Sunteți pe pagina 1din 13
Differentiating the Concepts of "yoga" and "tantra" in Sanskrit Literary History Samādhipāda: Das erste

Differentiating the Concepts of "yoga" and "tantra" in Sanskrit Literary History Samādhipāda: Das erste Kapitel des Pātañjalayogaśāstra zum ersten Mal kritisch ediert. The First Chapter of the Pātañjalayogaśāstra for the First Time Critically Edited. Texte und Studien, vol. 9 by Philipp André Maas; The Khecarīvidyā of Ādinātha: A Critical Edition and Annotated Translation of an Early Text of Hatḥayoga. Routledge Studies in Tantric Traditions by James Mallinson Review by: Gerald James Larson Journal of the American Oriental Society, Vol. 129, No. 3 (July-September 2009), pp. 487-498

Published by: American Oriental Society

Stable URL: .

Accessed: 10/02/2014 13:04

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .


JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact


information about JSTOR, please contact . American Oriental Society is collaborating with JSTOR to

American Oriental Society is collaborating with JSTOR to digitize, preserve and extend access to Journal of the American Oriental Society.

This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions

Differentiating the Concepts of "yoga" and "tantra" in Sanskrit Literary History




James Larson

of California,



of themost puzzling yet important terms in current research in South Asian


are the terms yoga and tantra, and the two books under review are important and welcome

contributions to gaining some clarity regarding the meaning and significance of both. Both volumes are revised versions of Ph.D. dissertations, the former for the University of Bonn

and the latter forOxford University. Philipp Andr? Maas

of thefirstP?da of the P?ta?jalayogas?stra,

P?da and the bh?sya attributed to Vy?sa, which he dates to the period of

provides the first critical edition

of the Sam?dhi

325 to 425 ce.

including both the s?trap?tha


provides the first critical edition of a well-known Hatha Yoga

text, the

Khecarlvidy?, many of the verses of which

the Yogakundall Upanisad, the fourteenth century ce.

text are also to be found in theGoraksasid

and the Matsyendrasamhit?,

or somewhat earlier. Both texts use

of the term tantra.What


he dates to perhaps yoga and both are

and a text that

the term

important for understanding the meaning


striking, however, is that the two terms yoga and tantrahave two distinctly differentmean

ings in the respective traditions towhich theybelong, and before discussing

these two new

critical editions it is essential tomake some historical and textual distinctions regarding the

relation (sambandha)

deal of popular as well as scholarly confusion regarding these terms thatneeds to be clari

fied and sorted out.

between yoga and tantra in these two environments. There is a great

P. V. Kane

in his massive History of Dharmas?stra


the following observation:


are really only

two main


of Yoga,


the one


in the Yogas?tra

and its Bh?sya by Vy?sa and the otherdealt with in suchworks as the Goraksasataka, the

Hathayogapradlpik? of Sv?tmar?mayogin with the commentary called Jyotsn?by Brahm?nanda.

Briefly, thedifferencebetween thetwo

is thatthe Yoga of Pata?jali concentratesall efforton the


and freedom


of the mind, while

from diseases.








of Dharmas?stra,

itself with

the body,

its health,

its purity







P?tanjala Yoga appears in the first centuries of the Common Era in a swira-compilation

known simply as the Yogas?tra. It is presented, at least in its principal bh?sya (attributed to a certain Vy?sa)?a bh?sya that in nearly all manuscripts, printed or handwritten, appears with

the s?trap?tha?as

a sam?na-tantra



(an 'explanation of S?mkhya'),


or, perhaps

in other words, a classical



a s?mkhya

system of Indian phi

losophy (darsana). Tradition links the compiler of the s?tra-p?tha, Pata?jali, with the famous


is a review article of: Sam?dhip?da:


ediert. The First Chapter

of the P?ta?jalayogas?stra

erste Kapitel

des P?ta?jalayogas?stra

zum ersten Mal


for theFirst Time Critically Edited. By Philipp Andr? Maas.


The Khecarlvidy? of ?din?tha: A Critical Edition and Annotated Translation of an Early Text ofHathayoga. By

James Mallinson.

Indiens. Texte und Studien, vol. 9. Aachen:





2006. Pp.


lxxxiii +






inTantric Traditions. London:


viii + 299.

Journal of theAmerican Oriental Society 129.3 (2009)


This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions

488 Journal of theAmerican Oriental Society 129.3 (2009)

grammarian, Pata?jali,

of the Mah?bh?sya.

Tradition likewise links the study of the self

(?tman) ormind (citta) in P?ta?jala Yoga with the two other principal "sciences" (tantras

or sastras) of the classical period (ca. third through thefifth centuryce.) in north

Indian in

tellectual history. The two other sciences (or tantras) are, of course, the science of medicine

(Ayurveda) and the science of grammar (Vy?karana), both of which are also associated


with thename Pata?jali,

and both of which were becoming mature sastras in the early

turies ce.

In addition to the association with thename Pata?jali, all three tantras or 'sciences' like

(2) sys

independence from religious authority.Regarding

wise share three important features, namely, (1) an empirical evidentiary database,

tematic pragmatic experimentation, and (3)

that the focus of

the latter feature, Imean

the subject matter in each instance is not depen

dent upon any particular sectarian orientation (Saiva, Vaisnava,

although, of course, the practitioners in each of the three tantras are often associated with

various sectarian orientations. Sectarian orientation, however, is neither a necessary nor suf

ficient condition for thework in these tantras.

My own view is that the traditional linkage of P?ta?jala Yoga, Ayurveda, and Vy?karana

is essentially correct both historically and intellectually, so long as one updates the his torical data in the light of recent research. In this regard, the major issue has to do with the

in the

second century b.c.e.; but thedate of the Yogas?trap?tha attributed to Pata?jali

a good deal later.The Yogas?trap?tha

name Pata?jali.

S?kta, Buddhist, or Jaina),

The grammarian Pata?jali,


tomost researchers, worked

is apparently

is probably tobe dated no earlier than the fourth cen

tury ce.

The principal reason for the later date of the s?trap?tha

is the extensive incor

poration of Buddhist notions and terms in all fourbooks of the Yogas?trap?tha, notions


terms that can be traced plausibly only to the first centuries of theCommon Era. In


regard, Louis de la Vall?e Poussin's "Le Bouddhisme et le Yoga de Pata?jali" (M?langes

chinois et bouddhiques

notions in all fourP?das of the Yogas?tra toVasubandhu's

5 [1936-37]:


inwhich he traces some fifty terms and


and Bh?sya,


to the point in attempting to date the Yogas?trap?tha. anachronism can be explained in either of two ways. Either therewere two

still very much The obvious


one the grammarian and the other the compiler of the Yogas?trap?tha. Most

scholars tend to accept this explanation, but not all. Or a portion or some of the s?tras, for

example, the Yog?nga "section" (YS II.28-III.5, or, according to J.W. Hauer, YS 11.28

section," can be traced to the grammarian Pata?jali.

III.55), the so-called "eight-limbed Yoga

Other s?tras were then collected by an unknown compiler (for example, someone such as


sake of legitimating thenew learned Yogas?stra and for the sake of highlighting theobvious

intellectual affinity of the three tantras. The latter explanation, I suspect, will eventually be

the S?mkhya teacher) with thewhole being attributed to Pata?jali, both for the

shown to be correct when sufficientevidence emerges, but currently either explanation is

plausible. For a full discussion see Gerald James Larson and Ram Shankar Bhattacharya,

eds., Yoga: India's Philosophy of Meditation,


of Indian Philosophies, vol. 12

(Delhi: Motilal


2008), 54-70.

In any case, as mentioned

earlier, there is a natural affinityamong the three tantras in

termsof an empirical evidentiary base, systematic pragmatic experimentation, and indepen

dence from religious authority. In the case of Yoga, thedatabase includes the study of bodily postures, breathing mechanisms, sensing andmotor functioning, the analysis ofmental states,

ego awareness, and general cognitive performance. The experimental component includes careful daily practice of physical exercises and cognitive meditations under the guidance of a

recognized expert or experts. Yoga in this sense of P?ta?jala Yoga lends itself to any number

This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions


"yoga" and "tantra" inSanskrit Literary History


of sectarian orientations, butmost often its primary affiliation iswith S?mkhya philosophy,

as clearly indicated in the colophons of all four sections of the Yogas?tra. In this regard, as

will be discussed below, the evidence is overwhelming in all printed texts and handwritten

manuscripts, where the colophons read, "iti pata?jale

yogas?stre s?mkhyapravacane




In the case of Ayurveda, the database

includes detailed classifications of symptoms, de

tailed categorization ofmateria medica (herbal medications or "pharmaceuticals" of all sorts),

and theuse of yukti or pragmatic reasoning in the identificationand treatmentof disease. The

experimental component includes the extensive and ongoing seminars or

symposia having

in a trial-and-error or

pragmatic manner, as evidenced in theCarakasamhit? and Susrutasamhit?. Here again, as

to do with the application of themateria medica

to various diseases

in Pata?jala Yoga, Ayurveda

independently of

is employed widely in sectarian and "secular" contexts quite

religious or sectarian authority. I have discussed this in detail in my

"Ayurveda and theHindu Philosophical


Systems," Philosophy East and West 37 (1987):

In the case of Vy?karana,

the database and systematic experimental component


the systematic description of the phonetic system of Sanskrit, the elaborate analyses of word formation, detailed citations of standard usage, the extensive meta-rules devised to describe

all aspects of Sanskrit in a comprehensive manner, and thedetailed and painstaking lists of

verbal roots and derivatives, all of which were derived from empirical observation and listen

ing, together with the construction of a

ture of the language. Again, as with Pata?jala Yoga and Ayurveda, sectarian affiliations,

of words as


theoretical framework thatwould exhibit the struc

the formation and meaning

while certainly pertinent in terms of analyzing

these may appear in sectarian contexts, in no way shape or determine either themethod

substance of what is being studied.

The terms yoga and tantra in these environments are clearly products of an elite intellec

tual milieu, made up of literate pandita communities, most likely innorth and northwestern

South Asia, that is, the Gangetic plain region (in and around present-day Varanasi) and the

Gandh?ra, Kashmir, and Punjab regions, in the early centuries of theCommon Era. Learned

traditionswere already taking shape, of course, a good deal earlier, in the time of theVedas

and Upanisads, and early Buddhist and Jaina traditions, and in the early epic period up

through the Mauryan period and the reign of Asoka. The Moksadharma portion of theMa


centuries before the beginning of theCommon Era, as is the grammatical theorizing

is symptomatic of the levels of intellectual sophistication achieved in these last


in such works as Pata?jali's Mah?bh?sya.


the consolidation established in the northwestern region of the subcontinent



Ganges River basin and the Gangetic plain (ca. 320-550),

(ca. 100 ce.),

together with the imperial Gupta


fashioned on the

an even more prolific cultural and

intellectual creativity emerged. The

various technical traditionsof Indian philosophy (Hindu,

Buddhist, and Jaina)

all begin to take shape in this period, each one centering around a

"founding" figure and a collection of utterances (s?tra) or verses (karika). Vaisesika, Ny?ya,


Ved?nta, Jaina, and Yoga

traditions all develop swira-collections associated with


the respective lineage figures of Kanada,


Jaimini, B?dar?yana, Um?sv?ti,


S?mkhya, M?dhyamika,

and Ved?nta have primarily faznfc?-collections with the

lineage figures of Isvarakrsna, N?g?rjuna, and Gaudap?da. Systematic and comprehensive

treatments in a variety of other areas of intellectual endeavor are developing, including grammar, medicine, poetry, astronomy, and law. It is precisely in this creative and systematic era that the terms yoga and tantra (and the

term s?mkhya as well) begin to be widely used. The termswere used earlier, of course, but

This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions

490 Journal of theAmerican Oriental Society 129.3 (2009)

for themost part they are late in appearing in classical Sanskrit literature.The term yoga

first appears only

course, it appears widely in the epic

and pur?nic literature.The term tantra appears only once in the Rg Veda in the sense of a


in the Taittirlya Upanisad

(II.4.1) and then in theKatha

and Svet?svatara

(II.3.11 and II. 11 respectively). Thereafter, of

'loom' and the fabric on a loom (cf. Grassman's W?rterbuch zum Rig Veda) and nowhere, so

far as I can find, in the early or "principal" Upanisadic

literature (cf. Jacob's Concordance).

Panini notes two roots yuj in his Dh?tup?tha,

the first in IV.68 as yuj sam?dhau


yuj in regard to concentration) and the second inVII.7 as yujir yoge (root yuj in regard to

yoking or harnessing or uniting). Vacaspatimisra

Yogas?tra. In

other words, in P?tafijala Yoga voga does notmean 'union'; it means, rather, sam?dhi. Panini

likewise notes the term tantra in two places, firstunder Astadhy?yi VII.2.9 in regard to the

suffix tra, meaning a 'vehicle for something', and then also inV.2.70 in regard to tantra as

points out thatboth he and Vy?sa under

stand theword yoga in the former sense

and not the latter sense in Pata?jali's

a cloth just removed from a loom (which presumably derives from theold Vedic reference).

The root tan originally meant 'spread' or 'stretch', in part in the sense of spreading or stretch

ing cloth on a loom.

In terms of Sanskrit intellectual history, itwould appear that the term tantra is first a


for 'extending' or 'stretching'. In its usage in learned, scientific environments, it is

a learned system or s?stra inwhich all of the component parts are placed

systematic place. M. Monier-Williams this sense:

in their proper

offers the following entries for the term tantra in



[the original


part, main

ory, scientific work,






feature, model,

such a work

the warp

; the leading

or principal

or essential

type, system,




the 1st section

a treatise


on astron




The old S?mkhya philosophy is, in this sense, called a tantra, namely, Sastitantra (that is,

a learned system having sixty components, possibly referring to a text by that titleor simply

a name for the old S?mkhya). It is not only philosophical

tantras in this sense. There are also tantras or learned traditions for many other intellectual

inquiries. There are, for example, the learned traditionsof Ayurveda, Dhanurveda, Darsana,

systems such as S?mkhya thatare

Vy?karana, Jyotisa, Ganita, Dharma, R?janiti, Nrtya, N?tya, K?vya, and a host of others. Generally speaking, in Indian intellectual history, tantra is a category or classification notion,

and as has been mentioned above, three learned traditions are characteristically referred to as

salient examples of tantra in this sense, namely, P?ninian

grammar or Vy?karana,

the classical S?mkhya (the Sasti


Indian Ayurveda of Caraka and Susruta, and, of course,

tantra) and P?tafijala Yoga, described as s?mkhya-pravacana. Furthermore, a tantra has certain essential components called tantra-yuktis, involving

thevarious methodological


traditionor science. Such yuktis are found already in theP?ninian system of grammar

and are discussed in detail in Kautilya's Arthas?stra, themedical

and the intro


and Susrutasamhit?,


four, or

devices thatare tobe used in composing or devising a tantra or

treatises, Carakasamhit?

and V?gbhata's

and Ast?ngahrdaya,

to the Yuktidipik? commentary on the S?mkhyak?rik?. Altogether thirty-two,thirty

thirty-six such yuktis are cited as "devices" tobe used in constructing a



of analogy"

include, for example, "reference to past authority" (atikr?ntaveksana),

(atidesa), "clarifying the topic of discussion" (adhikarana), "citing exceptions to general rules" (apavarga), "mention inbrief" (uddesa), "mention in detail" (nirdesa), "completing

an expression or ellipse" (v?kyasesa), and perhaps most notably "coherence and consistency

in presenting a subject matter" (yoga).

A useful survey account of thevarious lists of tantra

This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions


"yoga" and "tantra" inSanskrit Literary History


yuktis in this sense, together with detailed references from the classical Sanskrit texts, may be

foundW. K. Lele's

The Doctrine

of the Tantrayuktis:Methodology

of Theoretico-Scientific

Treatises inSanskrit (Varanasi: Chaukhamba

Surabharati Prakashan,

1981), 19-32.

Thus far, it is quite clear what yoga and tantramean in classical Sanskrit literature up

through the fourth and fifthcenturies of theCommon Era, and the social reality inwhich the terms are used is reasonably clear, namely, learned pandita communities in north and

northwestern South Asia in the time of theKus?na and the Gupta imperia. Then, however,

what appears to be a sectarian turn occurs, especially in certain Buddhist and Saiva (and to

a lesser degree, Vaisnava) environments inwhich the terms yoga and tantra come to have

dramatically different meanings. Quite possibly, of course, these sectarian traditionswere

prevalent in various incipient forms prior to thefifthor sixth century, but there is notmuch

textual evidence inSanskrit prior to themiddle of thefirstmillennium ce. Again, M. Monier


entry under tantra is interesting. Immediately following the first citation of the

meaning of

tantra as mentioned

above, Monier-Williams

continues as follows:

a class

of works


teaching magical


and mystical



1. the creation,


and Durg?

said to treat of five subjects,

in the form of dialogues

2. the destruction,



worship of the gods, 4. theattainmentof all objects, esp. of 6 superhuman faculties, 5. thefour



of union with

the supreme

spirit by meditation



in the sixth century of theCommon Era and thereafter, there is an explosion of

literatureknown as tantra in thisnew sense of "magical and mystical formularies." As Mark


has commented,

Although itisnot possible to sayexactly when thefirst ?gamas were written, thereisno concrete


to suggest

that any existed much


the sixth century

If our dates

are correct,

it seems thatthe Saiv?gama

that by thetimewe reach Abhinavagupta [975-1025] andhis immediate predecessors who lived

proliferated to an astonishingdegree at an extremelyrapid rate so

in ninth-century




in their works








Saiv?gamic literature. (Mark S. G. Dyczkowski, The Canon of the Saiv?gama and the Kubjik? Tantras of theWesternKaula Tradition [Albany: StateUniv. ofNew York Press, 1988], 5.)


comment applies primarily to the Saiv?gama


Tradition) branch, but the texts of other

(or theWestern Kaula

its so-called "Kul?mn?ya"

traditions of tantra in this

new sense?for example, theP?ficar?traVaisnava materials, the southern Siddh?nt?gama and

Sri Vidy?

also evidently

notmuch earlier than the sixth century, and in many instances are considerably later.Much

traditions, theBuddhist tantra traditions and the Jaina tantra?are

work is now becoming available

regarding the history and development of these various

traditions of tantra in north and south India, for example, in thework of David G. White

(Kiss of the Yoginl [Univ. of Chicago

("Saivism: Krama

Press, 2003]), Alexis


Saivism," "Saivism:

ofReligion, ed.M. Eliade


in Kashmir," and "Saivism: Trika Saivism,"

in Encyclopedia

[Univ. of Chicago Press, 1986], 13: 14-17), and Geoffrey Samuel

(The Origins of Yoga and Tantra

[Cambridge Univ. Press, 2008]).

"T?ntrika" in thisnew

sense is usually contrasted with "Vaidika"

(which latter termhere

is not somuch "Vedic" as something like "conventional" or "traditional"). Just as tantra in

the older sense, as we have seen, has certain characteristics, known as tantra-yukth, so it


appears that tantra in this new sense has some distinctive features, some of which are

following: (a) bubhuksu (desire for worldly experience) rather thanmumuksu (desire for

release); (b) focus on themodalities of "desire" (k?ma)\ (c) liberation while living (jlvan

mukti)', (d) quest

for extraordinary "powers" (siddhis); (e) transgressive or antin?mi?n ritual

practices (either in fantasy or literally) involving onanism, coitus, cunnilingus, and fellatio

This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions

492 Journal of theAmerican Oriental Society 129.3 (2009)

together with the exchange of sexual fluids; (f ) what Andr? Padoux has called a "swarm

ing pantheon with its fearsome deities" (both fearsome and benevolent, and both male and

female); (g) meditation practices that link the body of the practitioner (microcosm) with the body of the cosmos (macrocosm); and (h) theuse of sacred sounds, that is to say,phonemes,

syllables, mantras, and other ritual utterances in the context of prayer and/or magic. Per

haps thebest summary characterization of the "T?ntrika" Weltanschauung

comment by Andr? Padoux:

is the following




an attempt

to place


not to sacrifice

this world


in every



the word,

for liberation's

sake, but to reinstate

in the service


it, in varying ways,

within the perspective of salvation.This use of k?ma and of all aspects of thisword to gain both

worldly and supernaturalenjoyments(bhukti) and capacities, and toobtain liberationin thislife (jlvan-mukti), implies a particular attitudeon the part of theTantric adept towardthe cosmos,


he feels

integrated within

an all-embracing

system of micro-macrocosmic


(A. Padoux, "Tantrism," in Encyclopedia ofReligions, ed.M. Eliade, 14: 273; cited inD. G.

White, Kiss of the Yoginl,p. 15. See also Padoux's excellent discussion of "Tantrism" inhis

V?c: The Concept of theWord inSelectedHindu Tantras [Albany: StateUniv. of New York

Press, 1990], esp. 30-85.)



related to these new Tantra traditions, at least in theirSaiva and S?kti-Saiva for

is a new kind of Yoga, namely, Hatha Yoga

(literally 'Exertion-Yoga'), attributed

by and large to thework of two mah?siddhas,

and twelfth centuries either in the northwestern or northeastern margins of South Asia:

said to be the founder of theN?tha

thatboth Matsyendran?tha

legendary or mythical figures, but there seems tobe a growing

torical figures even though, of course, much hyperbole has come to surround their exploits.

is spelled out in summary form in the Yogatattva


who worked somewhere between the ninth

and Goraksan?tha

are only

consensus that they were his

that is



and his near-disciple Goraksan?tha,

order. Some have maintained

The nature of this new Hatha Yoga

oddly enough a Vaisnava

text but containing an account of Hatha Yoga

equally relevant for other sectarian groups. Four types of Yoga are listed in the Yogatattva;

is briefly characterized

as having twentycomponents, eight of which derive from the "Yog?nga" portion of theYS and an additional twelve:

R?ja Yoga, Hatha Yoga, Laya Yoga, andMantra Yoga. Hatha Yoga





the eyebrows, and Sam?dhi_(Thus)




of Hari

in the middle


is Yoga said tobe of eight stages. [These, of course, are

the well-known








of Yogas?tra



posture], Mah?bandha







the tongue


into the cranial

the central

area having


and Khecari


cut the frenum);


in the


restriction'or lock],Uddiy?na ['upward stomachrestriction'or lock], and similarly M?labandha




or controlling

the breath];




tionof the sacred syllable'], also Siddh?nta-sravana ['listening to the doctrines']; Vajroli [re

absorption of semen

one's urine and using theurine as a nasal douche], and Sahajoli [collecting urinebut not drink

after ejaculation, mixed with the female discharge], Amarol? [drinking

ing itor using itas a douche], considered as three aspects; theseconstitutethetwelvedivisions

of Hatha-yoga.





in The






Murti, tr.T. R. S. Ayyangar [Adyar:Adyar Library, 1938], 306.)

Hatha Yoga,

thus, has twentycomponents, the "eight-limbed" practices from the Yogas?tra,

and the twelve additional practices, plus, of course, the physiology of the cakras and/or


the theory of n?d?s, and the notion of the kundala (or kundalin?).

This content downloaded from on Mon, 10 Feb 2014 13:04:33 PM All use subject to JSTOR Terms and Conditions


"yoga" and "tantra" inSanskrit Literary History


Even though Pata?jala Yoga differs dramatically fromHatha Yoga,

it is possible

the latter from the former, at least to some degree. The locus appears to be inBook

to trace

III (the

Vibh?ti P?da) of the Yogas?trap?tha,


and specifically III.29, "When the circle of the nave