp. 705-708; (B) recension, Aiyar (1997), pp. 61-80.

⁹⁸ Bouy (1990a), pp. 114-116. See Bouy’s chart for the breakdown of verses. Summarily,

TBU verses 1.1-14 derive from the primitive (A) version of the text. Verses 1.15-51 derive from the

Aparokṣānubhūti, most of the remaining text of chapters 2-6 are derived or inspired by the Ṛbhugītā

(RG). The former source was also employed in the expansion of the YŚU and NBU.

Aparokṣānubhūti probably dating to between the the eighth and thirteenth centuries. It is often

attributed to Śaṅkara. The Ṛbhugītā is a long text (2900 ślokas) of unknown date, attested in late

medieval south India. There may be further research on this text that better assesses its dates and

location. This issue needs further research.

⁹⁹ Aiyar (1914) actually classes this text as a Vedānta text instead of a yoga text due to its

contents lacking much reference to yogic practice. Others include the TBU with the Yoga

Upaniṣads, either by bundling it with the other bindu titles, in other words because of its title

instead of its contents, or they classify it as a yoga text without explanation.

¹⁰⁰ Cf. Deussen’s introduction to the text. (1997), p. 705.

167

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

verses 1-2, the difficulty of meditation (dhyāna); verses 3-4, the requirements of one

qualified for meditation. Verse 5-8, the place of Brahman as the object of

meditation; verses 9-11, the enigmatic nature of Brahman; verses 12-14, the nature

of jīvan mukti.101 From Deussen's notations on the (A) text and the Telugu edition,

in combination with a review of the (B) recension (Adyar Library Series), it

appears that verses 1-14 of the Tejobindu Upaniṣad in all recensions are heir to an

"enormously corrupt text-transmission."102

Verse 1 describes the tejobindu: "Highest meditation aims at the radiant

point supremely enthroned in the heart: subtle, blissful, radiant, gross, then fine,

then superfine." From the context of the other bindu Upaniṣads, this radiant point

is probably the anusvāra of om that denotes Brahman. However, there is no further

mention of the bindu or any other comment on om or on mantra yoga. Verse 4b

states, "He who steps through the three gates becomes the haṁsa dwelling in the

three worlds."103 This verse may refer to the gates of body through which the

prāṇa (as om) travels in the practice of meditation.

The southern (B) recension contains 465 stanzas in six chapters. The six

chapters appear to be drawn from different sources, chapters 1, 2 to 4, and 5 to 6

101 Deussen's summary from his introduction to the text (1997), p. 705.

102 Deussen (1997), p. 705.

103 Cf. Deussen's translation (1997), p. 706. Deussen notes that the "three gates" are

explained by the commentary as renunciation, patience, and respectfulness toward ones teacher.

Ayyangār (1952) reads the verse as "three stages" without comment. It seems likely that the three

gates are the internal gates of the body mentioned in the ANU 26, although the verse is not clear.

168

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

are self-contained expositions.104 Chapter 1 includes the fourteen verses of the (A)

text, then adds an exposition of a fifteen-limb yoga system, focusing on meditation

as interpreted by non-dual Vedānta metaphysics. The fifteen-limb division

includes the eight limbs of classical yoga with several additions, including tyāga

(renunciation) and mauna (silence).105

Chapters 2 to 4 take the form of a dialogue between Paramaśiva and his

son, Kumāra. Chapter 2 begins with a question from Kumāra regarding the

"indivisible one essence," to which Paramaśiva gives a long list of attributions

(verses 1-23): the world, ātman, karman, jñāna, elements, principles, and so on. The

text presents a doctrine that all things are essentially consciousness. Knowing all

of this, one will know, "I am Brahman." Chapter 3 continues the dialogue with a

new question concerning the nature of ātman. Next follows a series of declarative

statements, "I am . . ."—the form of Brahman, bliss, knowledge, and so on.

Further repetitions of Vedānta non-dual assertions continue until verse 60, where

a list of the cures and results of the mantra "I am Brahman" is enumerated

(destroys sins, faults, and so on). Chapter 4 continues the dialogue with

expositions of jīvan-mukti and videha-mukti, states of liberation within and beyond

104 Although the RG is the sources of all of these chapters, the source itself must be a

compilation of many different currents.

105 yama, niyama, tyāga, mauna, deśa, kāla, āsana, mūla-bandha, deha-sāmya, ḍṛk-sthiti, prāṇa-

samyama, pratyāhāra, dhāraṇā, ātma-dhyāna, samādhi. Adyar Library Series TBLU 1.15-16. This section

is derived from the Aparokṣānubhūti 102-142 (not always by direct quotation), attributed to Śaṅkara.

169

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

the body, respectively.106 The characterizations of the two states involve similar

lists of Vedānta non-dual superlatives.

Chapters 5 and 6 contain a dialogue between the muni Nidāgha and the ṛṣi

Ṛbhu.107 Sage Ṛbhu explains the nature of ātman and anātman. He begins with an

explanation of vāc and Brahman, nāda and kalā, and continues with non-dual

metaphor pairs and negations denoting ātman. He rejects any possibility of the

notion or reality of anātman. The text continues with dozens of further non-dual

metaphors and superlatives.

Unlike many of the expanded Yoga Upaniṣads, this text does not contain any

haṭha yoga materials and is more of a non-dualist philosophical exposition than it is

a yoga text. The unknown sources of the (B) recension are thoroughly Śaiva

throughout. They read as Āgama philosophical texts interpreted through non-dual

Vedānta.

3.2.15. Triśikhibrāhmaṇa Upaniṣad ("Secret Teachings to the Brahmin with Three

Tufts of Hair")

This is a southern text of 165 stanzas of mixed prose and verse. The

Triśikhibrāhmaṇa Upaniṣad appears to be a compilation text of Vedānta, Āgama, and

106 The NBU also employs the term videha-mukta, see above.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

haṭha yoga (particularly the Haṭhayogapradīpikā) sources. It is unknown until the

listing of the 108 Upaniṣads, in which it appears forty-fourth, and attached to the

Śukla Yajurveda. As a compilation and due to its lack of earlier attestation, this text

probably dates from the seventeenth century. In most cases, the direct quotes

found in other expanded Yoga Upaniṣads are lacking, but the compiler of this text

was conversant with several Nāth texts and possibly the South Indian

Yogayajñavalkya as well.

The text begins as a discourse by the Sun (in ādiyaloka) to a Brahmin with

three tufts. It begins by stating that everything that exists is in reality only Śiva.

Appearance of difference follows from the devolution from Brahman to avyakta to

mahat to ahamkāra to the five tanmātras to the five mahābhūtas to the whole world.108

The text further describes the divisions and functions of the forms and phenomena

of creation and the nature of reality and the need for yoga and jñāna.

Verse twenty-one begins an exposition of yoga, briefly describing karma yoga

and jñāna yoga, afterm which it begins a longer enumeration of aṣṭāṅga yoga

(consonant with the classical eight-limbed system) with a non-dual Vedāntin

attitude. When the discussion comes to the subject of āsanas, the text shifts to

107 Ṛbhu is a Vaiṣṇava sage whose non-dualist teachings are mentioned in the Viṣṇu Purāṇa

1.15-16. The source text, the “Song of Ṛbhu” (RG) attributes much of its content to this sage. This

sage also appears in the frame dialogues found in the Varāha Upaniṣad (see 3.2.16).

108 This devolutionary or emanational terms, starting with Brahman, mean the Absolute,

the umanifest seed of all phenomenal reality, eminent cognition, ego consciousness, subtle

elements, and gross elements.

171

History, Text, and Context of the Yoga Upanisads: Chapter 3

sources from the traditions of hatha yoga. The text does not provide direct quotes,

but the āsanas described closely follow the language and order of the first chapter

of the Haṭhayogapradīpikā.

Haṭha yoga sections detail seventeen postures, kundalini, breath control, and

other aspects of esoteric physiology. The descriptions of the nāḍis and prāṇas are

quite detailed and evidence original elaboration on the part of the author. The

presentation of pathways and seats within the body describes animal as well as

human subtle physiology. For example, verses fifty-six to fifty-seven declares, "In

the middle of the body is the seat of śikhi (fire), lustrous like molten gold, [which

is] triangular in shape in the case of bipeds like humans, quadrangular in the case

of quadrupeds, circular in the case of birds, hexagonal in the case of snakes, and

octagonal for insects." The remaining characterizations follow this pattern, with

animals and birds also described.

The discussion of the umbilical knot (kanda) the nāḍicakra of twelve spokes

(probably navel here), and the eight-coiled kunḍalini are reminiscent of the subtle

body as described in the Amaraghāsāsana of Gorakhnāth.109 The following (verses

66ff) verses describe the nāḍis and prāṇas according to a mixture of patterns found

in the Gorakṣaśataka, the Śāṇdilya Upaniṣad (or the Yogayājñavalkya directly; cf.

Śāṇḍilya Upaniṣad above). However, the descriptions contain some original

109 See Silburn (1988), pp. 121ff, for discussion of this text and selected verse translation.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

language such as śubha nāḍi instead of kuhu nāḍi, as well as others. The

descriptions of the locations and functions of the prāṇas closely parallel those

found in the Jābāladarśana Upaniṣad (another late medieval Upaniṣad).110 The

sections of exposition continue throughout the remaining text, seamlessly drawing

from various traditions. The text ends with a decidedly Vaiṣṇava orientation, a

shift from the Śaiva character of the early verses.

3.2.16. Varāha Upaniṣad ("Secret teachings of the Boar")111

The Varāha Upaniṣad is a southern compilation text of 273 stanzas in five

chapters concerning non-dual Vedānta philosophy and the practice of yoga. The

text is ninety-eighth on the list of 108 and is associated with the Kṛṣṇa Yajurveda.

The Varāha Upaniṣad is unknown in the earlier textual tradition, and is attested

only with the compilation of the 108 Upaniṣads. Therefore, it probably dates from

the seventeenth to early eighteenth century. It reads as a compilation text, like the

other late Yoga Upaniṣads. It includes similar content to the Pāśupatabrahmā

Upaniṣad, the (B) recension of the Tejobindu Upaniṣad, and the Triśikhibrahmana

Upaniṣad. It incorporates scattered citations from the Upāsanāsārasamgraha, the

110 Not to be confused with the similarly titled Jābāla Upaniṣad.

111 Additional translations: (B) recension, Aiyar (1997), pp. 167-184.

173

History, Text, and Context of the Yoga Upanisads: Chapter 3

Hathayogapradipikā, and the Svāmaprakāśikā attributed to Śaṅkara.112 It may also

contain materials from the Ṛbhugītā, as this sage is present in its frame dialogues

and the contents are similar in many places to the parts of the Tejobindu Upaniṣad

derived from the Ṛbhugītā.113 These sources account for only part of the contents

of the Varāha Upaniṣad.

Of the five chapters, chapters 1 and 2 are framed with a dialogue between

the sage, Ṛbhu and the Bhagavān, the Lord, in the form of a boar (varāha, a form of

Viṣṇu).114 Chapter 3 has no frame dialogue and only a few sectarian references

("Śiva is all" and a reference to the sun as Hari). Otherwise, the language

concerns the ātman, māyā, and ākāśa of pure Vedānta. Chapters 4 and 5 are

presented as a dialogue between Nidāgha and Ṛbhu, similar in character to the

dialogue between these two sages found in the Tejobindu, discussed earlier. In

these chapters, Ṛbhu explains the nature of jīvan-mukti (liberation while living)

and describes the subtle body and practices of haṭha yoga. The first two chapters

112 Concerning the Svāmaprakāśikā of Śaṅkara and the VU, see Bouy (1990a), pp. 123-124. I

have not discovered dates for the Svāmaprakāśikā. If Śaṅkara actually wrote the text it would date

to the eighth or ninth century. If like many other text attributed to the great non-dualist, it is

written by his school or in his style, then it might date to any time after him during the medieval

era.

113 This is likely based on side by side analysis of the TBU and VU. Bouy (1990a) analyzes

the relationship between the RG and TBU, but does not comment on the presence of the similar

material in the VU. This hypothesis requires further research. Dependence on the RG might

however not be great, as the parts similar to the TBU are only a small portion of the text.

114 For the readers who are following the discussion by referring to Ayyangār’s English

translation of the VU, he translates Varāha as ‘porpoise’ instead of ‘boar.’ I am not familiar with any

such usage of the term. Ṛbhu is a Vaiṣṇava sage who appears in the frame dialogues found in the

Tejobindu Upaniṣad (derived from the Ṛbhugītā, see 3.2.14).

History, Text, and Context of the Yoga Upanisads: Chapter 3

are Vedānta with a Vaiṣṇava tone, whereas the three latter chapters are Vedānta

and yoga with a Śaiva orientation. In general, the whole text has the tone of a

Śaiva Āgama, whereas the Vaiṣṇāva frame is not pursued within the main body of

the text.

In chapter 1, the Lord as boar explains the twenty-four tattvas of the

Vedāntins, the thirty-six tattvas of the Śaivas, and the ninety-six tattvas of the

Sāṃkhyins. Chapter 2 includes several sections espousing Vedānta philsophy: the

nature of ātman, māyā, karma, ākāśa, the truth beyond varṇa and āśrama, and so on.

Verses 2.66-74 include a discussion of the evolution of knowlege and how karma is

overcome (moving through the stages of jñāna to vijñāna to samyaktattvajñāna).

The final section of chapter 2 explains unsteadiness and steadiness of the mind

according to alchemical analogies. For example, the mind is unsteady like

mercury is unsteady; both, when bound, lead to power in this world. The chapter

closes with recommending the nāda only as the object of meditation.

Chapter 3 generalizes in Vedāntin terms and describes the citta as the cause

of all things and states. Chapter 4 provides the discussion of jīvan mukti and

elaborates several analogues and allofoms for the letters of the om mantra

modeled on the discussions of the older bindu (A) texts. Stages of the "seven states

of knowledge" are enumerated. Verses 31-33 equate everything with Śiva (guru,

deva, veda, Self, and so on). Verses 34-42 explain instant liberation and liberation

175

History, Text, and Context of the Yoga Upanisads: Chapter 3

by stages through a story of the birds and the ants. The birds follow the path of

the sage, Śuka, and become satya muktas, liberated beings. The ants pursue the

course of Vāmadeva and after many rebirths become krama muktas, liberated-by-

stages. The Śuka path is attained by one who gains direct intuitive and

experiential realization of his own ātman, whether by samādhi, jñāna, or cognition.

The gradual path is associated with the practice of haṭha yoga and austerity. Both

paths are considered auspicious, one fast and the other slow.

The final chapter is exclusively devoted to haṭha yoga practice. Three verses

(6-9) partially quote and interpret Hathayogapradīpikā 3.56-60. Twenty-five verses

(50-75) derive from chapter seven of the Upāsanāsārasaṃgraha.115 It states that the

body is composed of five the elements (earth, water, and so on.). These elements

circulate and fill five maṇḍalas. For example, the vāyu maṇḍala, by blowing in the

body, causes the 21,600 breaths in a day (a concept found in other Yoga Upaniṣads).

Like other Yoga Upaniṣads, the text calls the practitioner a paramahaṃsa. The text

recommends uḍḍīyāna bandha (quoting part of the verses from the

Haṭhayogapradīpikā, but without the description found in Haṭhayogapradīpikā 3.57,

59. and adds that it is difficult to accomplish.116 The text then states that the yogin

115 Bouy (1994), p. 92. The Upāsanāsārasaṃgraha is an anthology of 24 chapters compiled in

the sixteenth to seventeenth centuries. Upāsanāsārasaṃgraha is also utilized in the expanded

versions of other Yoga Upaniṣads, the BVU, DhBU, and the YŚU.

116 uḍḍīyāna bandha is the haṭha yogic upward-lock practice in which the abdominal muscles

are drawn back and upward, after deep exhalation. Often the practice involves fanning, the

contraction of the muscles by which the stomach is ballooned outward and then contracted

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

should know mantra yoga, laya yoga and hatha yoga, in that order, and asserts that

these three yogas each comprises eight limbs (like the classical formulation).

The text describes each limb through various details: tenfold yama, tenfold

niyama, eleven postures (ten from the Hathayogapradīpikā, adding the cakrāsana and

its description), and so on. In its discussion of prānāyāma, the text inserts material

on the measurements of the body and its limbs that parallels the discussion found

in the Triśikhibrāhmaṇa Upaniṣad (see 3.2.15). Its description of the nāḍīs follows the

Nāth characterizations but lists the fourteen found in the Ṣabaladarśana Upaniṣad.117

The only cakra that is mentioned is the navel cakra where all the nāḍīs meet. The

text describes specific meditations (verses 5.31ff). For example, with breath

control and the use of the praṇava (oṃ), completed with the śrī bija,118 the yogin is to

visualize his ātman as “Śrī bathed in nectar.”119 This meditation is called

forcefully upward into the ribcage and backward toward the spine repeated in quick succession.

Whether under this name or other, this fanning practice is fundamental to many techniques for

raising the kuṇḍalinī and for controlling the suction techniques used by men or women in tantric

sexual yoga. These latter techniques involve addition training in other “locks” and “seals” (bandha,

mudrā). The fanning technique used with the lock is also akin to the nauli or laulikā “rolling”

exercise wherein the abdominal muscles are “rolled” from side to side. Descriptions of these

practices differ in some details from one text to another. In a formal sense, the lock might be

considered a static posture, whereas the rolling might include any variations of the abdominal

kinetic exercises.

117 These verses appear to be compiled from various sources, whereas after enumerating

fourteen veins, the text then talks about them as “the twelve” nāḍīs.

118 Use of the Śrī-bīja is indicated for pushing the prāṇa down to meet the drawn up apāna.

The commentary says śrīṁ is the bīja mantra referred to herein.

119 The ātman is visualized as “śrī.” This refers to the śrī bija, or to the goddess Śrī (Vaiṣṇava

Śakti), or as they can be understood as the same, then both.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

kālavañcana ("cheating time")or āyustambhana ("controlling the duration of life").120

Several subsequent verses give further generalizations about breath control. The

catuspatha bandha is then prescribed. The text designates all these practices as

samputa yoga121 (ending with verse 5.49).

A long quote from the Upāsanāsārasaṃgraha constitutes the remainder of the

text (with the exception of two closing verses that praise the kundalini and reassert

the Vaiṣṇava framework). The quote includes a summary of the standard six

cakras, bindu, nāda, Śiva as bindu, and Śakti as kuṇdalini. There is another summary

of breath control and praise of three haṭha yoga locks (bandhas). Vedhaka yoga is

then recommended for the piercing of the three knots (granthi) of Brahma, Viṣṇu,

and Rudra. Pranava mantra yoga is also recommended through which the

practitioner realizes the bindu, reaches the brahmarandhra, and even goes beyond to

the dvādaśānta.122 Finally, a series of four stages is enumerated for yogins to reach

the state of jīvan mukti.

120 Here the text again employs language that can refer to alchemy as easily as yoga.

121 Samputa yoga means "bowl union" and has several technical uses in tantra. It means

sexual union between male and female partners in tantric ritual. It also designates embedding one

mantra within another mantra. There are additional philosophical meanings as well. In this

context, it appears to be part of the general strategy of interpreting tantric or alchemical terms to

mean processes internal to the physical body during the practice of haṭha yoga.

122 dvādaśānta, "the end of the twelve" refers to at least three different locations in the

Tantras, especially those of Kashmir. The term can mean the limit of the breath flowing in and out

of the nose, the brahmarandra (cranial suture), or the cosmic "thousand petaled" mandala or pīṭha

that exists outside and above the subtle body but is connected to it. See Silburn (1988), pp. 30f,

passim.

178

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The Varāha Upaniṣad is largely a derivative text, yet its compiler took care with organizing its contents. Even with direct borrowing, he alters the text with practical insertions. Although the text incorporates other sources, there is considerable continuity across sections. The compiler's modifications of Haṭhayogapradīpikā verses convey the impression that he was a practitioner of yoga, not just a textual theoretician.

3.2.17. Yogacūḍāmaṇi Upaniṣad ("Secret Teachings of the Crest Jewel of Yoga)

This South India text was probably compiled in its present form by the redactor of the 108 Upaniṣads by employing the Baroda (B1) manuscript of the same name.123 The Adyar Library Series (B2) text has 121 stanzas (and three prose passages preceding verses 72, 74, and 75) that discuss haṭha yoga from a Vedānta point of view. With the exception of verses 1 and 72-84 of the Adyar Library edition (including the prose passages), the content of the Yogacūḍāmaṇi Upaniṣad is found entirely in the Nāth Siddha text, the Gorakṣaśataka (or under the alternate title, Gorakṣapaddhati).124 Thus, the Yogacūḍāmaṇi Upaniṣad is almost entirely a borrowed work from the root texts of the Nāth haṭha yoga tradition. It is forty-

123 Bouy (1994), p. 36-36. Bouy argues that the Baroda manuscript in grantha characters is probably the original manuscript of this text. This short text would therefore by the (B1) recension. It originates in south India and expounds the practices of mantra yoga focusing on om and haṃsa. The Adyar Library Series edition thusly should be seen as the (B2) recension of this text.

179

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

sixth on the southern list of 108 Upaniṣads and is associated with the Sāmaveda.

The (B1) recension's dates are unknown, but it is more primitive than the Adyar

Library Series (B2) text and is similar in many ways to the earlier (A) recensions of

other Yoga Upaniṣads, especially the Hamsa Upaniṣad. Although the exact dates

being unknown, the contents of the (B1) text suggest that it was composed around

the fourteenth century or later. The expanded (B2) text most likely dates to the

seventeenth or early eighteenth century (the date of the compilation of 108

Upaniṣads).

Verse 1 introduces the text by giving the title and declaring the result of this

yoga to be kaivalya. The root verses of (B1) (= verses 72-84 of B2) describe a

detailed science of the praṇava (om), including Vedic homologies, various Sāṁkhya

equations, analysis of phones, the haṁsa mantra, and other details. This root text of

nine verses parallels the northern (A) Yoga Upaniṣads' focus on mantra yoga and om,

but with more distinctly Śaiva Āgama language. When added to these original

verses, the material from the Gorakṣaśataka fleshes out a picture of a six-limbed

haṭha yoga method that prepares the yogin to practice silent recitation of the mantra

om. Many of the same quotations from the Gorakṣaśataka also appear in other Yoga

Upaniṣads (for example, the MBU and the Yogakundali Upaniṣad).

124 Bouy (1994), pp. 99-100. See Bouy's chart comparing the Adyar Library Series edition,

the Baroda manuscript, the Nowotny edition of the GŚ and the Bombay edition (1967) of the GP.

The YCU is comprised of 107 of the 201 verses of the Nowotny GŚ.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

3.2.18. Yogakuṇḍali Upaniṣad ("Secret Teaching of the Yoga of the Coiled One [kuṇḍalini-Śakti]"). 125

Like many of the other yoga texts, the Yogakuṇḍali Upaniṣad has two recensions. In this case, both recensions are South Indian. An early recension of seven stanzas (B1) was known before the later expanded (B2) text of the tradition of 108 Upaniṣads. The brief recension (B1) of seven ślokas is attested by the quotations and commentary of Lakṣmidhara, a sixteenth-century figure important for South Indian Śrvidyā (a Vaidika form of Śākta Tantra). 126 Available evidence suggests that the (B1) Yogakuṇḍali Upaniṣad was not written earlier than the thirteenth century or after the sixteenth century. The (B2) Yogakuṇḍali Upaniṣad is associated with the Kṛṣṇa Yajurveda and is eighty-sixth on the list of 108 Upaniṣads. This latter text was likely expanded with the formation of the canon of 108 Upaniṣads in the seventeenth or early eighteenth century.

The (B1) text offers a summary of kuṇḍalini yoga with limited details. It includes variations on the standardized terminology of the Tantras. This brief text describes the Śakti named kuṇḍalini as radiant and like a lotus thread. She bites the root of the mūla kanda lotus in contact with the hole of the brahmarandhra. Through contractions of the sphincter while sitting in the lotus position, the vāyu and agni

125 Additional translations: (B) recension, Aiyar (1997), pp. 199-208; (B) recension, Varenne under the title, Yogakuṇdalinī Upaniṣad, (1971), pp. 93-105.]

181

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

go upward to the navel. The kuṇdalinī pierces the three knots—brahmagranthi,

viṣṇugranthi, the rudragranthi—and finally pierces the six lotuses. Then Śakti is

happy with Śiva in the thousand-petal lotus, sahasrārakamala. The text then

declares that his is the highest avasthā,127 the only source of bliss. The text does not

use the term cakra, nor does it use the term nāḍi. The impressionistic tone of the

text assumes knowledge on the part of its readers of the meanings of its technical

terms. This short recension generally reads as derivative of the Tantras.

The expanded South Indian text (B2) is not attested before the redaction of

the 108 Upaniṣads, suggesting compilation during the late seventeenth to early

eighteenth centuries. It contains 171 stanzas in three chapters (seven (B1) verses

plus 164 additional verses). This text provides an Advaita Vedānta interpretation

of kuṇḍalinī yoga, and is also entitled Yogakuṇḍalini Upaniṣad. Like many of the

other South Indian texts, the (B2) Yogakuṇḍali Upaniṣad is predominantly a

derivative text. Chapter 1 of the Yogakuṇḍali Upaniṣad reproduces selected verses

from a South Indian recension of the Gorakṣaśataka.128 Chapter 2 is derived entirely

126 Bouy (1994), pp.38ff. Bouy dates Lakṣmīdhara, and his Saundaryalaharistutiovākhyā to

the sixteenth century. See also Brooks (1999), pp. 27ff for portrait of Lakṣmṣdhara and his context.

127 “State,” “stage,” or “condition.” Avasthā has several technical uses in yoga texts. It can

mean stages of practice (HYP 4.69), the Vedāntin stages of consciousness (waking, dreaming, sleep,

and the “fourth”), and stages in the practice of prāṇāyāma.

128 Bouy (1994), p. 102. GŚ is a thirteenth century text. The manuscript used by the author

of the YKU was probably the same as the extant Madras recension. Cf. White (1996), pp. 140ff and

passim, for discussion of the GŚ and other Nāth works attributed to Gorakhnāth.

History, Text, and Context of the Yoga Upanisads: Chapter 3

from a South Indian recension of the Khecarividyā.129 Part of the content of chapter

3 derives from a South Indian recension of the Candrāvalokana.130 The source of the

final verses of chapter 3 is unknown. All of the expanded quotations and

paraphrases add more specific details to the practice of raising the kundalini

through posture, breath, and meditation. Like many of the other expanded texts,

the meditation practices associated with the "seal of the aviator" (khecari mudrā)

are especially praised.

3.2.19. Yogaśikhā Upaniṣad ("Secret Teachings of the Yoga of the Pointed

Flame")131

The title of the text can be interpreted in at least two different ways. "The

Yogaśikhā from which this Upaniṣad gets its name is either to be understood as 'the

peak, the highest result of the Yogic meditation' or [more likely] as 'the pointed

flame' in the heart, in which the Yogin sees the highest being."132 The Yogaśikhā

Upaniṣad has northern (A) and southern (B) recensions represented by series of

129 Bouy (1994), p. 102. The KhV, "the arcane science of flight" or the "aviator's science"

dates from the fourteenth century. The text employed by the redactor of the 108 Upaniṣads was

most likely similar or the same as the Madras manuscript. See also White (1996), pp. 169-170. The

southern recension of this text comprises only the first chapter of longer northern recension. An

expanded rendering of the KhV is also found in the Matsyendra Samhitā.

130 Bouy (1994), p. 102. The Candrāvalocana / Candrāvalokana of Matsyendranāth date from

before 1450, as it is quoted by the Haṭhayogapradīpikā. It is a dialogue between Śiva and

Matsyendra.

131 Additional translations: (A) recension, Deussen (in English by Bedekar and Palsule),

(1997), pp. 709-711.

132 Deussen (1997), p. 708. See same for brief but detailed introduction to the (A) text.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

variant manuscripts. Recension (A) of the Yogaśikhā contain ten to eleven ślokas.

Like so many of the other southern variant texts, there are earlier southern

editions (B1) that are virtually identical to the northern (A) texts. However, the

(B1) texts contain enough variants to suggest that South India had its own

transmission history leading to the expanded (B2) texts. In the case of a number of

Yoga Upaniṣads, there have been no studies of (B1) texts; even though there is

evidence in the (B2) text to support a different transmission via a hypothetical

(B1). The Yogaśikhā Upaniṣad is an exception, for there are available (B1) texts that

constitute the variant earlier texts that are similar but not identical to the extant

northern (A) manuscripts.133 The (B1) manuscripts of the Yogaśikhā Upaniṣad have

not been critically edited.

It is most probable that the (B1) Yogaśikhā of ten verses was expanded to

form the (B2) text by the compiler of the 108 Upaniṣads in South India in the

seventeenth or early eighteenth century.134 The northern Atharvanic redaction of

this text was contemporaneous with or after Śaṅkara (eighth or early ninth

century) and before Śaṅkarānanda (ca. 1300). Therefore, the (A) recension was

written between the ninth and thirteenth century.135 The (A) text is first attested

by title between the twelfth and thirteenth centuries. The (A) recension is twenty-

133 Bouy (1994), p. 103.

134 Bouy (1994), p. 106.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

second in the Atharvanic collections. The southern (B2) text is sixty-third on the

list of 108 and is attached to the Kṛṣṇayajurveda.

The (A) recension recommends om recitation, and describes the movement

of breath up through the body, where it escapes through the cranial fissure. The

later verses state that the same benefits that derive from yoga practice can also

accrue from recitation of the text (a claim that is probably an interpolation).136

Verse 1 states that the yogaśikhā is the highest knowledge and that the limbs of the

yogin who meditates on the mantra137 do not tremble.138 Verses 2-3 summarize the

preparations for practice: seated in lotus posture, his gaze fixed on the tip of his

nose, with hands and feet and manas controlled, the yogin should meditate

continuously on the mantra om, enshrining the highest god in the heart. Verses 4-7

provide a summary of the results of meditation, comparing the human body with

a temple.139

135 Bouy (1994), p. 31. Although the ninth to eleventh centuries are not impossible, the (A)

recension of the text most likely dates to between the twelfth and thirteenth centuries.

136 The Telugu (1883) recension reads: "He who goes through this meditating thrice a day

reaches, from out of laziness and inattentiveness, to the pure bliss." Deussen (1997), p. 711. Cf.

Adyar Library Series 1.76-77 that reads virtually the same as the Telugu. The B1 (A) editions

reading that bhakti-style recitation of the text replaces the actual practice of mantra yoga, is so out of

character with the style and mode of this and the other Yoga Upaniṣads, that the later attested

Telugu variant appears to have maintained the original verses 8-9. This is an example of how the

dating of these texts and the numerous manuscript variations was a continuing process.

Additionally, this demonstrates that typing texts as (A) "older" and (B) "younger" is an

idealization, not an indisputable presentation of historical facts.

137 I read om

138 (B) states that practice of the mantra causes the body to tremble. The (A) reading seems

more probable but "trembling" and "not trembling" can both result from meditation practices.

139 Reading verse 4 as a Locative Absolute.

185

History, Text, and Context of the Yoga Upanisads: Chapter 3

One shall seek the highest in the body-temple having one post,140 nine

doors,141 three columns,142 and five gods.143 Therein glows a sun

surrounded by flame-like rays. In its midst is a fire, which burns like

the wick of a lamp. Its pointed flame, as large the highest god,144 is

there. Practicing yoga repeatedly, the yogin penetrates through the

solar mandala. Zigzag he goes upwards through the shining door of

the susumnā; breaking through the cerebral dome, he finally sees the

highest one.145

In the same fashion as the other early Yoga Upaniṣads texts, the (A)

recension of the Yogaśikhā Upaniṣad presents an impressionistic description of the

subtle body, rich with metaphor and suggestive language but without the

standardized language of kundalini yoga or haṭha yoga. The southern tradition

explicitly interpreted its tradition of the short (A or B1) recensions according to the

common symbols of Tantrism, as will be discussed at length in chapter 4.

Surrounding the verses of the (A/B1), the expanded (B2) text inserts materials

140 ekastambhe. Probably analogous to a sacrificial post, here referring to the spinal column.

141 navadvare. The conventional terminology for the nine openings of the body: eyes, ears,

nostrils, mouth, urethra, and anus.

142 tristhūne. The three columns, or even the three cords. Probably allofoms here for the

three primary nāḍis (iḍā, piṅgalā, and suṣumnā). Different commentaries interpret the three columns

as the three guṇas or three forms of karma.

143 pañcadaiwate. The analogy is not entirely clear. It could imply the five elements: earth,

air, fire, water, aether. Deussen suggests the five senses. The southern Adyar Library Series

translator takes the verse at face value, identifying the gods as Brahmā, Viṣṇu, Rudra, Īśvara, and

Sadaśiva, following the Śaiva orientation of the Yogasiddhāṃta/Yogabija source-text.

144 Parameśvara. The later southern tradition clearly interpreted this reference and

expanded the text based on the āgamic/tantric equation of Parameśvara with Paramaśiva or

Sadaśiva. Remaining consistent with the Upanisadic symbols and metaphors of the (A) tradition,

parameśvara refers here to the ātman as small as the tip of a flame enshrined in the temple of the

heart.

145 The Adyar Library Series (B2) text interpolates these verses with insertion of technical

terms from kuṇḍalini yoga of the Nāth source text. The verses are translated here according to

Deussen’s (1997) reading of the B1 edition and are adapted from Deussen with consultation of the

Adyar Library Series 1.72ff (p. 404).

186

History, Text, and Context of the Yoga Upanisads: Chapter 3

from the Nāths and the Tantras (the latter often via late yoga texts whose content is

tantric).

Like the Dhyānabindu and Nādabindu Upaniṣads, the (B2) recension of the

Yogaśikhā Upaniṣad includes variants of the (A) edition’s ślokas but with enough

differences to suggest that a different manuscript tradition from the extant (A)

recension was transmitted in South India.146 The longer (B2) texts fill in specific

details of practice that explain the allusions found in the short (A) text. The (B2)

recension of the text has 390 to 392 ślokas in six chapters (adhyāya) on mantra yoga

and haṭha yoga. Much of the text has a frame dialogue between Śiva and

Hiraṇyagarbha.147 Recension (B2) incorporates the contents of (A/B1) in its first

chapter as vereses 1.69c-79b, but with variants and elaboration. The text is

somewhat unified in its subject matter, but many chapters repeat information

from others with minor differences in focus or technical language. Even from a

cursory examination, all the chapters, with the exception of chapters 2 and 3, read

as separate texts.

146 Bouy (1994), p.31. See also Bouy (1990a) for detailed discussion of the variance between

the northern recensions and the root texts of the southern tradition. See also Deussen’s (1997)

introduction and notes on the text. In some cases, the southern (B1) tradition appears to have

maintained an older recension than the extant (A) texts. See above and the notes following.

147 Hiraṇyagarbha, the golden embryo, is an epithet of the creator god, Brahmā, and has an

is used as a technical term for the original creator from the Vedic period onward.

187

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Christian Bouy has traced the sources of virtually all of the contents of the expanded (B2) text.148 Chapter 1's expanded contents derive almost entirely from the fundamental Nāth text that is called the Yogasiddhāmṛta or (more often) the Yogabija.149 Chapters 2, 3, and 5 derive most of their content from the Yogasāra, a dialogue between Śiva and Śri-Devi.150 A few verses from chapter 5 quote the Haṭhayogapradīpikā. Chapter 4 quotes the Aparokṣānubhūti attributed to Śaṅkara.151 Chapter 6 is predominantly drawn from chapters 5 and 6 of the encyclopedic Upāsanāsārasaṃgraha.152

Chapter 1 of (B2) demonstrates that it is a yogic Śaiva Āgama work that also incorporates Vedānta teachings, emphasizing the importance of both jñāna and yoga. Its frame dialogue is between Śaṅkara-Maheśvara and Hiraṇyagarbha.153

The first chapter praises kaivalya, describes ātman, and preaches that a jīva must

148 Bouy (1994), pp. 35-36, 102-106, notes and index III.

149 See Bouy (1994), p. 102ff for comments on the first chapter of the expanded (B) YŚU. The Yogasiddhāmṛta (or Yogabīja) was written no later than 1450 as it is quoted in the Haṭhayogapradīpikā. The text could date to the thirteenth century. The text is attributed to Gorakhnāth and shares considerable overlap with the other root texts of the Nāth tradition. See White (1996), pp. 140-141, but also throughout for discussion, dating, and details concerning the haṭha yoga works attributed to Gorakh. For additional materials concerning the writings of Gorakh, see also G. W. Briggs (1998), H. Dvivedi (1981), A. K. Banerjea (1999).

150 More than one text bears the name Yogasāra. This Yogasāra is of unknown date but is quoted and cited by the sixteenth or seventeenth century text, the Upāsanāsārasaṃgraha. Its contents are drawn from multiple sources showing common content with Vedānta, Āgamas and Tantras.

151 The Aparokṣānubhūti dates to between the eighth and thirteenth century and is often attributed to Śaṅkara, the eighth century scholar and mystic of non-dual Vedānta.

152 The Upāsanāsārasaṃgraha is an encyclopedic work on yoga that dates to the sixteenth or seventeenth centuries. It is a compilation text itself. For a chart of quotations in the canon of Yoga Upaniṣads, see Bouy (1994), p. 92, see also p. 36 and notes.

188

History, Text, and Context of the Yoga Upanisads: Chapter 3

become Śiva. It also describes many kinds of siddhis (psychophysical powers) that

result from yoga practice. It considers siddhis gained from mercury-alchemy, herbs

and medicine, or magic spells artificial and inferior to the siddhis resulting from

yoga. The text decries book learning without yoga practice. From verse 81 on the

text begins a detailed discussion of haṭha yoga practice. The mechanisms of

cleansing the body, raising the eight-coiled kuṇḍali, and piercing the three knots

(granthis) are described according to the tantric language of the Nāth yoga texts.

Chapter 1 includes detailed discussions of the four kumbhākās and three bandhas.

These discussions are similar to other citations of Nāth texts found in the Yoga

Upaniṣads canon. As mentioned earlier, these materials are drawn from the

Yogasiddhāmṛta, but they are also similar to the South Indian recension of the

Gorakṣaśataka quoted by the Yogakuṇḍali Upaniṣad.¹⁵⁴ The first chapter incorporates

other variations on standard tantric and Nāth themes via some unique and

interesting terminology. For example, there are several passing references to

¹⁵³ The source for this chapter, the Yogasiddhāmṛta, is a apparently a dialogue between Śiva

and Śakti, but I have not examined the text. Our text's modified frame often drops Hiranyagarbha,

simply employing "Lotus-born" as the epithet of Śiva's dialogue partner.

¹⁵⁴ In other words, the redactor(s) of the 108 Upaniṣads, who expanded the southern texts

utilized a vast corpus of yoga texts. They often borrowed similar passages from different works

that focused on the same practices. The variations from the original sources and multiple

interpolations suggest that the redactor(s) were practicing forms of haṭha yoga that encouraged

them to select certain material from multiple sources. However, they seem to ignore content that

did not agree with their guru(s) or practices. For example, certain techniques (for example, bandhas

or mudrās) are repeatedly explored in the Yoga Upaniṣads, drawing on northern and southern

textual sources. Whereas other mudrās or techniques are only briefly summarized or even ignored.

It is also the case that many of the text attributed to Gorakh appear to be elaborations of the GŚ, see

White (1996), p. 141 and notes.

189

History, Text, and Context of the Yoga Upanisads: Chapter 3

distinctly Nāth terminology, such as the term paścima dvāra (the back or western

door to the suṣumnā).155 This term (borrowed here from the Yogasiddhāmrta) is not

otherwise employed by the Yoga Upaniṣads canon of expanded texts. Unlike many

frequently borrowed concepts, such as khecarī mudrā, it seems unlikely that the

redactor(s) of the 108 Upaniṣads had any practical intention in using this concept.

The practices described in chapter 1 of (B2) are referred to as suṣumnā yoga.

Drawing from the Yogasiddhāmrta, the text celebrates different types of yoga. First,

mantra, laya, haṭha, and rāja yogas are described in an ordered sequence. The

definitions of laya and haṭha are simplistic and etomographic. Mantra yoga is the

practice that turns haṃsa recitation into so’ham. Rāja yoga is defined as the joining

of rajas and retas (menstural flux and semen) that leads to siddhis.156 All of these

yogas are considered one mahāyoga, "great yoga," and they all are said to employ

breath control. Later in the chapter the four yogas are repeated, with the addition

of bhāvana yoga and sahaja yoga crowning the previous fourfold list. Chapter 1

ends with a description of the standard six cakras. This description of the cakras is

combined with four additional subtle-body locations: the four piṭhas (Kāmarūpa,

Purnagiri, Jālamdhara, and Uddiyāna). The sahasrāra or equivalent is not

mentioned.

155 Concerning the paścima dvāra, see White (1996), p. 255, chapter 8 in general, and notes.

156 This is a thoroughly tantric (via the Nāths) definition of the classical term Rāja Yoga.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Chapter 2 expounds mantra yoga through a hierarchy of meditation states

drawn from the Yogasāra. The praṇava is described as the root mantra

(mūlamantra). It is described as the nature of the nāda liṅga, the sūtratva/sūcatva, the

piṭha, and bindu. The nāda of the praṇava is celebrated as the highest mantra. It is

unclear from the text whether the praṇava intended by the source is the oṃ mantra.

Although Upanisadic readers and redactors would certainly interpret the term

praṇava to mean oṃ, the Tantras employ other root mantras that they also designate

as praṇava, modeled after the oṃ praṇava.157

Chapter 3 relates mantra yoga to cosmogonic expression of the Word, Vāc,

drawing from the Yogasāra as well as the Tantras’ descriptions of the stages

through which Vāc unfolds (parā, paśyanti, madhyamā, and vaikharī).158 The text

then connects the levels of speech with each cakra, providing a system of

correspondence that links the microcosm of the human body to the macrocosm of

the cosmic Word.159 Through meditation on Śabdabrahman the yogin attains the

Parabrahman.

157 In the broadest sense, the praṇava as a term intends oṃ. However different Tantras also

used the term to refer to “haum” or whatever mantra was at the foundation of their systems. For

specific details and general background, see Padoux (1992), pp. 142, 380, 402-404, passim.

158 This section (or its source text) should be read with Padoux’s (1992) chapter 3 (and

throughout) for it draws heavily from the tantric conceptions, especially those relating to the levels

of speech discussed throughout.

159 Here the presentation is also according to a relatively standard pattern as seen

throughout the Tantras. See Padoux (1992), in general.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

Chapter 4 presents generic Vedānta, describing the illusory nature of the

world and the body. As mentioned earlier, large portions of this chapter are

drawn from the Aparokṣānubhūti attributed to Śaṅkara. The illusory character of

the phenomenal world and the true nature of ātman are explained.

Chapter 5 provides descriptions of esoteric internal physiology drawn from

the Yogasāra. This chapter has a Vaiṣṇava orientation, although with many

influences from Śaiva language and concepts. The body is called the temple of

Viṣṇu and the abode of bindu, nāda, the great liṅga, Viṣṇu, and Lakṣmī. It

describes the six cakras and four piṭhas, presenting a system that is virtually

identical with the system found in chapter 1 that is drawn from the

Yogasiddhāmṛta. It also presents the nāḍis, five agnis, kuṇḍalini, and khecarī mudrā,

and gives its own list of the yogic siddhis.

Chapter 6 also presents considerable information on esoteric physiology,

but drawn this time from chapters 5 and 6 of the Upāsanasarasamgraha, which is

itself a compendium.160 Kuṇḍalini, the nāḍis (especially suṣumṇā), śaktis, haṃsa

recitation, and breath control are all discussed again, with minor variations from

the preceding descriptions of the same material.

160 The Upāsanāsarasamgraha is an encyclopedic work on yoga that dates to the sixteenth or

seventeenth centuries. For chart of quotations in the canon of Yoga Upaniṣads, see Bouy (1994), p.

92, and index.

192

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

As in other Yoga Upaniṣads, the redactor(s) borrowed materials covering the

same topics from different sources to flesh out their root texts, similar to a second-

order commentary.161 Like the other southern Yoga Upaniṣads compilation texts,

the materials from older sources are organized and manipulated by the redactor(s)

with various flourishes and elaborations. In other words, they do not always

appear as direct quotes. The redactor(s) use the sources with rhetorical flair to

meet their own interests and fulfill their own philosophical and programmatic

needs.

3.2.20. Yogatattva Upaniṣad ("Secret Teachings of the Principles of Yoga").162

The Yogatattva Upaniṣad has northern (A) and southern (B) recensions.163

The (A) text has fifteen ślokas. The southern (B) recension presents approximately

142 stanzas on yoga from a non-dual Vedānta position. The (A) recension dates to

the period of the majority of other northern texts, being attested between the

161 The materials read even more like the contemporary "report" genre that gathers results

from similar studies into one source. It follows the pattern of ‘cakra’ - Nāth cakra study 1a, tantric

cakra study 1b, āgamic cakra study 2a, etc. Whereas yoga texts have long been read and analyzed

for the philosophical contents, it has not been reinforced that these texts are also laboratory texts,

reporting the experiments of humans seeking perfection or immortality through yogic, tantric,

(al)chemical means, or a combination of all three.

162 Additional translations: (A) recension, Deussen (in English by Bedekar and Palsule),

(1997), pp. 713-716. (B) recension, Varenne (in French), (1971), pp. 49-70.

163 The YTU has recieved considerable scholarly attention. It has (A) and (B) recensions, yet

few of its commentators have differentiated the two. Eliade, Feuerstein, Beck, and Varenne all

comment on the longer (B) recension of this text. Eliade summarizes and comments on Adyar

Library Series (B) edition of the text, and considers it one of the most important of the Yoga

193

History, Text, and Context of the Yoga Upanisads: Chapter 3

twelfth and thirteenth centuries. The (B) recension, like the other southern

expanded texts was probably revised with the formation of the canon of 108

Upanisads in the seventeenth or early eighteenth centuries. Both versions begin

with the same two introductory verses. The (B) text incorporates the second

stanza of recension (A) twice and includes verses 3–13 of recension (A) near the

end of its text. The (B) text skips verse 14 of (A) and then includes verse 15 at the

very end.164 The (A) text is twenty-third in the North Indian collections of the

Atharvaveda. The southern (B) text is forty-first in the list of 108 Upanisads and is

attached to the Krṣṇa Yajurveda. The (B) text draws many of its additional verses

from the Yogasiddhāmṛta, the Yoga Śāstra of Dattātreya, and the Yogayājñavalkya.165

Recension (A) of the Yogatattva shares common content with the

Dhyānabindu Upaniṣad, suggesting that the two texts depend on a common source

Upaniṣads. However, his mis-association of the dates of (A) with the content of (B) invalidates

much of the thrust of his analysis.

164 Bouy’s analysis of the Tanjore (T) manuscript nearly follows the breakdown of verses I

have provided based on the Calcutta BI edition used by Deussen. Bouy does not evaluate the BI

edition but his analysis provides numerous additional details. See Bouy (1990a), pp. 97–98.

165 See Bouy (1990a), 109. Yogabija (YB) (known in one manuscript variant as the

Yogasiddhāmṛta) dates to no later than 1450 as it is quoted by the Haṭhayogapradīpikā. It is a root-text

of the Nāth haṭha yoga tradition, often attributed to Gorakhnāth. Thus it was probably written

between 1100 and 1450. The Yoga Śāstra of Dattātreya was written around 1300. The Y YV was

written in south India between the tenth and early fourteenth centuries. For in-depth study of the

Y YV, see Geenens (2000). The YB shows considerable influence from the root Nāth text the GŚ.

The YSD and the Y YV contain some virtually identical overlapping content. This demonstrates

that it is not only these Yoga Upaniṣads that follow the strategy of expanding old or writing new

texts based on a earlier yoga texts.

194

History, Text, and Context of the Yoga Upanisads: Chapter 3

tradition of Upanisadic yoga.166 Verse 1 praises yoga and the emancipating value of

listening to and reciting the text.167 Verse 2 calls Visnu the great yogin, the lamp of

truth, and the highest Purusa. Verses 3-5 describe the cycle of samsāra via various

metaphors. Verses 6-8 praise the all-encompassing nature of om. The threefold

syllable is equated with the three worlds, three Vedas, three times, three gods,

three fires, and three gunas. The half syllable (anusvāra-bindu) pervades all and

goes to the highest place.

In verses 9-11 the text presents a picture of the subtle nature of the heart as

a lotus, similar to depictions found in the other Yoga Upanisads.

The mind has its seat in the lotus flower that occupies the space in the

heart, calyx down, stalk above. With the u-sound it becomes luminous,

with the m-sound it resounds. Motionless is

the half-sound.168

Verses 12-13 describe the withdrawal of the hands, feet, and head and the

control of breath. Next the nine doors (body openings) are closed and kumbhaka

(retaining the breath) is indicated. Verse 14 returns to the images of the related

texts with a description of liberation. "Till, breaking, as through a lotus petal, the

impetuous wind leads him, whom they know as the guileless, between the brows

166 Deussen echoes this conclusion in his analysis. See Deussen (1997), pp. 713-716,

especially his introduction and notes. Verses 1-2 appear in some recensions of the DhBU but make

better sense here. Verse 8 appears as DhBU verse 7, more consistent with its context therein. Verse

9-11 agree with DhBU 12-14 (and Mahānārāyana Upanisad 11.8).

167 Cf. the notes on variant verse 8 of the YṢU, in section 3.2.19.

168 For Deussen's (1997) translation of the same verses, see p. 715.

195

History, Text, and Context of the Yoga Upanisads: Chapter 3

and the forehead.169 Verse 15 closes the text with the supremacy of the yogin and

the importance of calm, remote location.170 Thus, the (A) text presents a brief and

impressionistic summary of the Upanisadic mantra yoga that leads to escape from

saṃsāra.

The long (B) text is Vaiṣṇava in tone and defines an eight-limbed system of

four yogas: mantra, laya, haṭha, and rāja (following the Yoga Śāstra and the

Yogayājñavalkya).171 The (B) text begins with variants on verses 1-2 of (A) and

original transition verses in 3-4. Verses 4c-18b draw on Yogasiddhāmṛta 3c-28b.172

This section presents a Vedāntin interpretation of Nāth yoga, describing the jīva

ensnared by the web of māyā, the jīva as the paramātman in its true nature, and the

necessity of both jñāna and yoga. From verses 19 to 130 the text draws from the

Yoga Śastra and Yogayājñavalkya, presenting its four yogas of eight limbs. The text

describes the merits of each yoga. With the four yogas understood as a hierarchy of

practices appropriate to certain types of individuals (from least able to most

adept). Limbs, mudras, and bandhas are listed and some are described. Measures

of time for breath control are recommended. The text describes each limb as well

as the siddhis that come from practice. The end of the text replicates the (A) text

with elaborations and variants. The borrowings from other sources are cursory,

169 Following Deussen's (1997) translation. Cf. DhBU 23.

170 Contrary to Deussen's note 2, read more like ANU 17.

196

with lists, terms, and some brief descriptions. There is generally less detail than is

found in the Nāth texts.

3.2.21. Cūlikā Upaniṣad ("Secret Teachings of the Crest-Jewel").173

This text of twenty-one ślokas is one of the oldest of the Yoga Upaniṣads,

attested between the ninth and eleventh centuries. The Cūlikā Upaniṣad is fifth on

the lists of Atharvanic Upaniṣads. Among all the Yoga Upaniṣads, it has the most

decidedly Vedic character and is not haphazardly associated with the

Atharvaveda.174 The “crest-jewel” in the title is a characterization of yoga as the

peak of the pillar of Sāṁkhya. This Sāṁkhya is not the classical system but a system

more like that found in the classical Upaniṣads, especially the Maitrāyaṇīya

Upaniṣad. Reading it as a practical yoga text is open to interpretation, but it does

contain materials helpful for understanding the metaphysics of yoga found in

other early (A) recensions of Yoga Upaniṣads associated with the Atharvaveda.175

171 This picks up the reference from the primitive text (verse 2), and fleshes it out with the

content of the two Vaiṣṇava yoga texts (the YSD and YYV).

172 Bouy (1990a), p. 109.

173 Translation: (A) recension, Deussen (in English by Bedekar and Palsule), (1997), pp. 677-

681. I have not evaluated the southern recension (titled differently as the Māntrikā) of this text as

no one considers it a yoga text.

174 Bloomfield (1899), p. 19. See also Goudriaan & Schoterman (1994), pp. 1-5.

175 I have provided summary here, not because the CU has extensive practical yogic

content. Some scholars have classed it with the other Yoga Upaniṣads. I have analyzed it for

readers who, aware of its classification, want to understand its relationship to the other Yoga

Upaniṣads. The homologies listed below are widely utilized and explored in the canons of minor

Upaniṣads, especially those of the yoga variety. Additionally, other Yoga Upaniṣads contain little

practical yogic content as well, as seen in the summaries above (northern TBU, southern MVU).

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The text contains extensive quotations from the Atharvaveda to expound its Upanisadic Sāṇkhya-yoga. The Cūlikā Upaniṣad begins with a description of reality as the sun-bird (verse 1): radiant, eight-footed, three-stranded, eternal jewel, flaming, wandering twofold, seen and unseen. This characterization (along with that in verse 2) plays on the old sacrificial imagery of ātman and Puruṣa, and employs images of the Sūrya and the fire altar. The sun is visible whereas the ātman is hidden, but both are of the form of the bird (the hamsa of the Vedas and the later Vedānta).

This sun-bird, identified as Puruṣa, creates the world with prakṛti-māyā. Then the Puruṣa as babe, sucks at the breast of māyā (enjoying the world of the senses), just as all lesser beings suck at the breast of the mother of creation.

Puruṣa, as Brahman, is not ensnared by his enjoyment of māyā's breast, whereas all other beings are ensnared. The brahmin can be like Puruṣa by knowing the true nature of reality, Brahman. This knowledge comes through correspondences, the science of upaniṣad. The text follows with a long list of homologies for Brahman drawn from the Atharvaveda (the esoteric teachings of the mantras): brahmacārin, vrāta, skambha, Rohita, kāla, prāṇa, bhagavān-ātman, Puruṣa, Īśvara, as well as others.

The characterizations continue with Brahman as the twenty-six or twenty-seventh (allusions to the metaphysical numbers of the Sāṇkhya system). These are references to the nirguṇa Puruṣa as the "tip" of this Upanisadic Sāṇkhya-yoga

198

system (compare text title with verse 14). The text describes the cycle of rebirth. It

then concludes that if Brahman is proclaimed (ritually) then food (merit, reward)

will accrue to the brahmin and the ancestors. Better still, whether one is a brahmin

or not, if one knows (experiences directly) Brahman, then “he disappears, merged

in the ocean of Brahman.”

The Cūlikā Upaniṣad is thus not so much a practical yoga text as one that

presents the metaphysics and esoteric connections that are explored in later yoga

texts. Most of the homologies found in the text, although drawn from the

Atharvaveda, are the raw materials that give rise to new interpretations and

symbols in the other Yoga Upaniṣads.

3.3. Concluding Remarks

These synopses of the twenty-one Yoga Upaniṣads provide a window into

the complex history of the mantra yoga expounded in the northern recensions

between 800-1300 and the syncretistic haṭha yoga elevated in the southern canon

redacted around 1600 to 1700. The northern texts show an emergent tradition that

is impressionistic, experimental, and not completely systematic. The northern

recensions do rely on certain systems and technical language that predate them,

generally drawing from classical Upanisadic Vedānta combined with certain

elements of yoga, medicine, alchemy, and tantra.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The impressionistic northern texts present overlapping perspectives on a tradition that practiced breath control and meditation. The threefold breath

control of pūraka, kumbhaka, and recaka, which is not found in the classical Upaniṣads is present in these minor Upaniṣads. The unnamed nāḍis of the classical

Vedānta have been replaced with systematic terminology of three primary nāḍis and the widespread homologues of sun, moon, and fire. The term, Kuṇḍalini does

not appear as a technical term in the early Yoga Upaniṣads, but maṇas, śikhā, and prāṇa are described in terms familiar to kuṇḍalini yoga.176 Moreover, no system of

cakras is evident in the northern recensions. Esoteric yogic body locations that are similar to the cakra systems are called marman ("vital point") and dvāra ("gate").

Flowers (usually the lotus) are also employed as images for the "centers" of the yogic body. The imagery of the "gates" of the body resonates with the imagery of

the three granthis ("knots") associated with Brahmā, Viṣṇu, and Śiva in yoga and tantric texts.

The metaphysics of these northern traditions is developed but unsystematic. It appears to be a combination of the unsystematic Sāṁkhya and

the Vedānta of the classical Upaniṣads. There is no predominant sectarian

176 This is not unusual as kuṇḍalini appears to have gained terminological ascendancy only slowly. Prāṇa is used in this way in the Vijñānabhairava Tantra and śikhā is often used in the texts of the Western Transmission. See also Flood's comments (1993), pp. 256ff.

200

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

theological emphasis in these early texts, and although gods are mentioned,

goddesses are absent.

The yoga practices described in these northern texts primarily focus on

recitation of the mantra om. The purpose of silent meditation (with inner ear and

eye) is to focus the yogin’s consciousness into a force that can escape the body at

death through the cranial suture. This escape of the breath as ātman from the body

is held to bring salvation and liberation from suffering and rebirth. These texts

provide images that are similar in many ways to the images of the Buddhist and

Hindu Tantras. Images of breath, mind, or pointed flame that move through

pathways (nāḍis) are central to the developed systems of the Tantras. The

importance of the central pathway that leads to liberation is repeatedly presented

in these texts. This system of pathways is as old as the classical Upaniṣads and

appears to be shared by pan-Indic schools of yoga, tantra, and alchemy. However,

the names of the structures of the subtle body are not standardized across

traditions. Systematizing yoga according to limbs appears in some of the early

northern texts, as we have seen. In all cases, the system is six-limbed. The six-

limbed system of yoga, ṣaḍaṅga yoga, is first found in the Maitrāyaṇīya Upaniṣad,

and is later associated with the systems of Buddhist tantric yoga, as well as with

the yogic teachings of the twelfth-century yogin and Nāth Siddha, Gorakhnāth.

History, Text, and Context of the Yoga Upaniṣads: Chapter 3

The expanded southern recensions of the Yoga Upaniṣads reflect the

systematization and standardization of yogic language and practice over the

intervening centuries between the thirteenth and seventeenth centuries. Various

yoga texts embody this standardization. The eight-limbed path of Patañjali was

transformed over time by elaboration into the eight-limbed path of Dattātreya or

Yajñavalkya. The elaborate six-limbed haṭha yoga of the Nāths, the eight-limbed

paths, and other traditions developed during this period a consistent vocabulary

of the subtle body and the practices and theories of yoga. The southern redactors

of the Yoga Upaniṣads appropriated and recast these systems and expanded the

northern texts according to a vast array of texts and traditions.

The southern canon of Yoga Upaniṣads quote from over two dozen known

sources. The eleven expanded texts are often eight to ten times the length of the

original northern recensions. The additional ten texts, which are only attested in

the southern collections, continue the pattern of the expanded texts, quoting many

of the same sources and expounding the same combination of om and hamsa-

centered mantra yoga together with the systems and practices of haṭha yoga. The

established sources of the expanded texts fall into five categories: Yoga texts,

medieval non-dual Vedānta texts (including the devotional Gītās), Tantras and

Āgamas, and classical Vedic and Upanisadic sources. The yoga texts include the

haṭha yoga texts of the Nāth Siddhas, yoga encyclopedias, and South Indian

202

History, Text, and Context of the Yoga Upanisads: Chapter 3

Vaisnava yoga works. Late medieval Vedānta sources include non-dual works

that generally follow the model of Śankara's philosophy, while the devotional

Gitās are late medieval encyclopedic works expounding theistic non-dual Vedānta.

Tantras and Āgamas are less often cited directly but constitute the textual

background materials for most of the yoga sources upon which the southern

recensions draw, including yoga texts associated with Gorakhnāth, Yajñavalkya,

and Dattātreya. Finally, true to their genre, the Yoga Upanisads often allude to and

cite the classical Upanisads and other Vedic materials.

Almost half of these quotations and paraphrased materials come from the

Nāth Siddha traditions—particularly from the writings attributed to Gorakhnāth,

such as the Goraksaśataka, but also the later Nāth derivative Hathayogapradipikā. A

quarter of the remaining citations come from other kinds of yoga texts such as the

Kashmiri Laghuyogavaśiṣṭha, the encyclopedic Upāsanāsārasaṃgraha, and the eight-

limbed paths of the Yogayajñavalkya and Yoga Śāstra of Dattātreya. The remaining

sources are a collection of other late medieval compilations that include diverse

materials. In the remaining two chapters we will turn to an analysis of selected

elements of the mantra yoga and hatha yoga found in the Yoga Upanisads and will

examine the historical development of the texts and the communities that

produced them.

203

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

Chapter 4

The Traditions of the Yoga Upaniṣads:

Mantra Yoga and Haṭha Yoga

The six plus one cakra system of Hindu haṭha yoga is one that

we are so familiar with as to assume that it emerged, fully

formed, like Athena out of Zeus's forehead. This of course

was not the case.

—David White1

White's statement, quoted above, reflects a fundamental difference in

approach between scholars of tantra like White and many scholars of yoga. White

and other scholars of tantra have been acutely concerned with the diachronic

history, development, the specific features of medieval Indic religious life. Most

scholars of yoga, in contrast, have been more concerned with synchronic ahistorical

generalizations concerning the essential nature of yoga. White's observation

concerning "the six plus one cakra system" provides an opportunity for us to

grapple with several important issues regarding the nature and history of the Yoga

Upaniṣads and the history of scholarship focusing on yoga.

Many features of the Yoga Upaniṣads deserve detailed study. In this

chapter, I will examine the subject of the "yogic body"2 in both the northern and

1 White, "Yoga in Early Hindu Tantra," p. 2.

2 I employ the term "yogic body" for the theoretical body of the yogin in the Yoga

Upaniṣads. This theoretical body includes concepts both material and ethereal. The Tantras and

Alchemical texts employ such terminology as subtle, gross, enjoyment, causal or other epithets for

bodies. These terms do not appear in the northern Yoga Upaniṣads. For the southern Yoga

204

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

southern canons of Yoga Upaniṣads. Two different images of the yogic body were

created and explored by the mantra yoga of the northern texts and the haṭha yoga of

the southern texts. The yogic body depicted in the northern texts does not

represent a standardized system. The images and metaphors that the northern

texts employ are used in the service of meditation and appear to be visualizations.

During recitation of mantras, the yogin visualizes the morae or phones of the

mantra as well as gods, flowers, and other objects inside of his body. Phones,

gods, and other constructs are projected into the areas of the navel, heart, throat,

and head of the yogin. The objects and locations of these visualizations depict a

subtle inner world created through visualization. These visualizations construct a

yogic body. A combination of the yogic body of the northern Yoga Upaniṣads and

the subtle bodies of the Tantras provides the twofold foundation for the nāḍi and

cakra subtle body system of the southern Yoga Upaniṣads.

Unlike the unsystematic northern presentation, the southern texts present a

standardized system: the six plus one cakra system. This six plus one cakra system

was presented systematically in a ninth to tenth century Tantra of the Yoginī

Kaula, the Kaulajñananirnaya of Matsyendra.³ There is also early mention of the six

Upaniṣads, these latter terms are appropriate, as the expanded texts draw from tantric and haṭha

yogic sources. See section 4.2 below.

³ White (1996), pp. 73, 134-135; See also White, “Yoga in Early Hindu Tantra,” p. 4. The

Kaulajñananirnaya presents six cakra and an eleven-cakra systems. This text is attributed to

Matsyendra or Macchanda.

205

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

plus one cakra system in the tenth century Kubjikāmata Tantra of the Western

Transmission.4 These systems are further elaborated in the haṭha yoga texts of the

Nāth Siddhas. An analysis of the transformation of the Yoga Upaniṣads' yogic

body from unsystematic visualizations into a standardized system of subtle

physiology offers the occasion to examine and explain the distinctions between the

two canons. In the concluding chapter, we will review the history that these

distinctions reveal.

Whether yoga systems comprise eight limbs or six limbs and are

Brahmanical, śramanic, or tantric, there are certain fundamental characteristics

found in most yoga systems. These fundamentals are breath control, withdrawal

of the senses, concentration, meditation, and enstatic absorbtion (prāṇāyāma,

pratyahara, dharaṇā, dhyāna, and samādhi).5 These practices' primary goal is the

control of the body and mind of the yogin. The goals of this training and control

include immortality, longevity, super-human power, and liberation. The yogins

elaborate and explain their system's limbs and goals with metaphors, declarations,

4 Sanderson suggests that the six cakra system first evolved in the Western Transmission

and was then adopted by the Kaula traditions. Following White's argument, it is more likely that

there were emergent concepts in the classical Upaniṣads, medical texts, and Saiddhāntika works

that were systematized in Matsyendra's Kaulajñānanirnaya and shortly thereafter adopted by the

Western Transmission. See Sanderson (1986), p. 164; Sanderson (1988), pp. 687-688; White, "Yoga

in Early Hindu Tantra," pp. 2-3, 10, and throughout.

5 The many additional limbs of any yoga system always lead up to or elaborate on these

five processes, sometimes under different names.

206

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

and descriptions of structures projected by imagination and visualization onto

their bodies.

The imagery of the body in the Yoga Upaniṣads often replicates the celestial

imagery of the classical Upaniṣads. The concepts of salvation in the classical texts

involved the following model of the cosmos.6 The classical worldview depicts a

world that was covered by bowl-shaped or vaulted heavens. The sun is the only

"lid" or doorway that leads outside of the vaulted cosmos. In the cycle of birth

and death the eternal Self rises up in the smoke of the funeral fire or escapes from

the eye or head or some other part of the body. Ritualists could attain a positive

rebirth by going to the moon, from which they would be born again. Upanisadic

sages who knew the secret homologies of breath, mantra, heat, and sun could

escape the known cosmos and its cycles of death and birth by escaping through

the portal of the sun.

The Yoga Upaniṣads map the interior of the yogic body according to this

classical model of the cosmos. The body of the yogin is not just guts, muscle, bone,

and blood. It is also a micro-geography modeled on terrestrial landscapes, the

cosmos, or geometric shapes. Maps of this subtle geography are part of the yogin’s

special knowledge. The yogin’s practices follow this guiding map because it

reveals the path to the treasures of power, immortality, and liberation. The

6See also chapter 2 of this study.

207

History, Text, and Context of the Yoga Upanisads: Chapter 4

pilgrim on this body-path is the very mind and breath of the yogin. The yogin

must focus his thoughts, perceptions, conceptions, and meditations toward

discerning (hearing, seeing, and feeling) his way to liberation through the hidden

paths in his body.7

It is as if the mind and breath are trapped in a maze. Normal humans

entrapped in ignorance diminish their lives with their breath-mind running into

dead ends and tangled passages. Ultimately, their energies are burned out by the

time span of a single life. Due to the physics of transmigration, the process repeats

itself infinitely. The yogin learns how to fight the fire of time with the fire of yoga.

Through learning and practice, the yogin learns the secret paths and gains the

mantra-keys to escape successfully the life-body-maze.

The control of the body and mind require the yogin to be adept in both

theory (jñāna) and practice (yoga). The theories of the yogic body and mind were

developed simultaneously through meditation experiments, visualizations,

visionary poetry, and intellectual systematization. All of these

processes—inductive, deductive, homological, and taxonomic—operated to define

yoga, giving it content as well as system. The yogic systems of body, as the yogin’s

instrument and his laboratory, became one of the most elaborated areas of

7 I use the masculine pronoun for practitioners of Upanisadic yoga because, although there

is some evidence for female practitioners of yoga (one of the meanings of “yoginī”), the systems of

Yoga Upanisads are specifically directed to male practitioners.

208

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

speculation and practice for the medieval yogins. The Yoga Upaniṣads, both

northern and southern, illustrate some of these processes and concepts of the

medieval yogic body.

The theories of the body are elaborately detailed and developed in the

traditions of Tantrism, yoga, medicine (āyurveda), and alchemy (rasāyana). As

mentioned earlier, the northern Yoga Upaniṣads do not reveal a yogic body that is

fully systematized or consistently portrayed. These texts include details but not

the elaboration and consistency that one finds in the Tantras or haṭha yoga texts.8

The southern Yoga Upaniṣads redress this impressionistic yogic body, bringing its

description fully into agreement with the elaborate subtle body systems of

Tantrism, especially the six plus one cakra system.9

In the following sections I will describe and analyze the yogic body of the

northern Yoga Upaniṣads. I will then turn to an analysis of the subtle body

described in the southern tradition of Yoga Upaniṣads. These systems of body

largely depict the "territory" that the ideal Upanisadic yogin must control.

8 Being visionary and experimental, consistency is not essential to any of these systems.

9 Throughout, I have referred to the contents and systems of texts called Tantras or to

tantric yoga. The term 'Tantrism connotes the cultural milieu of medieval India: its texts, sects,

traditions, yoga, and other characteristics. See White (1996), pp. 1-2, and notes. See also Padoux

(1984), pp. 273. The subtle body is elaborately mapped in dozens of different arrangements.

Medieval south Asian mystics created a vast array of theories, systems, and visualization

meditations concerning the inner physical body and the invisible subtle body. These generated

models used medical and tantric symbols, metaphors, and metaphysics and showed considerable

variation. Throughout I have referred to the six or six plus one cakra system. These are the same

systems. It is sometimes even referred to as the seven cakra system.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

4.1. Mantra Yoga and Haṭha Yoga: Body-Mind Yogas

The canons of Yoga Upaniṣads explore the images of the yogic body in the

contexts of their practical and theoretical paradigms. At the beginning of the

present study, I raised the question: What is the yoga of the Yoga Upaniṣads?

Mantra yoga is an appropriate label for the practices of the northern canon of the

texts. The oṃ mantra is the primary focus of breath control and meditation.

Mantra is the energizing agent that actualizes power, immortality, and freedom for

the Upanisadic yogin.

These practices of the northern Yoga Upaniṣads include more than the

simple recitation of mantras. Posture, breath control, and meditation guide the use

of the mantras. In many ways, this northern yoga appears at first to represent little

morethan a combination of the paradigms of the classical Upaniṣads combined

with the eightfold system of Patañjali's Yoga Sūtra. There is more. The northern

Yoga Upaniṣads include clues to suggest that the yoga of the Tantras has already

altered the paradigms of the old Upaniṣads. A primary demonstration of this

influence is the transformation of the threefold oṃ recitation. The classical

threefold recitation was intended to free the Upanisadic yogin from the cosmos of

suffering and repeated death. The authors of the northern Yoga Upaniṣads

transformed this practice into fourfold and twelvefold mantra recitations aimed at

leading the medieval yogin through the maze of his subtle or yogic body.

210

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

The yogins of the northern Yoga Upaniṣads are represented as seeking

solitude and practicing breath control and seated postures. Thereupon, they

embark on a practice of repeating the om mantra, visualizing its letters and hearing

its sounds moving with their breaths. As the buzzing reverberation (nādabindu) at

the end of the mantra dissolves (laya) into silence (mauna), the yogin attains bliss.

Repeatedly practiced, this mantra yoga trains the breath in the proper ascent

pattern so that it exits the body at death through the cranial fissure. Piercing the

solar disk at the top of the skull, the yogin is declared free from evil, debt, and the

bondage of repeated death and birth. Like the principal Upaniṣads, northern Yoga

Upaniṣads generally depict a yogin freed from the body and the world at the time

of death. It is not until the redaction of the southern Yoga Upaniṣads that we find

references to jivan muktas who are liberated-while-alive and immortal yogins.

The northerntexts are brief and impressionistic. Their ambiguity allows for

multiple interpretations of their practices. These practices involve the sequence of

intoned mantra-buzzing-fading-silence, and then the process is repeated. Some

clues suggest that this mantra yoga accompanied nyāsa-like ritual meditations.10

These texts praise the value of silence, as it punctuated the mantra repetitions.

10 See especially the Kṣurika and Hamsa Upaniṣads, but several of the northern texts show

some evidence of nyāsa-like practice. Eliade calls nyāsa "ritual projection." It involves the

enlivening of both animate and inanimate objects but projecting vital breath, mantras, or gods into

them. Often this takes the specific form of projecting divinities into various parts of the body, on

the model of bringing a yantra, mandala, or image alive. See Eliade (1973), pp. 210-211.

History, Text, and Context of the Yoga Upanisads: Chapter 4

Alternately, the spoken mantra constitutes the preliminary practice that subsequently is replaced by a silent repetition sequence, wherein the sounds are imaginatively created within the body through visualization meditations. These two types of mantra practice can be seen as either independent methods or as part of a sequentially developed yoga method. The ambiguity of the texts is probably the result of a collection of similar practices being interpreted with minor variations by different gurus and their lineages.

In the southern Yoga Upanisads, hatha yoga and post-Śankara non-dual Vedānta transformed the mantra yoga as represented in the northern texts. Even after being transformed and elaborated in the southern texts, mantra recitation remains the fundamental practice prescribed in both canons. The southern texts include all of the materials of the northern recensions and therefore prescribe and praise the same practices of om recitation. In addition, these texts praise and recommend the practice of the hamssa mantra. The southern texts describe om and hamssa recitation as supported by a large number of hatha yoga practices and theories. Reciting mantras in combination with yogic postures, breath control, and meditation is the fundamental practice of all of the Yoga Upanisads, both northern

11 Variations on these types of meditation are still practiced in India.

12 Hamsa (the migratory gander mantra) recitation is described in the Hamsa Upanisad of the northern tradition. The HUI is the youngest of the northern texts written in the twelfth or thirteenth century. Hamsa recitation receives significantly more attention in the southern texts than it does in the northern Yoga Upanisads. Hamsa symbolism appears in both collections and predates all of the minor Upanisads. See chapter 2 and 3 concerning the "gander" mantra.

212

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

and southern. The differences found in the expanded southern texts are

essentially ones of quantity and variation. The sheer number of mantras and

postures and the addition of mudrā and bandha and other techniques make the

southern Yoga Upaniṣads a far more elaborate corpus.

The mantra yoga of the Yoga Upaniṣads was certainly a work in two parts as

practical yoga combined with voiced and silent recitations of words of

power—particularly om but also hamsa. This Upaniṣadic yoga constitutes a

discipline for controlling the body and mind. This disciplining involves

reproducing the constituent parts of the mantras in the body and mind of the yogin.

Mantras are always considered far more than human products, not merely

meditation props, mnemonic devices, or poetic fancies. Instead, mantras are held

to be truth and reality condensed into words or sounds of power. These power

sounds, through sounded and silent recitation, serve to catalyze the liberation of

the yogin.13 Like om in the classical texts, the mantras are literally the keys that

unlock the yogin's body escape hatch(es), so that the breath-self can properly

escape and the yogin can thereby achieve liberation. Alternatively, on the model of

the Tantras, the mantra is held in the heart or forehead in order to produce yogic

super powers and embodied immortality.

13 Cf. Padoux (1992).

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

Even a mastered and perfected yogic body necessitates a trained and restrained mind. Control and perfection of the mind require complex and repetitive exercises of concentration (dhāraṇā) and meditation (dhyāna). These practices result in a state of altered consciousness that is often called samādhi, the "integrated state of enstatic absorbtion." While the Yoga Upaniṣads sometimes use the term, samādhi, yet they also simply describe the goal of their yoga as release from bondage (mokṣa), embodied liberation (jīvanmukti), or disembodied liberation (videhamukti).

Most of the practices of meditation presented in the Yoga Upaniṣads are not explained in detail, since, like other yoga traditions, these details were generally learned orally by small groups of students from a yogin-guru. However, the texts do contain some descriptive details. In the context of the discussion of the yogic body in the Yoga Upaniṣads, I will explain selected features of meditation in northern and southern Yoga Upaniṣad traditions. The yoga of the Yoga Upaniṣads is a yoga for controlling the body and mind and attaining liberation through the practices of mantra. This yoga in two parts operates as a continuous blending of mantra and yoga whose goal is some form of power or salvation, whether the state of jīvanmukti or videhamukti. The yoga of the Yoga Upaniṣads is thus a blend of classical Upaniṣadic theory with the theories and practical methods of the medieval tantric yogins.

214

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

4.2 The Yogic Body in the northern Yoga Upaniṣads

As discussed earlier, the yogic body in the mantra yoga of the northern Yoga

Upaniṣads is not the standardized subtle body described in later Hindu traditions.

This standardization, wich finds fruition in the modern period, developed slowly

and with many variations. I have employed the term "yogic body" in the

following analysis because body terms such as gross body (sthūla śarīra) or subtle

body (sūkṣma śarīra) are not employed by the northern Yoga Upaniṣads. Images of

the body in the northern Yoga Upaniṣads are not at all systematic or particularly

self-conscious. The texts do not refer to the "body" or "bodies" in general,

although they do refer to the heart, toes, thighs, head, or other body parts and to

bodily functions such as the breath or thinking.

Eight of the eleven northern Yoga Upaniṣads include some clues as to the

imagined structures of the yogic body. The Amṛtanāda Upaniṣad and Haṃsa

Upaniṣad offer special examples of the differing pictures of the yogic body. The

evidence from the other northern texts reflects a complex system but does not

explain or give detailed descriptions. The Brahmavidyā, Dhyānabindu, Kṣurika,

Tejobindu, Yogaśikhā, and Yogatattva Upaniṣads provide certain images that

supplement the descriptions of the Amṛtanāda Upaniṣad forming a picture that is

more complete if still unsystematic.14 We will focus in this section on these six

14 BVU, DhBU, KṣU, TBU, YŚ, and YTU.

215

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

Yoga Upaniṣads and the Hamsa Upaniṣad, after which we will turn in the following

section to an analysis of the Amṛtanāda Upaniṣad.

The Hamsa Upaniṣad represents a special case, as it adopts the standardized

system of six cakras from the Tantras of the Yogini Kaula and the Western

Transmission.15 The six cakra system developed from Upanisadic, medical, and

āgamic sources.16 Variant systems of yogic body centers and pathways predate

this system.17 The sixfold system is first mentioned in the eighth century drama

Mālati Mādhava of Bhavabhūti. The term cakra probably first occurs in the

Kaulajñānanirnaya (ninth century), which provides the earliest extant systematic

presentation of the six cakra system. From the Yogini Kaula, the system was

15 It is also likely that the verses containing the six cakra terminology are later additions to

this text. See chapter 3, section 3.2.7. The HLU does not use the term cakra, but does employ the

standard names, ādhāra, svādhiṣṭhāna, maṇipūraka, anāhata, viśuddhi, and ājñā, found in the tantric

and later yogic systems.

16 The “heart” and “veins” appear in brief references in the classical Upaniṣads (see chapter

2). The medical texts describe the notions of the heart, pathways (siras, srotas, and dhamani), vital

points/joints (marmans), and other bodily constituents and winds. These are explained in the

major medical texts. The three major medical texts are the first century Carkasamhita, the fourth

century Suśrutasamhitā, and the six to seventh century Aṣṭāṅgasamgraha of Vāgbhaṭa the Elder. For

dates of these titles, see White (1996), p. 52. The Aṣṭāṅga Hṛdaya of Vāgbhaṭa the Younger,

Śārṅasthānam 3 and 4 elaborates veins and vital points. In general, see White, “Yoga and Early

Hindu Tantra,” throughout. For elements from the Saiddhāntika influences (Āgama) in

Matsyendra's Kaulajñānanirnaya, see White, p. 9; Sanderson (1996), p. 35 and his citations. None of

these earlier sources contains fully developed yogic systems. Tantric ritual and meditation

traditions borrowed from these traditions and systems for the creation of their yogic bodies.

17 See White, “Yoga in Early Hindu Tantra,” p. 2 and throughout. Buddhist Tantras

mention four-center systems, such as in the eighth century Caryāgitī and Hevajra Tantra. The eighth

century Hindu Bhāgavata Purāṇa enumerates six “places” that correspond to the later six cakra

system. The ninth to tenth century Kaulajñānanirnaya presents a systematic presentation of six and

eleven cakra systems. The tenth century KMT of the Western transmission contains systems of five

and six cakras. See also Snellgrove (1988) section III concerning the Buddhist systems. See also

Heilijgers-Seelens (1994), throughout for the five cakra system in KMT 14-16.

216

History, Text, and Context of the Yoga Upanisads: Chapter 4

adopted into the Kubjikā tradition of the Western Transmission. It was

subsequently adopted by later Kaulism, for example in cult of Tripurasundari, and

by the Nāth haṭha yoga traditions of Gorakhnāth. Although it has been often

assumed that the six cakra system was integral to the various tantric systems, this

is incorrect. Even so, the six cakra system did become more widely known over

time, through texts such as the northern Hamsa Upanisad, late haṭha yoga texts,

early modern tantric compilations,, and the southern Yoga Upanisads. In this way,

these texts "over-wrote" much of the diversity of the earlier multi-vocal

traditions.

The twelfth to thirteenth century northern Hamsa Upanisad is a transition

text between the two collections, northern and southern. Most of the northern

Yoga Upanisads acknowledge some system of the body's internal pathways, the

nāḍis. Channels of the vital powers (prāṇa) and breath (also termed prāṇa) appear

to constitute the oldest and most widely shared aspect of esoteric physiology in

18 These texts are derivative from the Nāth Haṭha yoga texts. They are the fifteenth century

Haṭha-Yoga-Pradīpikā, the late seventeenth century Gheraṇḍa Saṃhitā, and the late seventeenth to

early eighteenth century Śiva Samhitā.

19 Such as the late Mahāniroāṇa Tantra.

20 See White, "Yoga in Early Hindu Tantra," throughout. See also Sanderson (1986), p. 164

and Flood (1993), pp.256ff. This issue is further examined in this chapter, and in chapter 5 as

Western scholarship also contributed to the standardization of the six cakra system as THE system

of cakras. The HU could have interpolated the six cakra system from any number of texts. The HU

probably did not adopt the six cakra system directly from the Yoginī Kaula or the Western

Transmission. It seems more likely that the system came from later intermediary sources.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

Indic systems.21 The later medieval systems all agree that there are three primary

nāḍīs: the left channel (iḍā), the right channel (piṅgalā), and the central channel

(suṣumnā).22 The central channel is most important in these systems, as it is the

path to liberation. Several of the northern Yoga Upaniṣads mention or allude to

these three channels. Numerous other channels are also acknowledged but

without being named.

Brahma Vidya Upaniṣad 11-12 describes the oṃ mantra penetrating the sun-like

central channel that breaks through the head of the yogin and brings immortality.

This is a cosmological allusion to the classical Upaniṣads’ worldview. In this

worldview the realm of human cycle of birth, suffering, death, and rebirth is the

visible earth, air, and the vault of the sky. The sun is represented as an exit

doorway through which the ātman can escape phenomenal existence to the eternal

realm of Brahman. In the Yoga Upaniṣads this cosmological system is condensed

into the microcosm of the yogic body. The yogic body is partially modeled on this

Upanisadic idea of the cosmos. The classical Upaniṣads had already experimented

with such microcosmic interiorizations that are further elaborated in the Yoga

Upaniṣads. The Tantras also adopt and elaborate this Upanisadic cosmos and its

interiorization. From this example alone, it is unclear whether the Brahma Vidya

21 See chapter 2.4 for classical conceptions of the inner pathways.

218

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

Upaniṣad is extrapolating from its classical predecessors or from the yoga of the Tantras.

The picture is somewhat clarified by the references in the Brahmavidyā Upaniṣad to the “conch.” The Brahmavidyā Upaniṣad depicts the phones of om as seated in a “conch.” Deussen suggests that the conch is the brain, but the text does not describe the location.23 This observation is consistent with the use of the term in medical parlance. The medical descriptions of the body include two vital points (marman) above the ears that are called śaṅkha, conch shells.24 These are the in the region of the temporal bones. The term conch was in use for these points and for the frontal bone, the bone of the forehead.25 The grammatically feminine related term śaṅkhini is important to subtle body constructions in the Tantras and Nāth Siddha texts where, it is the name of a nāḍi possessing doors in the cranial vault and at the base of the body; it is also a name for kuṇḍalini.26

In the Brahmavidyā Upaniṣad, the three morae or phones of om are homologized to the moon, sun, and fire. These are also common homologues for the three nāḍis and for certain maṇḍalas in the head. The Amaraghasāṣana,

22 Thousands of nāḍis are mentioned as early as the classical Upaniṣads (72,000 being a common number). Although the three primary channels are especially privileged, tantric and yoga texts developed systems of ten, twelve, or fourteen important pathways.

23 Deussen (1997), p. 669.

24 These and other such marmans are enumerated in Vāgbhaṭa the Younger’s Aṣṭāṅgahrdayam, Sārtrasthāna 4.1ff.

25 Monier-Williams (1988), p. 1047.

26 Silburn (1988), pp. 124ff; White (1996), pp. 254f.

219

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

attributed to Goraknāth, describes the brain case as containing the fire, sun, and

moon mandalas. Silburn analyzes the appropriate verses from the

Amaraughaśāsana:

The bindu, a point of concentrated power representing virility and

situated above the center between the eyebrows, explodes and

expands, thus giving birth to the mastaka and brahmarandhra. The

orifice at the base of the linga, shaped like a small bell (ghanṭikā), is

the rājadanta or elephant tusk; this is also the aperture of the

śankhinīnādi [synonyms for the uvula or the roof of the palate]….

The braincase contains three circles: the fire, sun, and moon

mandala. In the middle of the latter, somamandala, a stream of

nectar, released by the rupture of the linga, starts to pour into the

cranium; then the śankhini, whose tenth door opens into the

thousand-petalled lotus (the sahasrāra), draws this nectar from the

moon circle, and gaining control over it infuses it into the median

way.27

Silburn furthermore ties these images to the practice of khecarī mudrā. This

practice, through blocking the two side channels, cause the sixteenth kāla to move

upward to the place of the triangle, the three-pointed thunderbolt, and the om

mantra. This description employs the specialized vocabulary of the Nāth texts, but

it also provides a connection to the Upanisadic images by linking the sun, moon,

and fire with the head and with the om mantra.

The Amaraughaśāsana contains many more details than the Brahmavidyā

Upaniṣad, but both texts are describing a similar picture of the inside of the yogin’s

head. In the Brahmavidyā, the pointed flame (śikhā, the half mora, the buzzing

27 Silburn (1988), p. 131. See also White (1996), p. 231.

History, Text, and Context of the Yoga Upanisads: Chapter 4

reverberation) is described as penetrating the sun-like channel28 that breaks

through the head, which recalls the Amaraughaśāsana's description of the stream of

nectar pouring in the cranium and becoming infused into "the median way." The

metaphorical fields and terminology are different in the two texts, but both

present the following picture. For both there is a conch in the skull. Both texts

describe a process that occurs in the area between the temples or behind the

forehead. Both constructions appear to see the sun, moon, and fire as present in

the cranium, and both represent penetration of the channel in the skull as leading

to blessing or bliss.

One difference between the two traditions is especially interesting: the

privileging of the sun in the Brahmaoidyā Upaniṣad, and the privileging of the

moon by the Nāths. The Nāths seek, as part of their practice, the rain of lunar

nectar that would spill from the upper conch because of this type of meditation.

The northern Upaniṣads do not use the image of the lunar nectar—although the

evidence is not conclusive because maṇḍala, as "orb," can refer to the moon as

easily as it can refer to the sun (or any orb-like celestial body). The images in the

northern Yoga Upaniṣads are vague, but generally solar imagery is privileged,

especially regarding the maṇḍala in the head or at the top of the head.

28 This characterization is unclear. It could refer to the right channel or the central channel.

The sun is usually associated with the right channel. The logic of the verse expects that the

emphasis would be on the central channel, and this is how Deussen translates the verse.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

It is not yet possible to determine a specific relationship between the

northern Upaniṣads and the Nāth texts. The dates of the Amaraughaśāsana are

uncertain. The text was written before 145029 and is attributed to Gorakhnāth,

who probably lived in the twelfth to early thirteenth century. Thus, the text was

most likely written between 1150 and 1450. The title Brahmavidyā Upaniṣad is

attested between the ninth and eleventh centuries, although its extant manuscripts

date to later centuries. The pertinent descriptions in the Brahmavidyā Upaniṣad,

along with similar materials in the other northern Yoga Upaniṣads, appear to be

meditation visualizations, as they are in the earlier Buddhist and Hindu Tantras.

With the haṭha yoga developments promoted by Gorakhnāth and his followers,

what were visualizations for the earlier Tantras take on the characterizations of

actual body structures, even if subtle.30 The Brahmavidyā Upaniṣad appears to rely

on classical Upaniṣadic models while at the same time employing language and

images that correspond to the yogic practices of the Tantras that are later

elaborated by the Nāths.

The primary image of the Brahmavidyā Upaniṣad is a tantric one, even if it is

laden with Vedānta symbolism. In a conch between the temples, the phones of om

shine like the moon, sun, and fire. The bindu of om is a pointed flame burning

29The text is quoted by the HYP. The HYP dates to approximately 1450.

30These concretizations of yogic-body-visualizations into subtle body structures is further

examined in the following sections that focus on the southern Yoga Upaniṣads.

222

History, Text, and Context of the Yoga Upanisads: Chapter 4

above the conchshell like a flame above a lamp. Here we see the resonance of om

as a śikhā burning at the peak of the yogin’s head. As we have seen, śikhā provides

a double metaphor for this image, as it means both “peak” and “pointed flame.”

We find a similar image in the Tantras.

This peak is located at the end of the stream of divine sound (nāda)

which resounds throughout the micro/macrocosm until it merges

into the Silence of the Transcendent at the highest level of being at

the End of Sound (nādānta) in the Twelve-finger Space. The

goddess is therefore said to reside on the peak of mount Kailāsa

and, as such, is the Goddess of the Peak (śikhādevī) and the Mistress

of the Wheels of Energy, which revolve in the cosmic body.31

In Kaulajñananirnaya 17.4 we find a list of the eleven centers of vital breath ending

with the brahmarandhra which is described as “brilliant” or “‘shining” at the peak

(śikhara) (of the head). This imagery resonates with the Brahmavidya Upaniṣad’s

description of the pointed flame burning above the brahmarandhra, an image also

found in the aptly named Yogaśikhā Upaniṣad.

The Kṣurikā Upaniṣad also contains images of the yogic body that are both

Upanisadic and tantric. The Kṣurikā Upaniṣad describes the central artery,

suṣumnā, as originating at the region of the navel. The central artery is first

mentioned by name in the classical Maitrāyanīya Upaniṣad. In the Kṣurikā Upaniṣad,

31 Dyczkowski (1989), pp. 91-92 and notes, summarizing from the Manthānabhairavatantra,

Yogakhanda 76a and 80b. Dyczkowski further confirms that the Western Transmission texts

repeatedly refer to the reverberation of mantras (but not always om), and the force later called

kuṇḍalini the term “sikhā,” the pointed flame that burns at the peak of the yogin’s skull. Personal communication, (2001).

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

the yogin must slip into this channel with the breath by means of the "razor or

knife of concentration," the meditation technique taught in this text. This

terminology resembles the terminology of the "bolt practices of the knife" taught

in the Kubjikāmata Tantra.32 These practices, taught by the fierce tantric Bhairava,

are ritual actions of literal cutting or piercing of body parts. Subsequently, there

are offerings of the practitioner's blood and flesh and other substances to the fierce

goddesses of the Western Transmission. These offerings are accompanied by

mantra recitations. In addition to these external rituals, the Kubjikāmata Tantra as

well of the earlier Kaulajñananirnaya, also contains internal piercing practices that

are yogic. It is precisely such tantric rituals that are internalized in the practices

and meditations of hatha yoga. The metaphor of the knife to designate the practice

of concentration derives from the use of real metal knives for cutting knotted

ropes or for sacrificial butchery. This is the general historical trend with yogic

interiorization. External rituals provide visual images, technical vocabulary,

sounds, and other stimuli that are the raw materials for images, metaphors, and

symbols of visualization and meditation.33

32 For this and the following descriptions of the Kaulajñananirnaya and KMT, see White,

"Yoga in Early Hindu Tantra," pp.5ff.

33 In any fantastic system, the patterns of influence can move in both directions. On the

one hand, rituals and day-to-day life can create images for metaphorical and symbolic usage. On

the other hand, imaginary and symbolic systems can influence rituals and procedures. The

primary metaphorical foundations of Indic religious life generally run from the literal to the

figurative. That is, butchery, wells, spiders, looms, oil, libation, fire, rivers, roads, and other

images from daily and ritual life are the basis of images of fierce and benevolent gods and internal

224

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

There are additional similarities between the Kṣurikā Upaniṣad and these

earlier Tantras. The phones or morae of oṃ (twelve divisions in this text) fix the

breath in different body parts: big toes, ankles, and calves. This practice is similar

to nyāsa, the yogic ritual used to enliven of certain parts of the body, either of a

person or a consecrated image. The practice follows with fixing the breath with

phones in the knees, thighs, anus, and penis. Afterward the yogin enters the

navel, the location of the breath. Here the guided breath slips into the central

channel. The breath then moves up to the heart, called the "little red lotus

flower." Penetrating through the heart, the breath ascends to the neck and seizes

the mind-razor therein. From the neck, the yogin descends the body, cutting the

vital points (marmans) throughout. The term "knot" is not used in the text, but the

vital points and channels (nāḍis) of the body are represented as clogged or

knotted. The texts of the Trika Kaula employ similar imagery, in which the body

centers (cakras) are tangled coils in non-practitioners, because they knot together

spirit and matter.34

The imagery that follows in the Kṣurikā Upaniṣad is vague, terminology is

unclear, and the text contains incorrect grammar. However, the imagery appears

to describe the return of the concentration-razor to the neck where all the severed

geographies. The evolution of yogic interiorization likely involved many trance practices and

ecstatic experiences, yet the "lotus of the heart" or the "inner libation" are signs from the external

world, not from some inner yogic eye.

225

History, Text, and Context of the Yoga Upanisads: Chapter 4

nādis provide a "support" or "pillow" that should be seen (visualized) as

resembling jasmine flowers.35 Unlike the importance of plantain or lotus flowers36

in other body imagery, the emphasis with this metaphor appears to be based on

the habit of jasmine stems: the jasmine is a multi-stemmed, climbing, vine-like

plant with many leaves and small fragrant flowers. Thus, the region of the neck is

figuratively full of tangles of nāḍi stems so thick that they look like a perfumed

pillow of jasmine vines.

The Kṣurika Upaniṣad continues with emphasis on the central channel. The

central channel is not cut by the razor practice. Then follows the statement that

the hamsa flies free. The text contains conflicting images and mixed metaphors.

However, it follows the model of the Tantras that blockages in the body—blocked

joints and tangled channels—are fetters that keep the essential Self (the hamsa)

trapped in the body. The image is of a bird tangled by vines, cords, or tethers.

34 White, "Yoga," p. 5, and notes. See Silburn (1988), pp. 25-35.

35 The ambiguous references to the central channel as the "pillow" (see chapter 3, 3.2.8) and

the practice of "stuffing" the pillow with karmic residue of all "good and bad states" is similar of

the tantric descriptions of filling the channel with the white and red drops (bindu) found in the

tenth or early eleventh century Buddhist Kalacakra Tantra (KT). See Daniel Cozort (1986), p. 127. In

our Upaniṣad, there is no sexualized symbolism, but this parallel demonstrates the commonality of

'stuffing' meditations relative to the esoteric channels. In the eighth century literary Mālati-

Mādhava of Bhavabhūti contains a similar phrase, "by the gradual filling of the channels,"

(nāḍīnāmudayakrameṇa). See White (199x), p. 11 quoting from Kale's translation.

36 In my opinion the plantain and the lotus are important images because of the habit of

their flowers. Both flowers are considered beautiful by aesthetic standards and both grow calix

down. They bloom by turning upward. The lotus also has other remarkable qualities that provide

potential images for human metaphor makers: many species close at night or during rain, then

reopen; they grow hidden in muddy bogs, etc.

226

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

When the cords are severed the bird-self flies free, symbolic of the yogin attaining immortality.

The Kṣurikā Upaniṣad, as discussed in chapter 3, dates to between the ninth and eleventh centuries, and was thus written around the same time as the Tantras to which it is similar. However, the text also contains primitive terminology—in particular nāḍi and marman—that predate the Tantras. When it describes the "little red lotus" and the central channel, the Kṣurikā Upaniṣad states that these are described in the Vedānta texts—an allusion most likely to the classical Upaniṣads or their commentary traditions.

The text's overall imagery is mixed and sometimes abstruse. Such imagery is much more consistent, complex, and elaborate in the Tantras. This suggests at least two possibilities. If the text was written during the ninth century, then it was part of the milieu that produced the Tantras, and as such, would represent apparently widespread methods of meditation that were elaborated in much greater detail in the Tantras. If the Kṣurikā Upaniṣad is an eleventh century text, then it is heir to these more complex systems of the Tantras. This latter interpretation seems more likely because the text's imagery and mixture of technical terms do not read as early formulations of later tantric notions, but rather as summaries of and allusions to various tantric theories brought together in this short text.

227

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

The Kṣurikā Upaniṣad (verse 16) mentions the iḍā and piṅgalā by name but concentrates on the central channel as the path to liberation. These technical terms predate the Kṣurikā Upaniṣad. There were numerous Tantras during this period that elaborate and explained the type of meditation described in this text. Among the other Yoga Upaniṣads, only the Brahmavidya and Cūlikā Upaniṣads date to this same ambiguous period (ninth to eleventh century), but these texts do not contain these types of meditation details. Thus, the Kṣurikā Upaniṣad would be the only known Upaniṣadic example of this kind of meditation that might possibly predate similar materials in the Tantras. However, the text's descriptions of this razor-meditation and intermixing of images make it more likely that the Kṣurikā Upaniṣad derives from the Tantras, not the other way around.

The Dhyānabindu Upaniṣad does not mention the primary nāḍis by name but does allude to them by their common homologues, the sun, moon, and fire.³⁷ Verse 22 is even more explicit: "Using the half phone [the nādabindu of om] as a rope, draw out of the well³⁸ of the heart-lotus the manas upward along the path of the nāḍi between the brows (bhruvor madhye), where it melts away."³⁹ This text

³⁷ See White (1996), throughout chapter 2 for the symbolism of moon, sun, and fire. See also Silburn (1988), p. 131.

³⁸ See White (1996), pp. 242ff, for the complex symbolism of the "well" in yogic and alchemical texts.

³⁹ Translation adapted from Deussen (1997). Southern ALS reads differently than the northern BI edition. The ALS (v. 39) does not employ the metaphor of water drawn from a well. It has the half mora raised from heart to between the eyebrows by means of "garland of letters"

228

History, Text, and Context of the Yoga Upanisads: Chapter 4

does not depict the escape of the breath-self through the cranial fissure. Instead, it

recommends that the mind (manas) reside in the forehead at the root of the nose,

called the "great resting place." This is an important distinction, because the

Upanisadic model of liberation generally has the breath-self escaping the body

through the cranial fissure. Again, one can turn to the Tantras for explication.

Several of the Tantras and later yogic texts recognize the importance of the

forehead. The place, between the eyebrows at the root of the nose, is the place of

esoteric importance in the Kaulajñananirnaya, Vijñanabhairava Tantra, Kubjikāmata

Tantra, as well as most of the other Hindu Tantras that describe the esoteric body.40

It possibly designates the same location as the śaṅkha of the Brahmavidyā

Upaniṣad—or, more precisely, the śaṅka is inside the head behind the forehead.

Some of the other northern Yoga Upaniṣads mention the breath ascending

through the body to the head or through the head, but they generally do not

describe its path(s). The Yogaśikha Upaniṣad is of particular interest in this context

because it portrays (in verse 7) the breath as following a "zigzag" path: "Then

zigzag the Self [as mind or breath] goes upward through the shining door of the

(mālamātreṇa) a reference to meditation on the letters of oṃ, or meditation on the letters of the

Sanskrit alphabet.

40 With the same characterization in Kaulajñananirnaya 17.3. See White, "Yoga," p. 4;

Silburn (1988), pp. 28, 67; Padoux (1992), p. 144f. This is also the location of the Ājñā cakra in the

systematized version of the six cakra system.

229

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

suṣumnā; breaking through the cerebral dome, he finally sees the highest one."41

The "zigzag" path of the breath resembles the early references to the force that is

later called kuṇḍalini: "impel the crooked on upward,"42 "where rests the coiled

one," "coiled three or three and one half times."43 Although the Yogaśikha

Upaniṣad is not explicit about what is happening in the yogic body, the śikha and

prāṇa play a role later in the text similar to that assigned to the serpent power,

kuṇḍalini-śakti, in the early tantras.

A number of northern Yoga Upaniṣads allude to the presence of esoteric

"centers" in the yogic body. The Kṣurkiā Upaniṣad refers to nāḍis, and its order of

meditation follows the sequence of navel to heart to throat. It refers to the heart as

the "little lotus flower" that the Vedānta texts describe. Otherwise it mentions the

marmans that must be cut by the razor of concentration. The imagery of the text is

of a body full of channels that are blocked at different vital points. Meditation

must cut through these vital points for the yogin to attain immortality. The text

41 This observation requires more research due to the variations in the texts. See Deussen's

translation, (1997), p. 710. As with the DhBU, the ALS reads differently than the northern texts

printed in the BI. The ALS (v. 75a-76b) reads bhindanti yoginah surya yogābhyāsena vai punah;

dvitīyam suṣumnādvāram pariśubhram samarpitam. kapālasampuṭam pitṭva tatpa tatpadam. "By

practicing yoga again and again, yogins pierce the sun; accompanied by opening the shining door of

the suṣumna. Having drunk from that skull bowl, he envisions that realm." Also see Mahābhārata

5.33.52 for a similar statement, which Sontheimer mistranslates in Pastoral Deities, p. 203.

42 KMT 23.114a, see White "Yoga," p. 7.

43 This last example is of interest because it parallels the three or three and one half

divisions of om that are repeatedly explored in the Yoga Upaniṣads.

230

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

does not refer to any specific "centers," beyond the central significance it ascribes

to the navel, heart, and throat.44

The Amṛtabindu Upaniṣad describes the annihilation of the manas in the

heart but does not otherwise refer to any yogic body locations or structures. The

Dhyanabindu Upaniṣad provides flower and god allofoms for body locations and

also relates them to the breathing practices of inhalation, holding, and exhalation.

Inhalation is associated with the navel, flax flower, and Viṣṇu. Holding the breath

is associated with the heart, lotus, Brahmā, and the colors red and white.

Exhalation is associated with Three-Eyed (Śiva) in an eight-petaled45 plantain

flower in the forehead. A flower of one hundred petals46 is the location of "him"

who is beyond fire, moon, and sun—although the location of this flower is not

specified. The text thus describes a flowers-as-centers system that is somewhat

like the more developed cakra systems with associated lotus flowers, gods, and

colors. This simple system appears to map navel > heart > forehead > beyond (?).

The last two verses of the Dhyānabindu Upaniṣad present another perspective on

the yogic body, describing the mānas, by means of the half phone, as ascending

44 The text refers to several body parts, which are below the throat, as the objects of the

"cutting" concentration. The toes, ankles, etc. are all "cut off" because they are the locations of

tangles, or knots: the marman (vital organs). I do not think they have as specialized meanings as do

the navel, heart, or throat.

45 Flowers of eight leaves and of eight petals are found throughout the Tantras. In the later,

standardized six cakra system, the eyebrow center has a lotus of two petals.

231

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

from the heart and dissolving between the eyebrows, at the root of the nose. Thus,

the Dhyānabinḍu Upaniṣad provides two body-maps: first, navel > heart > forehead

beyond; and second, heart > forehead.

The Tejobindu Upaniṣad mentions the heart and three gates. Otherwise, it is

silent on references to the yogic body. The Yogaśikha Upaniṣad, as discussed

earlier, describes the śikhā (pointed flame) as breaking through the head. The

Yogatattva Upaniṣad associates the mind, heart, and lotus and describes the a, the u,

and the m of om as opening the heart lotus. This image is reminiscent of the

Tantras, in which the meditations and mantras serve to activate the body-center

lotuses. When this occurs, the down-pointing lotus turns upward and opens its

petals. Yogatattva Upaniṣad 14 refers ambiguously to the yogin breaking through

the lotus and being led by the wind to the place between eyebrows and forehead.

We find here the same model as in the Dhyānabinḍu Upaniṣad, where the forehead,

between the eyebrows, is the destination of the meditation wind.

4.3. The Mysterious "Gates" of the Amṛtanāda Upaniṣad

The preceding examples from the northern Yoga Upanisads depict body

systems as flowers in four locations: navel, heart, forehead, and beyond (probably

46 Kaulajñānanirnaya 5.27 mentions a cakra with one hundred leaves, but without location.

The DhBU one hundred petaled flower appears to be the highest center, thus equating it with the

later designation of the highest center, the "thousand petaled lotus."

232

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

the cranial fissure). The Amṛtanāda Upaniṣad contains body-center references, but

it does not use the flower imagery, and it also does not employ nāḍi terminology.

The text describes the breath as moving along a path or way (mārga) but otherwise

no channels or pathways are described. The Amṛtanāda Upaniṣad is concerned

rather with visualization of "gates" or "doorways" (dvāras) on this path. These

concepts resemble in many ways the cakras of tantric systems, although, as

discussed earlier, the term cakra is not used in any of the northern Yoga Upaniṣads,

other than the Hanṣa Upaniṣad.

The Amṛtanāda Upaniṣad provides a list of special "doors" inside of the

yogic body.⁴⁷ Amṛtanāda Upaniṣad verses 26 is one of the least understood passages

in the Yoga Upaniṣads. Every translator has interpreted this verse differently, and

the medieval and early modern commentaries are not especially helpful.⁴⁸ The

meaning of this important verse is ambiguous. In the following pages, I will

provide several alternative translations and analyses of this verse.

The verses that precede 26 are straightforward in their recognition of these

doors existing along the path (mārga) of the vital breath. These verses describe

what the yogin sees while practicing the voiceless or silent om: "By means of the

⁴⁷ Since the ANU is nearly identical in both northern and southern recensions, I have

consulted all translations of the text for the following analysis. See chapter 3, section 3.2.3 for

summary.

⁴⁸ ANU 26 (or 27), see Deussen (1997); Aiyar (1997); Ayyangar (1952); Varenne (1971);

Feuerstein (1998). The verse is translated as v. 27 in some of the translations.

233

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

sound [om] he sees that path along which the prāṇa goes; for this reason he should

constantly practice going on that very path." Next follows verse 26. The verse

reads as follows in the Adyar Library Series and Aiyar Sanskrit editions:

hrddvāram vāyudvāram ca mūrdhadvāram athāparam |

mokṣadvāram bilaṃ caiva suṣiraṃ mandalaṃ viduḥ || 26 ||*

Continuing from the previous verse, these gates are on the path of the vital

breath. The later verses say that this path leads to liberation: "no matter where he

dies, he is not reborn." There are several possible translations of verse 26.

Translation 1:

The heart-gate (hrd-dvāra) and the wind-gate (vāyu-dvara),

next the head-gate (mūrdha-dvāra) having nothing beyond

[it]; In this way, they know the passage of liberation (mokṣa-

dvāra) as the cave (bila), the hollow (suṣira), and the circle

(mandala).

This reading of the passage interprets it to mean that there are three esoteric gates.

Each gate is named twice in the text, in which a descriptive term is associated with

a secret. For example, the gate of the heart is called the "cave." The system of the

text could then be mapped as follows:

gate of the heart = the cave (bila)

[heart]

gate of the wind = the hollow (suṣira)

[throat]

• Key terms: dvāra (n. door, gate, passage opening, aperture of the human body); hrd-dvāra

(gate of the heart); vāyu-dvāra (gate of wind); mūrdha-dvāra (gate of the head, gate of the upper-most

/ top-most); mokṣa-dvāra (the liberation passage, the gate of liberation); bila (n. cave, hole, opening,

aperture); suṣira (n. hollow, hole, cavity); mandala (n. circle, disk); viduḥ (a. from vid, they call them,

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

gate of the head = the circle (mandala) [cranial suture]

This reading assumes that the term mokṣadoāra means the "passage to liberation."

As such, it is a collective name for the three gates, or it is a nāḍi-like passage that

connects the three gates.50

Equating the heart with a cave (bila) is consistent with older associations.

The heart is referred to as the cavity or cave (guhā) in the classical Upaniṣads. The

wind-gate and the hollow (suṣira) are not common characterizations, but these

images evoke logical associations with the throat. The Kaulajñānanirnaya refers to

the throat as the cakra of breath and utterances.51 Alternatively, it is possible that

the hollow refers to the navel area, an association that would bring it into

agreement with the importance of the navel in other esoteric physiologies. Wind

and hollow could thus equally characterize either the navel or throat. If the term

suggests navel, then the gates are not listed in logical order. This may be of no

consequence, but the Tantras and yoga texts usually present a sequence of internal

structures in ascending order, as in the six cakra system.52

which they know as; bhu from vidu, adj., intelligent, wise; m., the hollow between the frontal globes

of an elephant, i.e., the top of an elephants head).

50 It is possible that this reading describes a four "gate" system in which the first three

gates are named twice and there is no synonym for the liberation-gate.

51 White's translation from Kaulajñānanirnaya 17.3a. "Yoga," p. 5.

52 We find a notable exception in the Buddhist system described in Hevajra Tantra (HT)

II.iv.54-55, where the "centers" are not arranged in normally ascending order. See David

Snellgrove (1987), p. 251. The ninth to tenth century Kaulajñānanirnaya lists of seven and eleven

cakras follow linear order, as do both the six and five cakra lists of the KMT.

235

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

There is another possible interpretation of the esoteric path of the breath, if

the wind gate is the navel instead of the throat. This model would have the breath

coming in through the nostrils and then resting in the heart, consistent with

representations found in the classical Upaniṣads). The yogin’s science would then

move the breath downward from the heart to the navel. There the breath would

enter the “gate that leads upwards” (mūrdha-dvāra). From the navel-gate breath

would move upward to the head and beyond. From this perspective, the gates

would not be in a nonlinear order.

Calling the gate of the head a circle (mandala) is consistent with verse 38 of

the text: “That man is never reborn wherever he may die, whose vital wind

escapes at the top of the head after piercing through this mandala. That man is

never reborn.” There are also variant readings of verse 38 that do not support this

analysis. For example, Feuerstein translates verse 38, “He whose [life force],

having broken through this region (mandala) of the wind, rises to the head.”53

Deussen translates the verse “In whom, breaking through this ring [mandala], the

vital breath ascends to the head.”54 Both of there translations read the verb yāti (i =

“to go”) plus locative mūrdhani to mean: the breath “goes into the head.” in the

sense of rising to or ascending to the head. I suggest an alternative reading:

53 See Feuerstein (1998), p. 420. “Vital wind,” mārutah is nom. case, making Feuerstein’s

translation unlikely.

54 Deussen (1997), p. 698.

236

History, Text, and Context of the Yoga Upanisads: Chapter 4

"goes/escapes at the top of the head." I understand mūrdha in this context in the

sense of topmost, roof, uppermost, peak, not simply as head, forehead, roof of the

palate, or skull.55

We have seen two models of liberation in the Yoga Upanisads. The older

model follows the classical Upanisads and maintains that liberation occurs when

the breath goes out/escapes at the topmost part of the head. However, a later

tantric model also appears in the corpus of Yoga Upanisads—for example, the

notion found in the Dhyānabindu Upanisad, that the forehead, or somewhere inside

the head, is the location where the breath-self attains bliss or liberation. Thus, all

of the following phrases can make sense in either an Upanisadic or tantric context:

(1) "The vital air enters into the head;" (2) "The vital air moves into the

uppermost;" (3) "The vital air goes into the roof of the palate;" (4) "The vital air

enters the forehead." (5) "The vital air escapes at the forehead;" (6) "The vital air

escapes at the topmost part of the head."

Another interpretation of verese 26 appears in translation 2:

The heart-gate, the wind-gate, and the head-gate; next

moreover is the liberation gate, which they know as either

the cave, the hollow, or the circle.

55 All of these readings are normative usage of the locative (in, on, at, into, onto), and

mūrdha includes the meanings topmost, roof, uppermost, and peak as well as head, forehead, roof

of the palate, and skull.

History, Text, and Context of the Yoga Upanisads: Chapter 4

This second plausible reading is that there are four gates. In this reading the

fourth gate, moksa-dvāra, is the only gate that is renamed. The text is signaling that

"they know it" by three different terms: cave, hollow, and circle. Such a system

can be mapped as follows:

gate of the heart [heart]

gate of the wind [throat]

gate of the head [forehead]

gate of liberation = cave, hole, circle [cranial suture]

As in Aiyar's reading of the verse,⁵⁶ the cave, the hollow, and the circle may

all be alternative names for the gate of liberation. This translation gives virtually

no meaning to the terms atha apara, which is possible at times when such words

are used as expletives or to fulfill the requirements of poetic meter. This would

mean that apara is not used in the sense of "having nothing beyond or after," but is

rather either untranslatable or means something like "moreover." Like the

preceding interpretation, the wind-gate could refer to either the navel or of the

throat. I read this as a four gate system: heart, throat, forehead, and cranial suture.

According to this reading, this fourfold system ignores the lower body centers,

particularly the navel. This is conceivable since the oldest Upanisadic models

focus their attention on the heart as the location of vital breaths and the beginning

⁵⁶ Aiyar (1997), p. 165.

238

History, Text, and Context of the Yoga Upanisads: Chapter 4

place of channels, breaths, and senses. Early Buddhist tantric systems have only

four centers, although these are usually navel, heart, throat, and head.57

A third interpretation of verse 26 appears in translation 3:

The heart-gate, the wind-gate, and the head-gate; next

moreover is the liberation-gate: these are the same as (eva)

the cave, the hollow, the circle, and the elephant's crest.

Translation 3 departs from the standard interpretation of one of the words of the

verse, reading viduh not as the verb vid ("to know") but rather as the masculine

noun vidu, "the hollow between the frontal globes of an elephant" (i.e., the

forehead or the top of an elephant's head). This reading is grammatically

problematic but anatomically sensible. If this verse were read as a nominal list,

standing alone, instead of as an accusative list connected with verse 25, then

translation 3 presents a fourfold system that can be mapped as follows:

The gate of the heart = cave

[heart]

The gate of the wind = hollow

[throat]

The gate of the head = circle

[forehead]

The gate of liberation = elephant's crest

[cranial suture]

57 Navel, heart, throat, and head are the locations of the "nerve centers" for many of the

tantric Buddhist systems. If the ANU privileges three or four 'centers' instead of seven, then the

system of this Yoga Upanisads is more like Buddhist four-"center" systems then it is to the five, six,

or more centers of the Hindu Tantras. For the Buddhist four-center systems, see Snellgrove (1987),

p. 248, Dasgupta (1950), pp. 163ff, and Eliade (1973), p. 244, passim.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

The image of the elephant's crest is not found inother yogic or tantric texts as a characterization of an esoteric center of the subtle body. There is nothing conclusive to suggest this reading, yet it is consistent with other esoteric body systems.58

This does not exhaust the possible interpretations of this verse. Other scholars have translated verse 26 in different ways. Aiyar translates the verse as follows:

It is through the opening (or hole) of the heart, through the opening of vāyu (probably navel), through the opening of the head and through the opening of mokṣa. They call it bila (cave), sushira (hole), or mandala (wheel).59

Aiyar's reading is sensible, but even in English the meaning is ambiguous. "They call it" does not specify an antecedent but suggests that the "opening of mokṣa" be variously called the cave, hole, and wheel. Aiyar's translation is possible, but his notes presuppose that there are four openings in the body, suggesting that his reading was influenced by the commentarial tradition and his

58Although it seems unlikely, I have provided this third reading because there may be other "elephant" imagery in the Tantras that has not been recognized or carefully examined. Many Sanskrit terms do have secondary meanings related to elephants. The Nāth texts sometimes refer to the uvula as the "elephant's tooth." The importance of the uvula for these yogic systems appears as early as the Maitrī, although without the elephant terminology. It is not conclusive, but Kaulajñānanirnaya 6.23 uses the term "rājada" (rājada[anta]?) in a way similar to that of the later Nāth usage. Rājada does not have any apparent meaning, although it could be a name. I am inclined to read this as rājadanta. It may simply mean the eyeteeth here, as the verse says that the bindu stands between the two rājada. This may be an alchemical reference where in a "pill" is held in the mouth, under the tongue (= at the front teeth, between the eyeteeth, etc.) to generate the amṛta nectar. See Kaulajñānanirnaya 6.23-25.

59Aiyar (1997), p. 165. Parenthetical comments original to Aiyar's translation.

240

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

expectation of a four-opening system. Thus, such a reading may or may not be

accurate and, at best, is ambiguous like the original text. In addition, he suggests

that the wind-opening is probably the navel. This interpretation is certainly

possible, as discussed earlier, but it understands the second gate as lower in the

body than the first gate: heart > navel > head.

Deussen interprets his Calcutta edition (BI) differently than the Adyar

Library edition or other southern editions. All the other translations are based on

South Indian manuscripts. Deussen translates verse 26 in the following way:

Through heart-gate, wind-gate,

The gate, which leads upward,

And the opening of the gate of liberation,

Which they know as the open orb.60

Deussen’s reading is plausible. He reads mūrdha-dvāra as “the gate that leads

upward.” In his translation he reads bila as the “opening” of the gate of liberation.

He reads suṣira and as the adjective “open” modifying mandala, which he

translates as orb. His reading suggests a system of four gates. This is certainly

consistent with the earliest notions of centers in Buddhist literature. It also makes

easy use of the terms cave, hollow, and circle without the need of special

interpretive gymnastics. Deussen’s system would be heart-gate > wind-gate >

upward-gate > the gate of liberation. The opening of the gate of liberation is also

called the open orb. If one alters Deussen’s translation slightly to read, “They

241

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

know the opening of the gate of liberation as the pierced circle," then we see

remarkable parallels between this text's fourth gate and the brahmarandhra of the

Tantras.

Ayyangār renders verse 26 with a considerable number of interpolations

from the commentary.

The enlightened know the seven doors leading to attainment

of Ātman, viz., the door of the heart (leading to Virāj), the

door of Vāyu (leading to Sūtra), the door of the head

(leading to Bīja), and the door of Mokṣa (leading to the non-

differentiated Brahman), Bila (leading to Turya-virāj), Suṣira

(leading to Turya-sūtra), and Maṇḍala (leading to Turya-

bīja).61

Ayyangār's translation attempts to make sese of the text through interpolations

from Upaniṣad Brahmayogin's commentary. Whereas this interpretation

accurately communicates the 1751 interpretation of the text, it is not especially

helpful for understanding the older context, as none of the language—including

virāj, turya, and other interpolated terms—appears in the original text itself.

Varenne also follows Upaniṣad Brahmayogin's commentary. His

translation of verse 26 includes less interpolations but maintains the heading "the

seven doors." It reads:

Sept sont les Portes:

le Cœur, l'Air, la Tête;

au-delà : la Libération,

60 Deussen (1997), p. 696.

61 Ayyangar (1952), p. 14.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

le Trou, la Cavité;

enfin : le Cercle.62

Varenne's interpolation "seven are the gates," does not appear in any

published edition of the texts.63 This translation, even with the "seven gates" title,

is presented as a list, with no notes or other interpretation. However, when this

translation is combined with Varenne's other arguments,64 one finds that he

directly equates the gates of verse 26 of the Amṛtanāda Upaniṣad with the standard

seven (six plus one) cakras of the Tantras and haṭha yoga.65 The Amṛtanāda Upaniṣad

text itself employs none of the standard terms of the tantric six plus one cakra

system. Even if Varenne's reading is correct, he does not explain what logic this

ordering of gates would follow: heart > wind > head > liberation > hole > cavity >

62 Varenne (1971), p. 120.

63 Varenne interpolates the terms sapta-dvāra from Upaniṣad Brahmayogin's commentary.

64 See Varenne (1973), pp. 164ff. Another example in which Varenne bends the text to the

'standard' system of seven appears in his comment on ANU v. 19. He states: "The other interesting

point brought out by this quotation is the mention of the first center of the subtle body the

mulādhāra." This verse in the ALS does not mention the mulādhāra. The lowest cakra is only

referred to in the commentary of Upaniṣad Brahmayogin. It never appears in the text. The verse

actually refers to breath control: "Closing the nostril with one finger, [he] takes the air in with the

other, maintains agni [or blocks in himself agni] and meditates on the sound [oṃ]." The verse

indicates the maintenance of "fire" in the body. The most direct reading is that this verse simply

refers to the "heating" effects of holding the breath. It may additionally indicate a visualization of

fire, but the focus of meditation appears to be on the sound of the mantra. Considering all its

possible valences, it can only be equated with kuṇḍalinī meditation through a series of

interpolations and interpretations. Varenne's translations (1971, and 1973) rely heavily on

Upaniṣad Brahmayogin's commentary and perpetuate a general or standardized reading of the

texts that obscures the originality and ambiguity of the texts.

65 The six plus one (or seven) "centers" system is widely known in India and the West from

Arthur Avalon's (John Woodroffe) translation and interpretation of the Saṭcakranirūpaṇa (SCN) and

the Pādukāpañcaka (PP), and the through the popularization of the haṭha yoga of the Nāths, which

also employs the same system. See previous notes on the Kulaḷijñānanirnaya and the KMT, and

White, "Yoga," throughout.

243

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

circle. This ordering and its terminology cannot easily be correlated with to the

standard six plus one cakra system.

There are centers outside the body in Kashmir Śaivism, and there are secret

centers within the head in some Nāth traditions. While these traditions could

possibly be the sources of the Amṛtanada Upaniṣad list, this would still not explain

how a system borrowed from Kashmir or the Nāths would include no references

to the technical terms of these traditions. Alternatively, if this text does indeed

refer to seven gates, then this sevenfold system may constitute a divergent

tradition, unknown from any other source. If this is the case, then the seven gates

would not correspond to the six plus one system of cakras found in so many other

yoga texts that have been influenced by the six cakra system of the Yoginī Kaula

and the Western Transmission.

Like Aiyar, Feuerstein simply maintains the literal ambiguity of the text in

his translation. He translates verse 26 as follows:

[Yogins] know the gate of the heart, the gate of the wind, the

gate of the head, and likewise the gate of liberation, as well

as the "cavity" (bila), the "hole" (sushira), and the "circle"

(mandala).66

Elsewhere Feuerstein offers the explanation that this is a system of seven gates

that leads the yogin to liberation. He does not assert that this is the same system as

the six plus one cakra system. He states:

244

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

They [the seven gates] refer to diverse anatomical structures, though the author does not divulge anything about them. The last four are probably all esoteric loci in the head. These technical terms hint at the fact that the composer of this Upaniṣad was steeped in esoteric lore that was far more sophisticated than the text of his composition.⁶⁷

Feuerstein is probably correct concerning the complex systems that lie behind the text. However, he also admits that the meaning is unclear. Like Varenne, he accepts the reading of Upaniṣad Brahmayogin’s commentary in using the term sapta-dvāra to refer collectively to the seven terms.

This reading of verse 26 as designating seven gates suggests the possibility of a number of subtle centers in the head. Such a system would begin with the path of the breath in the heart going through the gate of wind (throat) into the head, where there are several centers in the head that concern the yogin: mūrdhan, mokṣa, bila, suṣira, and maṇḍala. Direct equations are not possible due to the originality of the terminology. However, there are a number of parallels in the Tantras that provide readers of the Yoga Upaniṣads with a fuller picture of the variant interpretations of esoteric locations in the head.

The Kaulajñānanirnaya recognizes the cooling knot of the uvula; the root (or tip) of the nose; the “end of twelve” (finger-spans) (dvādaśānta); the space between the eyebrows; the forehead (mūrdhan); and the cleft of brahman (brahmarandhra) at

⁶⁶ Feuerstein (1998), p. 419.

⁶⁷ Feuerstein (1998), p. 416.

245

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

the crown of the head. The location of the “end of twelve” center is not clear from

the text. Later traditions, such as the Trika Kaula, locate this center twelve finger-

spans above the fontanelle. However, the Kaulajñananirnaya does not provide any

explanation and the placement of this center in the otherwise linear list suggests a

location inside the head. Twelve finger-spans describes the measure of distance

that the vital breath reaches outside the body, from the tip of the nose.68 It is also

the approximate distance through the head from the root of the nose to the

occipital protuberance. This bump at the back of the skull is the object of a

meditation on an eight-leaf cakra in Kaulajñananirnaya 7.16.69 Mūrdhan probably

means forehead but could also mean the roof of the palate.

Abhinavagupta, the great exponent of the Trika Kaula, lists various esoteric

locations inside the head: bhrūmadhya, lalāṭa, tālu, triveni, brahmarandhra,

dvädasänta, and triśūla.70 The bhrūmadhya we have already seen as a literal

description of “between the eyebrows.” Lalāṭa is in the middle of the forehead.

Tālu is the uvula or the vault of the palate. This area is also called lambikā,

lampikästhāna, and catuspada. The triveni is the triangle that constitutes the

68 Silburn (1988), p. 30.

69 Kaulajñananirnaya 7.16: “Meditate on an eight leaf cakra, white and resembling the moon,

at the juncture of the spine and the skull at the base of the hair,” (keśaskandhakapālasyaṁ). There is no

prima facia reason to connect the “twelve finger-spans” to this verse. I offer the observation that

“meanings” do not always accompany these terms, and that the same term may be used in

different ways from text to text.

70 White, “Yoga,” p. 4. See also Silburn (1988), pp. 28–33 for the locations and workings of

these points or centers in the head.

246

confluence of fire, sun, and moon; the three breaths; or the three primary nāḍis.

Brahmaraṇdhra is the cranial suture, the fontanelle. Dvādaśānta, discussed earlier,

can mean the limit of the breath "where the ordinary breath dies away."71 In the

Trika system, dvādaśānta has at least two additional meanings. It is a synonym for

the brahmarandhra, following the skull twelve finger-spans in an arc from the

bhrūmadhya. It is also the supreme dvādaśānta that is twelve fingers' breadth above

the head. The triśūla, trident is located at the level of the fontanelle.72

The Nāth texts offer even more names and terms for esoteric loci inside the

head, including fire, sun, and moon maṇḍalas, liṅga, ghaṇṭikā (uvula), lampikā

(uvula), rājadanta (elephant's tooth, also uvula), and upper door of the śaṅkhini

(conch). Nāths also employ the more common subtle body terms, such as ājñā

cakra, which is the same as the bhrū.73 There are also numerous knots between the

palate and the top of the skull, some of which are reminiscent of tantric texts (tālu,

nāsāmūla, and so on).

Returning to the Amṛtanāda Upaniṣad, verse 26 provides a list of seven terms

whose meaning and esoteric location are in the final analysis uncertain with the

exception of the terms heart and head74 in two of the gate names. As we have

71 Silburn (1988), p. 30.

72 See also Flood (1993), pp. 256ff. and his notes on Brunner (1974).

73 Silburn (1988), pp. 124ff. See also White (1996), chapter eight.

74 As explained above, the term mūrdhan may mean the "top" or "uppermost" and thus

may refer to different parts of the head.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

seen, interpretation of Amṛtanāda Upaniṣad 26 demonstrates considerable

ambiguity and polyvalence of meaning. I have followed this single example

through its multiple possible interpretations to demonstrate several points. Most

important, it is clear that Amṛtanāda Upaniṣad 26 does not list the six plus one cakra

system that later came to be seen as the system of esoteric body centers. My

primary concern here has been with the northern Yoga Upaniṣads, yet the

corresponding verses are identical in the later southern Yoga Upaniṣads. We must

thus question the interpretations of those translators who have attempted to

equated the three, four, or seven gates of the Amṛtanāda Upaniṣad with the six plus

one cakra system.

It is also clear that, the Amṛtanāda Upaniṣad does describe a yogic body

whose navigation parallels the yoga of the earlier texts of the Yogini Kaula and

Western Transmission texts (and less so the Trika). Of the many possible

readings, two probable hypotheses emerge from this detailed examination. First,

the Amṛtanāda Upaniṣad describes a system with four esoteric gates. Attached to

this list are three or four additional ambiguous terms. The gates system starting at

the heart matches the oldest Upanisadic models of the yogic body. A four gate

system parallels the four center systems of the early Buddhist Tantras, although

the actual centers described in the two systems do not correspond: heart-throat-

248

History, Text, and Context of the Yoga Upanisads: Chapter 4

forehead-fontanelle as distinct from navel-heart-throat-head.75 Second, however

one interprets the cave, the hollow, and the circle, no valid reading suggests that

this is more than a four gate system. Whether these terms are understood as secret

names for the four gates, or they are additional esoteric loci in the head, the

fundamental conclusion is that it is still a four gate system. Just as the special

inner cranial points of the other yogic and tantric systems are not counted as

primary cakras, so the three additional terms in the Amrtanada Upanisad do not

themselves constitute distinct gates.

4.4. Conclusions About the Northern Yoga Upanisads

All the clues in the northern texts provide enough evidence for some

conclusions both affirmative and deconstructive. Affirmatively, the northern Yoga

Upanisads demonstrate developments in continuity with the classical Upanisads.

The importance of the heart, channels, and "breaking through the head" continues

from the older traditions, but these concepts are substantially transformed in

details and terminology. The concept of internal pathways (nadis) undergoes

considerable development from the hita-vein references in the classic Upanisads.

These developments may have occurred in the schools of yoga, but the evidence

75 The Samnyasa Upanisads, that share many characteristics with the Yoga Upanisads, include

four "sites." Brahma Upanisad, chapter two, names four sites: navel, heart, throat, and head. See

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

suggests that it was the medical traditions that first elaborated these concepts. It is

probable that the schools of yoga borrowed these ideas from medical theory. Body

centers are not systematically portrayed in the northern Yoga Upaniṣads, although

importance is attributed to the regions of the navel, heart, and head. The throat

and the area between the eyebrows are also at times emphasized, but in fewer

examples.

Examination of the northern Yoga Upaniṣads provides several important

examples of the need for ongoing deconstructive analysis of scholarship on yoga

traditions. It is clear from the evidence in translations and scholarly commentaries

on the Yoga Upaniṣads that the northern tradition has been ignored or

misinterpreted. One crucial example of this confusion is the scholarly attempts to

standardize of the metaphors and images across traditions. Commentaries on the

Yoga Upaniṣads have often equated the gates and centers of the Yoga Upaniṣads

with the six cakra system of the Yoginī Kaula, the Western Transmission, and later

traditions of haṭha yoga. In addition, many scholars of the Yoga Upaniṣads have

relied exclusively on Upaniṣad Brahmayogin’s 1751 commentary on the southern

canon. This has led to the obsfuscation of the many important differences that

distiguish the northern and southern traditions of the Yoga Upaniṣads.

Olivelle (1992), p. 149. See similar reference in Parabrahma Upaniṣad chapter two. in the same

source, p. 267.

250

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

Examination of the details of the yogic body in the mantra yoga of the

northern Yoga Upaniṣads leads to the conclusion that there is not a systematic yogic

body in the northern canon—even though it is likely that the tantric systems

predate and probably influenced these texts. Some of this ambiguity results

because the images of a yogic body found in the northern Yoga Upaniṣads do not

appear to be structures of an esoteric or astral body at all. It is much more likely

that the references to esoteric centers are textual shorthand for visualization

practices rather than subtle body taxonomies. The images of the flowers, vital

points, gates, pathways, and other locations or structures'presented in the

northern Yoga Upaniṣads are all consistent with meditation practices of

visualization, in which such homologues and images are consciously constructed

to accompany the recitation of the mantra om. It is certain that modern discourses

of the Hindu Renaissance and academia associated the subtle body with the

physical body. They also see concepts concerning the subtle body as a system.

Even if the subtle body is ethereal, it is a body homologous to the physical body.

Visualization techniques, in which images are imagined within the body as part of

mantra meditation, do not require a taxonomic presentation of a system of subtle

body structures. Visualization techniques form part of the practical technologies

of meditation, whereas systematic taxonomies of the subtle body form part of

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

theories of the body. Meditation technologies and theories of the body may occur

together but neither one requires the other.

In the northern Yoga Upaniṣads, it is possible that references to the heart,

cranial fissure, and nāḍis point to actual structures in the subtle body. However,

in the case of all the other body images, the texts appear to present not maps of the

subtle body, but rather guides to visualization meditation.76 As we move into an

examination of the southern Yoga Upaniṣads, we will see that such a distinction is

not inconsequential.

Although it is impossible to ascertain from internal textual evidence, it

seems most likely that the northern Yoga Upaniṣads are heirs to tantric systems of

visualization meditations, since these are attested earlier than the Yoga Upaniṣads

and in much greater detail, including some of the same terminology (nāḍi, padma,

and others). Since the texts are thoroughly Brahmanical in their metaphors

(names, mantras, gods, and so on), it is difficult to determine the nature of the

tantric influences on the northern Yoga Upaniṣads. The Hamsa Upaniṣad, discussed

76 This agrees with more detailed descriptions of meditation found in the Buddhist and

Hindu Tantras and the works of Kashmir Śaivism. (See Flood (1993), pp. 256-267, and his sources;

See also Snellgrove (1988), sections III.14-15, and throughout, for Buddhist examples.) This point

may not be conclusive, but evidence continues to accumulate that subtle body systems developed

from a combination of visualization meditations (including interiorization of external rituals,

models, machines, beings, and celestial bodies) with various medical conceptions. Esoteric

manuals repeatedly insist that meditation “maps and models” were just tools, not representations

of reality. However, this insistence never overcame the equally powerful social forces of reification

and concretization. These processes of making “lazy” or temporary methods into theories of

252

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

earlier, provides an example of a transition text between the northern and

southern texts. The Hamsa Upaniṣad describes the yogin moving the wind through

the six locations: ādhāra, svādhiṣṭhāna, manipūraka, anāhata, viśuddhi, and ājñā.77 No

characteristics are ascribed to these names, and they are not called cakra. The text

further describes the yogin "sending the wind from the ādhāra to the brahmarandhra

(cranial fissure) while meditating on that one [om] which comprises three phones

and the nāda." Otherwise, the hamsa is meditated on in the heart of eight petals.

Like earlier Yoga Upaniṣads, the hamsa-bird-body is described with homologies as

having Agni and Soma as his wings, and so on. The body characterizations in this

text are self-consciously described as meditative visualizations. The image of

ascent accompanies the list of centers: the yogin should send the wind upward,

around, and through the centers. The breath-ascent is not described as going

beyond the throat, but rather the yogin is instructed to "maintain the breath in the

viśuddhi." According to the text, the eyebrow circle and cranial fissure are the

objects of "thinking" and "meditating" respectively.

Continuing the process evident in the Hamsa Upaniṣad, the southern Yoga

Upaniṣads solidified the connections between their yoga and that of the Tantras and

Āgamas. Whereas the northern texts are terse and often unclear, the southern texts

facticity are social tropes, found throughout institutional and bureaucratic rituals and in the

taxonomies of scientists, intellectuals, and philosophers.

253

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

borrow detailed materials from the Tantras to explicate their texts. The expansion

of the southern canon explicitly relies on such tantric textual materials to expand

the mantra yoga meditations of the northern texts into the recognizable systems of

haṭha yoga.

4.5. The Southern Yoga Upaniṣads and the Science of Haṭha Yoga

The expanded texts of the southern tradition of Yoga Upaniṣads significantly

go beyond the northern tradition in their presentations of the structures of the

subtle body. In the previous section, I suggested that the body images of the old

northern tradition were unsystematic examples of visualization meditations.

These images are supplemented and adapted in the southern Yoga Upaniṣads to

accord with the medieval systems of eight- and six-limbed yogas, especially the

haṭha yoga of the Nāths.

By the time that the southern Yoga Upaniṣads were written, the widespread

adoption of certain tantric notions via the haṭha yoga texts had slowly standardized

several conceptions concerning the powers of the yogic body. Liberation, power,

bliss, and knowledge were seen by many as the transformations resulting from the

77 Paragraph 3 in Deussen's BI translation. Verses 6-7 (expanded with elaborations) in

ALS.

254

History, Text, and Context of the Yoga Upanisads: Chapter 4

activation of the serpentine force, the kundalini, in the late medieval period.78 In

the Tantras, various ritual, alchemical, sexual, and yogic practices were prescribed

for the awakening of the dormant kundalini-energy that resides at the root of the

body in the lowest cakra. This lowest center was seen to be located at the navel or

lower at the root of the trunk in the ādhāra or mūlādhāra cakra, variously depicted

as at the base of the spine, the perineum, or anus. Over time, haṭha yoga became

the dominant system of practices for raising this serpentine power.79

When awakened, this energy is said to surge upward like fire through the

central channel (suṣumnā nāḍi) of the body. When it reaches the cranial vault, or

escapes the body through the cranial fissure, it brings explosive realization, super-

human powers, and yogic bliss. By the time of the redaction of the southern Yoga

Upaniṣads, the kundalini energy is as often depicted and imagined as an actual

power inside the body more often than as an image in a guided visualization.

The Nāth treatises on haṭha yoga represent an explosion in the details and

elaborations of body practices and body theories. The Tantras, especially of the

78 The conceptual framework developed for describing, explaining, and interpreting this

transformation involved considerable historical development, drawing on the ancient shamanic

and ascetic traditions of south Asia. Although heir to long historical development, a refined

Kuṇḍalini theory only appears in medieval alchemical, tantric, and yoga traditions. There is

extensive technical vocabulary associated with this process of transformation. Kuṇḍalini is the

most widely used term in general, western literature: compared to the extensiveness of the

tradition, singular use of this term is somewhat misleading. An inexhaustive list would include:

Kuṭilangi, bhujangi, Śakti, caṇḍāli, dombī yoginī, nariyamani, śikhi, and others. See also White, "Yoga,"

pp. 7ff; Silburn (1988), throughout.

Yoginī Kaula and Western Transmission, often contain unsystematic forms of

haṭha yoga or similar techniques under different names. The development of haṭha

yoga parallels the historical developments of the earlier yoga formulations. It is

clear that there were breath control and meditation traditions before the common

era, yet it was the Yoga Sūtra of Patañjali that brought a systematic yoga into

writing, and therefore into history. In much the same way, Gorakhnāth and the

other Nāths took the yoga practices of the Kaula and Kubjikā Tantras and

systematically invented the formal system of haṭha yoga, as it was forever named

thereafter. The earlier northern Yoga Upaniṣads were mostly attested by the time of

Gorakhnāth and like the earlier and contemporary Tantras, contain a mixture of

ingredients that were brought together more systematically in the haṭha yoga texts.

The physical and gymnastic quality of haṭha yoga is evidenced in all of its

texts. Many visualization meditations were increasingly “physicalized” and

materialized by the body systems of haṭha yoga. This was the general model of

haṭha-formulation, as it also physicalized alchemical processes and apparatuses.80

Over time, these interiorized “physicalizations” became yogic-medical knowledge;

they were generally accepted as physical descriptions of the human body. This

physics thus did have dimensionality in that the system recognized gradations

29 This raising is sometimes called kuṇḍalinī yoga, or other names, but is essentially the yoga

of the Tantras elaborated by the Nāths, especially Gorakhnāth. In other words, after the time of

Gorakh, all body yoga was ever after influenced by the forceful yoga of the Nāths.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

from the material to the ethereal, or gross to subtle. There are many examples of

this shift from an emphasis on images visualized in meditation to concrete

physical structures, whether gross or subtle. One example is the nāḍis. From the

earliest mention of these channels in the medical texts, the term nāḍī is a

polyvalent term designating various types of channels and other vein-like

structures, ranging from blood-carrying veins to nerves. In haṭha yoga texts the

concept of nāḍī is even further physicalized through assoication with purification

practices. Clearing the nāḍis in the older yoga traditions required simple breath

control and visualization practices. In haṭha yoga traditions, in contrast, these

pathways of the body are literally cleared through the many cleansing or washing

(dhautī) practices that include the sinuses, bowels, ears, urethra, and any other

accessible body orifices or passages.81 The southern Yoga Upaniṣads, through

incorporating haṭha yoga terminology and practices, display a similar tendency

80 White, "Mountains"; White (1996).

81 We see a similar increasing physicalization with the haṭha yoga version of the khecarī

mudrā. This term has several meanings. As a body practice for haṭha yoga it requires the severing

of the fraenum and physical stretching of the tongue so that it can be reversed and inserted into the

sinus cavity (kapāla kuhara). Earlier alchemical and visualization practices had this same name, and

it seems probable me that the body altering involved is part of the physicalization of, and in this

case sexualization of, the Gorakh tradition of physical yoga. We find additionally the cakras in the

Haṭha yoga as being associated with the spine, e.g., ŞU 4.10 states: "The suṣumṇā is known as

viṣuddhārini and mokṣamārga. Attached to the vertebral column behind the anus up to the crest of

the head where it is called the brahmarandhra." The earlier visualizations of internal body centers

appear to literally run through the middle of the body: navel knot, heart, center of skull, cranial

suture are all adjacent to the spine but not inside it or astrally overlapping it. The relationship of

the spine to the central channels and the centers was probably solidified in the Nāth texts and even

further elaborated in the nineteenth and twentieth centuries. Avalon's Serpent Power introduced

257

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

toward concretization and systematization. What was vague and experimental in

the northern Yoga Upaniṣads becomes taxonomic dogma in the southern texts.

4.6. The Yogic Body of Haṭha Yoga and the Southern Yoga Upaniṣads

The science of haṭha yoga includes both theory and practice concerning the

human psychosomatic system. Important conceptual models for yoga practically

function as medical-anatomical theories. Anatomical assumptions and

terminology proliferate in the teaching, explanation, and practices of haṭha yoga.⁸²

A few of these yogic techniques and anatomical terms are of special importance in

the southern Yoga Upaniṣads. This includes the basic techniques of the haṭha yoga:

breath control (prānāyāma), posture (āsana), muscle-lock (bandha), hermetic seal

(mudrā), and body cleansing (dhauti). These practices manipulate anatomical

constructs and forces. Some important constructs are the different bodies: gross

body (sthūla śarīra), body of enjoyment (bhoga-śarīra), and subtle body (sūkṣma

śarīra). Next are the vital structures, junctures, and centers: vital point (marman),

knot (granthi), energy wheel/center/circle (cakra), lotus (padma), circle/matrix/orb

(mandala), and channel (nāḍī). The animating forces of the body include the prāṇas,

the West to the six cakra system. In it he associated subtle structures with muscle-skeletal features,

making the associations wide spread in the subsequent secondary literature.

⁸² See Feuerstein (1998), pp. 505ff. Feuerstein devotes chapter eighteen of his general

introduction to haṭha yoga. Although his presentation is often hagiographic instead of historical, it

provides a useful introduction and overview of haṭha yoga and the Nāth Siddhas. See White (1996),

258

History, Text, and Context of the Yoga Upanisads: Chapter 4

the elements and senses, the mental constituents of the brain (mind,

consciousness, ego, and so on), and the liberating energy of the breath power,

called variously the pointed flame, the goddess energy, the serpent power and

others (śikhā, prāṇa, śakti, agni, kundalī, kundalinī, and others). Haṭha yoga

developed these notions as structures and forces of the human body.

As part of its science, haṭha yoga also interiorized sexualized symbols and

practices.83 Commingled sexual fluids provide the literal and metaphorical grounds

for many conceptions about animating body energies in haṭha yoga. The

combination of male and female body forces are modeled on feminine discharge

(dravyam), lineage nectar (kulāmṛta), true being-“good stuff” (sadbāva), vulva

essence (yonitattva), menstrual blood (śoṇita, artava), and male semen or sperm

(śukra, retas). When the “good stuff” of the female was conjoined with the semen

of the male, a great fluid (maharasa) was produced.84 In the tantric Kaula

for a detailed investigation of the Siddha traditions, including many specialized details of haṭha

yoga.

83 The kundalinī practices include numerous sexualized symbols, e.g., the feminine heat

meeting with the cooling seminal nectar in the heart or head of the yogin. The meditation practice

of the khecarimudrā in its Nath version involves inserting the lengthened tongue into the sinus

passage where it is milked to release the nectar of immortality. The sexual valences here are

obvious. Gorakhn and other Nath authors write various negative assessments of physical human

women. Unlike the Kaula Tantras, the haṭha yoga replaced the human females of old kaula practice

with gendered forces contracted, pumped, and milked within the body of the single male yogin.

See White (1996) and White, “Yoga.”

84 White (1996), p. 200. These are the body substances of the sexual rites of the Kaulatantra.

They have alchemical and subtle yogic body homologues. Sexual facts such as genital fluids and

actions such as coitus provide the objective models for various metaphors and symbols used in all

of these systems. Cosmological processes, terrestrial landscape, metallurgy, and embryology

supply additional metaphor fields and symbols.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

traditions, the intermingling of these "power-substances" was practiced literally in

both yogic and ritual settings. The haṭha yoga traditions transformed the theories

and practices of intermingling through homologizing these various substances

with forces and structures within the single yogin’s body.

This terminology of sexual fluids refers to qualities or elements of both the

gross and subtle body landscapes. Except for the basic constituent parts of the

human physiology found in the earliest literature—in particular Vedic texts and

medical literature—most of these structures and energies were inspired by the

rituals, yoga, and sexualized symbolism of the Tantras. The elements derived from

the earlier texts include the breaths, senses, elements, and mental faculties. Nāḍi

terminology apparently evolved next. Virtually all other terms and accompanying

concepts come from the medieval Tantras.

Tantric and yogic anatomical theory understood the body to be made up of

both gross and subtle constituents. The gross physical body (sthūla śarīra) is

composed of the five gross elements: earth, water, air, fire, and ether.85 It is subject

to physical and magical forces, and is impermanent. Mediating between the gross

and the subtle is the body of enjoyment (bhoga-śarīra), which “clothes” the subtle

body on the path of transmigration.86 It is more lasting than the gross body, but is

not eternal. The yogin’s ultimate concern is with the subtle body (sūkṣma śarīra), an

85 White (1996), p. 18.

260

astral body manifest as an invisible yantra or maṇḍala formed by geometric figures,

forces, and principles. This subtle body is beyond the range of ordinary sense

perception. According to this perspective, the physical body is not the model for

the subtle, but rather the subtle body is the matrix or blueprint (maṇḍala, yantra)

that is the prototype for the gross body. The meditation practices of the yogin

focus on the subtle body described as a complex matrix of channels, winds, knots,

props, gates, and centers. The subtle body is the body of liberation, in that

mastery over it results in immortality, power, and liberation.

The southern Yoga Upaniṣads occasionally employ such adjectives as gross

and subtle but they do not provide a systematic presentation of the subtle

physiology. The tantric concepts of yantra and maṇḍala, which are important in

haṭha yoga theories of the subtle body, are not explored and explained in the

southern Yoga Upaniṣads. Much subtle body theory is present in the southern

canon, but it is embedded in the general worldview of the texts, and is not a topic

of special interest or investigation. However, some of the constituent parts of the

subtle body—in particular the—nāḍis and cakras—are of interest to the compilers

of the texts.

86 White (1996), p. 18.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

4.7. Nāḍis and Prāṇas in the Southern Yoga Upaniṣads

The nāḍis are channels or pathways of the yogic body.⁸⁷ They are analogous

to, and as numerous as the veins, nerves, and arteries: 101, 72,000, or even

hundreds of thousands.⁸⁸ About a dozen of these nāḍis are of primary concern the

yogin, and, as discussed earlier, three are of major importance. The vital forces of

the body, the prāṇas, travel through the nāḍis. Through exercises of breath control

(prāṇāyāma), the yogin clears the nāḍis. Through various breathing and muscular

exercises, postures, and visualization techniques, the yogin awakens to an

awareness of all the body's paths.

The number of major nāḍis of the yogic body differs from one text to the

next. Drawing on evidence of a number of importatn sources, around twenty

major nāḍis are named:⁸⁹ suṣumṇā, iḍā, piṅgalā,⁹⁰ gāndhārī, hastijihvā, puṣā, yaśasvinī,

⁸⁷ There are numerous overlapping conceptual systems of the nāḍis. Nāḍis are ambiguous

structures relative to the concepts of gross and subtle. Some texts clearly see the channels as the

physical pathways of the body. High Tantrism and modern scholarship usually describe the nāḍis

as subtle channels. Indian medicine and the old Upanisadic notions are more material in their

estimations.

⁸⁸ See chapter two. These first numbers appear in the classical Upaniṣads. Medical and

yogic literature and the Tantras repeat these numbers. 72,000 is the most common ideal number.

Other numbers are also written: hundreds of thousands or even "countless." Gorakhnāth says

200,000 in several Nāth sources. TŚBU gives the number 80,000. The late haṭha yoga manuals ŚS

and GS state 300,000.

⁸⁹ Yogayājñavalkyam, Gorakṣaśataka, Siddhasiddhāntapaddhati, Yogaśikha Upaniṣad (B), Śandilya

Upaniṣad 4.9, Jabaladarśana Upaniṣad, Yogacūḍāmaṇi Upaniṣad.

⁹⁰ The iḍā runs from the left nostril, the piṅgalā from the right. The suṣumṇā runs through

the center of the subtle body. Earliest characterizations have the left and right channels simply

running down each side of the body. Later characterizations tend to show the two crossing each

other at joint/knot/or center points. In these systems, the iḍā and piṅgalā run back and forth

traversing the subtle body and intersecting in the cakras. All three nāḍhs terminate at the base of the

lowest cakra either at the perineum or between the genitals and navel. The suṣumṇā stretches from

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

alambusā, Kuhū, śaṅkhinī, sarasvatī, vāruṇī, payasvinī, śūrā, viśvodarī, saumyā, vajra, citriṇī, jīhvā, and rakā. The southern Yoga Upaniṣads are heir to so many sources

that they contain longer lists than most other texts: the Yogaśikhā and Śāndilya Upaniṣads include fifteen of these channels, whereas the Gorakṣaśataka includes

only the first ten. The classical Upaniṣads contain non-specific91 references to hitas, or vein structures, which form in the heart. This idea was replaced by multiple

references to the navel as the region of a bulb-like structure from which the many nāḍīs spring. These other major nāḍīs have various locations in the southern Yoga

Upaniṣadsand are characterized differently from one text to another. Usually starting in the navel, they terminate in such locations as the big toe, ears, eyes,

penis, palate, tongue, and other parts of the body.92

Nāḍīs carry breath or vital energy. They have numerous homologues: genders of male, female, and neuter, and other allforms such as rivers, colors,

gods and goddesses, and heavenly bodies. The terminology for three most important nāḍīs, iḍā, piṅgalā, and suṣumṇā—may predate the names of the other

the lowest cakra through the fontanelle, to the outside of the body. It reaches to the highest seat or center. See Eliade (1969), p. 236; Varenne (1976), pp. 160ff.

91 The term hita is used to mean “vein” or channel. The suṣumṇā is named only in the MaitU.

92 See ŚU 4.9-11; TŚBU 70-75; DU 4.6-20; YCU 18ff. See chapter three for the sources of these Yoga Upaniṣads. Much of their characterizations come from Nāth texts, as the YCU list demonstrates.

263

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

nāḍis. The texts present these three as the primary focus of yogic exercises.93 We

have already encountered these, especially the central channel, in the classical

Upaniṣads, Tantras, and the northern Yoga Upaniṣads. In the southern Yoga

Upaniṣads the nāḍis are not described in the context of meditation practice but are

rather presented in descriptive lists as structures of subtle physiology.

The southern Yoga Upaniṣads that deal with the ten to fourteen principal

nāḍis generally follow the pattern of their sources, presenting the principle prāṇas

in the same context. The old characterization of prāṇas from the classical Upaniṣads

has also been expanded. The southern Yoga Upaniṣads enumerate ten prāṇas:

prāṇa, apāna, vyāna, samāna, udāna, nāga, kūrma, kṛkara, devadatta, and dhanañjaya.

Like the variations in treatment of the nāḍis, the descriptions of the breaths differ

in some details.94 The first five, following the traditional functions, are generally

respiration, excretion, decomposition, digestion, and the ascending wind. The five

additional prāṇas govern different functions, such as spitting, blinking the eyes,

sneezing, yawning, pervading the body, hiccuping, and other windy functions.

The southern Yoga Upaniṣads describe the nāḍis and prāṇas as yogic body

93 Masters of haṭha yoga pay greater attention to the larger group of channels. For example,

advanced yogic practices, such as the ear splitting of the Kanphaṭa order of Nāths, is meant to

stimulate or fully open the ṭuṣā and yaśavini nāḍis, or other secret ear terminating channels.

Mastery of these other channels is often associated with gaining super-human powers, whereas

piercing the central channel leads to liberation. Not all yogins pursue liberation: some seek

advanced levels of embodied power.

94 See DU, 4.23-34, similar lists also appear in the southern recensions of the NBU, DhBU,

TBU, TŚBU, YCU, YŚU.

264

channels and winds. Although the secondary literature generally explains these

as subtle, the southern Yoga Upaniṣads tend to by descriptive and literal in their

descriptions, and, like medical texts, it is ambiguous as to whether the nāḍīs and

prāṇas are considered subtle or gross. Haṭha yoga tends to treat channels and

winds rather concretely, as discussed earlier, suggesting that the visualization

techniques of the earlier northern texts have developed into biomedical

descriptions by the time of the compilation of the southern Yoga Upaniṣads.

4.9. Cakras and Other Body Centers in the Southern Yoga Upaniṣads

The subtle structures and centers of the yogic body in the southern Yoga

Upaniṣads include wheels (cakras), lotuses (padmas) props (ādhāras), knots

(granthis), and vital points (marmans). The yoga systems of India describe various

numbers of body centers: five, six, nine, twelve, or even twenty-seven.95 Some

examples of these centers were encountered in our discussion of the northern

texts. Because of the variety of their sources, the southern Yoga Upaniṣads mention

several of these different types of body centers. However, the standardized six

95 The six cakra system is actually a center system of six plus one, or seven. The seventh,

highest center is without the general characteristics of the cakras because it is beyond normal

reality. Correctly speaking, it is not a “cakra” at all. In the Śaktatantraṁ, it is a seat or “pīṭhā.” It is

usually described as a lotus with thousands of petals. Other systems of cakras (discussed above)

described different numbers of subtle “centers.” Centers in the navel, heart, throat, and head

anchor these different systems. Variations and expansions add to these four.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

cakra96 system, originally developed by the Yoginī Kaula, dominates all of the

materials of the southern texts. The southern Yoga Upaniṣads adopted this system

primarily from texts attributed to Gorakhnāth and from the haṭha yoga manual, the

Haṭhayogapradīpikā (ca. 1450).

Cakras are centers of vital energy, literally "wheels," or "circles." Each cakra

has its own corresponding deity, element, maṇḍala, mantra, sphere of control, color,

and number of lotus petals.97 Hindu lists locate the cakras proximate in the gross

body to the anus or perineum, genitals, navel, heart, throat, forehead, and the

seventh, meta-center above the head.98 The standard names that appear in the

southern Yoga Upaniṣads are (mūla-)ādhāra,99 with four petals; the svādhiṣṭhāna, with

six petals; manipūra(ka), a ten petalled lotus; anāhata, with twelve petals; viśuddhi,

with sixteen petals; ajñā, with two petals; and sahasrāra, the thousand petalled

lotus variously described as at the end of the path of the brahmarandhra, as the

dvādaśānta, or as the dvādaśāra.100 In this last case, the meaning of the term

brahmarandhra has shifted. In the earlier texts, brahmarandhra was a term for the

96 Descriptions of cakras often include the term padma intermixed within the same lists. The

other terms (knot, vital point, prop) are generally not used to refer to the cakras, but to other classes

of body loci.

97 See Varenne (1976), pp. 167-72, for diagrams and full descriptions of each cakra. See also

Motoyama (1981), Avalon (1974).

98 Varenne (1976), p. 172.

99 Some of the southern Yoga Upaniṣads use the term ādhāra, others use the term mūlādhara

for the center at the base of the body, the perineum between the anus and the genitals.

100 These spellings represent the six plus one centers as they are presented in the southern

Yoga Upaniṣads.

266

hole at the fontanelle. In later texts, the term can refer to the fontanelle hole, to the pathway leading outside of the body from the hole, to the suṣumṇa, or to the thousand petaled lotus flower. Some of the texts name four piṭhas located in the body. The southern Yoga Upaniṣads mention particularly kāma or kamariupa, a four-

petaled lotus between the two lowest cakras. In addition to being described as subtle body centers, the cakras are also understood as psycho-mystical levels of transcendence or trance. In other words, the cakras can be understood as either energy centers in the subtle body or as expereinces of sequential levels of psychophysical perfection.

The cakras and other subtle body structures (knots, vital points) are connected by nādis. The southern Yoga Upaniṣads do associate these centers with the spine (merudaṇḍa or brahmadanda). A characterization that appears to have been derived from the Nāth texts. The cakras are enlivened by the awakening of the vital breath or the kundalini power. When the kundalini fire is raised, it moves upward via the central channel through the cakras. As it pierces or enlivens each cakra, new levels of awareness, yogic super powers, and other paranormal experiences occur.

4.10. Kuṇḍalini in the Southern Yoga Upaniṣads

According to the Tantras, kuṇḍalini is the power or entity that usually lies

dormant in the base of the human trunk, below the navel. She is referred to as a

snake, as a goddess, as a śakti, as a fire, or as some other kind of power.101 It is the

yogin’s task to awaken her, and once awakened, she rears up like a king cobra

spreading its hood or rushes up the central channel of the body like a snake out of

its hole. Tantric texts also describe kundalini as heat burning its way upward

through the body.102

Kundalini is also the microcosmic homologue of the macrocosmic creative

force.103 The various tantric theologies differ as to which god or goddess identify

with ultimate reality. Generally, the creation of the cosmos is envisioned as a

creative force that bursts forth from original, unified, eternal reality, which is

variously designated as Brahman, Paramaśiva, Parā, Siddhilakṣmi, among other

names). This creative force is sometimes described as the cosmic kundalini, which

generates the manifest universe. Following an emanational metaphor, the

kundalini bursts forth from unity into multiplicity. Following the metaphor of a

river she runs from unity to evolved multiplicity like water from a melting

101 Descriptions vary, but the serpent force lies dormant low in the pelvis: at the base of the

spine, just below the navel, at the perineum, and so on. See Silburn for extensive examination of

this subject. See also White, "Yoga." For general discussion see White (1996), Padoux, Flood,

Varenne, Eliade, Feuerstein, Svoboda, Evola in this chapter’s bibliography.

102 The many characterizations of this force exceed the parameters of the present

discussion. For example, there are voluminous details concerning sexual imagery, geographic

imagery, and the goddess imagery associated with Kundalini. For more detailed information on

the variety of interpretations and their contexts, see White (1996), especially chapter 8. See also

Kinsley (1997) and Silburn (1988).

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

mountain glacier that forms a delta and runs into the sea. Both metaphors—along

with other metaphors of wells, mines, and caves—depict a cosmic force dispersed

and hidden on the macrocosmic level in the manifest world and on the

microcosmic level inside the human body.

Methods of raising the kundalini involve combination of yogic body

exercises, concentration, visualization techniques, and meditation practices. Some

varieties of practice include verbal repetition of mantras and visualization of

yantras, combined with haṭha yoga breathing exercises, postures, muscular

contractions, and locks. All of these activities are intended to open and purify the

various subtle channels and centers and push the vital breath down into the

lowest cakra104 where kundalini lies dormant. The adept awakens her through the

force of breath moving into the lowest center. The traditions of the Tantras also

maintain that the kundalini can rise spontaneously due to the grace of the deity.105

In yogic practice, it is through breath control that the lowest body center is

heated, like a furnace and bellows. This heat awakens the dormant kundalini, and

she moves up the central channel through each cakra106 as a fire or serpentine

power. This ascent can be gradual, penetrating one higher body center and then

103 For the cosmic role of the serpent force, see Padoux (1992) and Silburn (1988),

throughout.

104 Whether in the navel or perineum.

105 This characterization is especially found among the Kashmiri Śaiva tāntrikas. Some even

favor spontaneous arousal of kundalini as the greatest form of the practice. Cf. Padoux (1992),

Silburn (1988), Murphy (1999).

269

History, Text, and Context of the Yoga Upanisads: Chapter 4

returning to the base of the body. In such cases, the serpent force is raised

repeatedly to different body centers until it finally reaches the highest centers.

Other characterizations include a force burning up everything in its path and then

blasting through the head like a volcano. Alternate descriptions have her going

up to the head, then down to the lowest cakra. In another image, she is Śakti who

rises up and unites with her male consort, usually a form of Bhairava or Śiva. This

union occurs either in the cranial vault or above the head in the highest center

outside the body.

Many of the southern Yoga Upaniṣads refer to the kuṇḍalini directly as well

as by implication. These references are often drawn from known Tantras or Nāth

yoga texts. Advayatāraka Upaniṣad verse 5 offers the following description of the

serpentine force and some details concerning the subtle body, the exact sources of

which are not known. The technical language in the passage suggests that it was

probably inspired by Nāth Siddha yoga texts.

In the middle of the body there exists the suṣumnā, the nāḍi of the

Brahman, with the form of the sun and the luminosity of the full

moon. Rising from the mūlādhāra, the channel goes in the direction

of the brahmarandhra. In the center of that channel is the celebrated

kuṇḍalini, with a radiance equal to myriads of lightning flashes and

subtle-membered like the thread of a lotus fiber. Having beheld it

with the mind, the practitioner is liberated because of the destruction

of all evil (pāpa). If he incessantly beholds her radiance (tejas) by

virtue of the effulgence of tāraka yoga in the maṇḍala of the lalāṭa

106However many their number might be.

270

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

(forehead), then he is a siddha.107

The Advayatāraka Upaniṣad account gives this brief description but otherwise

does not provide much information on the serpentine force. It is most likely

that the Advayatāraka Upaniṣad represents an experimental yoga text. In

other words, the author of the text actually practiced yogic meditation. The

meditations described in the text are almost exclusively devoted to

visualizations and visual manifestations. However, kundalini appears in the

text as a formulaic description of the forces of the sublte body, with little

relationship to the meditations and visualization practices actually

described in the text.

Triśikhibrahmaṇa Upaniṣad verses 59-65 provides another description

of the kundalini. Although there are no direct quotations, the textual

description in the Triśikhibrahmaṇa is similar to those found in the

Amaraughaśāsana of Gorakhnātha. Triśikhibrahmaṇa Upaniṣad 59-65 states:

They know the middle of the kāṇḍa (bulb) as the navel. In the navel

is a twelve-spoked cakra. On those spokes are Viṣṇu and the others.

I take my stand there and cause the cakra to spin by my own creative

power (māyā). The jīva spins among the spokes, one after another.

Just like the spider remaining in the middle of its cage of webs, the

jīva moves perched on the prāṇa. The jīva does not exist without the

prāṇa. Above the bulb is the place of kundalini, in line with the navel.

She is composed of eight different constituents and is a spiral of eight

coils. Being coiled around and beside the bulb, she is always

regulating the prāṇa and apāna, the passage of water, food, and the

107 See also Feuerstein (1998), p. 427.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

like. She covers completely the hole leading to the brahmarandhra.

When roused by air (marut) and fire (agni) of yoga, in the form of a

serpent, intensely effulgent, she shines in the space of heart.108

In this passage from the Triśikhibrahmaṇa Upaniṣad, as in the passage cited

above from the Ādvayatāraka Upaniṣad, the description of kuṇḍalini appears as a

passing reference. In the southern Yoga Upaniṣads, the references to the serpent

force appear to be derivative from earlier traditions and simply reproduce the

received traditions concerning the kuṇḍalini as a component of the subtle

physiology. The kuṇḍalini appears to play little role in the meditation practices

that are of central importance in the southern Yoga Upaniṣads. Breathing exercises

and mantra recitation are extensively described and explained in the southern Yoga

Upaniṣads, while most of the references to the kuṇḍalini are brief and formulaic,

derived from earlier sources, such as Nāth Siddha texts.

4.11. Conclusions and Problems

Scholars of yogic traditions often adopt generalized designations of yogic

elements, structures, and systems when evaluating Indic models. One example

would be in the use of the term subtle for certain structures of the

psychophysiology. Another example would be the use of the term cakra for

esoteric power centers within the human body. Awakening these centers is

108 See also Ayyangar (1952), pp. 101-102.

272

History, Text, and Context of the Yoga Upanisads: Chapter 4

fundamental to the practices of tantric yoga and hatha yoga. However, scholars fall

prey to the possibility of over generalization when they impose specific

terminology before examining the actual systems as described in particular texts.

As we have seen, the images of the yogin's inner body in the northern Yoga

Upanisads differ from the standardized terminology used to describe the subtle

body in tantric and hatha yoga texts. The southern Yoga Upanisads, in contrast, tend

to replicate the standardized discourse of the kundalini and the six plus one cakra

system.

Whereas moving from certain Tantras to other Tantras or alchemical texts

often allows one to examine different appropriations of the same terminology, in

different textual environments, one cannot make this jump from the northern Yoga

Upanisads to the Tantras or other genres of yogic literature without reservations.

Explanation of the yogic body in the northern and southern Yoga Upanisads

requires examination of the yogic cosmos and the distinctive concepts of

interiorization that are unique to these texts.

The body imagery in the Yoga Upanisads, as discussed earlier, often replicates

the celestial imagery of the classical Upanisads. Liberation in the classical texts

involved the essential Self, as breath or mantra, escaping the vaulted cosmos

through the doorway of the sun. The Yoga Upanisads transpose this old cosmic

liberation path onto the interior of the body of the yogin. This mapping process is

273

History, Text, and Context of the Yoga Upanisads: Chapter 4

often called interiorization (or internalization), it reflects and understanding of the

interconnections between microcosm, mesocosm, and macrocosm.109

Interiorization involves several different methods of meditation and practice:

visionary experiences, visualization practices, and visionary poetry.

The first of these method is the experimental yogic experience of

visionionary experiences. Some yogic texts describe or give clues regarding the

physiological results of yogic practices, including sweating, shaking, floating,

flying, seeing lights, and hearing noises as well as other feats. Depending on the

lineage and its gurus, some of these results are judged measures of

accomplishment, as hindrances, or as powers in their own rights. Siddhis, the

yogic superpowers appear at the refined end of this continuum of results. These

bodily experiences are considered the byproducts of the yogin’s altered states of

consciousness.

Altered states of consciousness, as practiced by yogins, received substantial

negative and biased assessments by Western scholars in the late nineteenth and

early twentieth centuries. In 1931 Jean Filliozat summarized the general trends of

the then-current scholarship concerning yoga.

The union of the soul with the supreme Being is achieved by

relatively unsophisticated practices: one hypnotises oneself by

holding some posture, concentrating one’s gaze, holding one’s

breath, concentrating on some sacred syllable (OṂ); one uses

109 See also chapter 2.

274

prolonged fasts, solitary living, or suffering. It is from these that

the Hindu fakirs were born.110

Filliozat continues by quoting Barth concerning yogic practices: "When these are

rigorously practised, they can lead only to madness and idiocy."111 Filliozat

worked against these kinds of judgments in his scholarship, attempting to bring

clarity and legitimacy to several aspects of yogic practice and thought. However,

even with the work of such scholars as Filliozat, these unhelpful and biased

approaches to yogic states persisted for several decades, especially in studies of

mysticism and studies based on psychoanalytic theories. The old

characterizations of "self-hyponosis" and "madness" were not scholarly

explanations, but were rather pejorative labels that ended any possibility for

fruitful investigation or conversation.

Recent scholarship has largely ignored yogic states of consciousness, which

are held to be beyond the reach of the science of religion, and has favored instead

less subjective research into the contents of texts and the cultures that produced

them. Such studies have yielded rigorous investigations of the philology, history,

and philosophy of yogic traditions. Discussion of yogic practices have generally

remained on the level of phenenological descriptions, with little attention to the

110 Filliozat in "Sur la ‘concentration oculaire’ dans le Yoga" (1931), pp. 93ff, quoting Roger

Bastide, Les Problèmes de la vie mystique (1931), p. 32. Reprinted in Filliozat (1991), p. 269.

111 Barth. "Les religions de l’Inde," in Oeuvres t. I, p. 83. Filliozat (1931), p. 93. Reprinted in

Filliozat (1991), p. 270.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

states of consciousness that are purportedly produced by these practices.112

Another approach to yogic states has been to use terms such as "ecstasy" or

"enstasy" to designate such states, without attemping to explain in detail the

meaning of these designations.113

Extensive description and explanation are beyond the scope of the present

study, but a brief discussion of yogic states of consciousness will assist readers in

their interpretations of the contents of the Yoga Upaniṣads. Generally, research into

altered states of consciousness has been undertaken not by Indologists and

historians of religion, but rather by physical anthropologists and cognitive

neuroscientists who have accomplished significant enough research in the last half

century to provide a general overview of the effects of altered states on the bodies

112 There is nothing problematic or anti-intellectual in this "hands-off" approach to yogic

states. I have summarized these approaches here descriptively to demonstrate that most historians

of religion simply do not address the psychophysiological issues of yogic practice. As such,

textual, historical, or philosophical approaches are focused in such a way as not to have to evaluate

or explain altered states of consciousness or trance. These are generally the purview of physical

anthropologists and cognitive neuro-scientists, not that of Sanskritists or historians of religion.

This is not to say that Indologists have written nothing about yogic states, but rather that

Indologist do not appear to have benefited from the more sophisticated research done in

biomedical fields.

113 Eliade (1973) coined the term enstasy to differentiate it from ecstasy. Enstasy as used by

him designate the focused withdrawal from the world as in yoga or Zen meditation. Ecstasy is

employed to refer to the states produced by austerities or other more active trance practices such as

shamanic trance dancing. In practice, there is considerable permeability between these two

categories. Feuerstein (1997) rejects Eliade's usage, employing ecstasy for any altered state of

human consciousness. Considering the forceful nature of many haṭha yoga methods, Feuerstein

may be correct in his use of the term ecstasy for all such states. Trance studies suggest that Zen

style sitting meditation does activate different body chemistry than active trance practices.

Activities such as the bandha or mudrā practices of haṭha yoga or other active practices, such as

dancing, and the techniques of shamans and trance healers or mediums produce different

psychosomatic responses than Zen sitting.

276

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

and brains of human beings. In brief, altered states of consciousness in yogic

practice appear to result from changes in the nervous system and chemistry in the

yogin’s psychophysiology. These changes are brought about by repetitive training

and stressing of the respiratory system, muscles, and nervous system through

body postures, muscle contractions and locks, breathing exercises, and

meditations.114

A number of research studies indicate that yogic and meditative practices

cause many body systems, including the hemispheres of the brain, to function

differently than they do in ordinary body-brain functioning. These studies

indicate that people who practice meditation and trance technologies for long

periods of time can gain conscious control of otherwise unconscious bodily

processes (pulse, chemistry, and so on) and can also tolerate much higher levels of

physical stress (heat, cold, pain, and so on.). In addition, meditation practices can

improve mental clarity as well as visual and aural acuity. A number of research

studies have also shown that regular practice of certain meditation techniques can

result in visionary experiences and to generate, control, and alter these visions.115

114 Scholarly studies of altered consciousness and religious practice are included in the

bibliography. See especially William Wedenoja (1990).

115 Directed dreaming is one such meditation of control known to the West. “Day-

dreaming” is a simple form of directed, mild trance that is familiar to Westerners as well.

Westerners have generally not cultivated the more complex forms of consciousness training found

among esoteric societies, such as yoga, qigong, or other systems.

277

History, Text, and Context of the Yoga Upanisads: Chapter 4

In addition to visionary experience, a second method of interiorization

involves visualization, which, in contrast to spontaneously emerging visions,

entails the controlled generation of images. In visualization practices the yogin

meditates and projects an image (object, god, force, flower) in his mind's eye or a

specific part of his own body. Such visualization involve creative acts of

concentration and imagination in the context of meditation. In a technical sense,

such practices result in altered consciousness, but this consciousness altering is

largely under the control of the yogin. The yogin carefully imagines a detailed item

or image in his mind's eye through concentration and disciplined thinking.

Visualization also includes imagining sounds, smells, and textures, as well as

strictly visual images.

Visionary poetry is yet another tool of the yogins. It is a literary device that

may or may not have any relationship to visionary expereinces or visualization

practices. Visionary poetry involves authors of yogic texts developing symbols,

images, metaphors, similes, and other poetic devices to communicate yogic truths

and teachings. Visionary experiences and visualization techniques require yogic

practice. Visionary poetry requires skill at manipulating words and images in

literary works, whether oral or written. Historical figures such as Gorakhnāth and

Abinavagupta appear to have possessed developed skills in all of these areas.

History, Text, and Context of the Yoga Upaniṣads: Chapter 4

None of the three modes requires the others, although certain authors move

seamlessly from one to the other without differentiating among them.

The northern Yoga Upaniṣads include all of these categories: visionary

expereinces, visualization practices, and visionary poetry. Their primary form of

meditation involves mantra recitation and visualizations. They also include

occasional references to visions and sound experiences during the practice of

meditation. In addition, these body-laboratory texts interweave poetic and

symbolic images from the Tantras and other sources.

The southern Yoga Upaniṣads include taxonomic descriptions of visionary

experience, but, in contrast to the northern texts, in most cases they do not

represent active experimental or laboratory yoga. There are a few examples in

which the compilers altered the source materials in such a way as to suggest that

they did practice yoga.116 However, for the most part, the southern texts rely on

images and systems inherited from older sources. They have standardized what

were once visionary experiences and visualization practices into yogic taxonomies

and yoga theory. This yoga theory is combined with dogmatic presentations of

non-dual Vedānta. In mapping the history of representations of the yogic body

from the northern Yoga Upaniṣads to the southern Yoga Upaniṣads, we thus discern

116 Such as altering the description of a posture, lock, or seal, or adding the aside, "this

practice should be done with an empty stomach."

279

History, Text, and Context of the Yoga Upanisads: Chapter 4

a major shift in textual milieus: from experimental body-laboratory texts to

taxonomic yogic compendia.

280

History, Text, and Context of the Yoga Upaniṣads: Chapter 5

Chapter 5

Social World of the Yoga Upaniṣads:

History, Text, and Context

In chapter 2, we examined the historical roots of the Yoga Upaniṣads in

classical Vedānta, and the long development of the genre of minor Upaniṣads. It

has not yet been possible to write a detailed social history of these texts. Many

details and facts remain unknown, and some facts are most likely unknowable.

Nevertheless, in the concluding section of this chapter I will attempt to reconstruct

a hypothetical history of the Yoga Upaniṣads on the basis of the data that is

available.

In chapter 3, I provided a description and analysis of the texts of the Yoga

Upaniṣads. Various scholars of yoga have translated, cited, and commented on

these texts. These comments and analyses have been incomplete, misleading, and

at times mistaken. Many previous problems in analysis and explanation were due

to a lack of understanding concerning the distinctions between the two separate

canons of Yoga Upaniṣads, which have different dates, dispositions, and

geographic origins. Some scholars have relied on second and third order analysis

and never examined the texts themselves. Others have depended extensively on

eighteenth and nineteenth century commentaries, causing them to miss many

unique, unusual, or unsystematic features of the texts. Historical and cultural

analysis can only be revealing if the subject matter is first understood

281