2

A History of the sad-

ahga-yoga of th e
Kdlacakratantra an d It s
Relation t o Other
Religious Traditions
of Indi a

A close look at the Kalacakratantra's six-phased yoga reveals its correlation and his-
torical connectio n t o earlie r forms o f the six-phase d 3>oga , foun d i n bot h Hin -
duism and Buddhism. Moreover, it also reveals the uniqu e character o f the practica l
applications an d implication s of the Kalacakratantra' s six-phased yoga. To the bes t of
my knowledge, the earliest reference to a six-phased yoga is found in the MaitrayarSya,
or Maitn Upanisad, whic h belongs to the branch of the blac k Yajur Veda and i s con-
sidered to be the last of the classical Upanisads. The sad-anga-;yoga of the Maitrdyaruya
Upanisad, Ch . 6 , v, 18, contains the following six phases: breath-control (f>raria;yarna) ,
retraction (prat^ahara) , meditative stabilizatio n (dhyana), concentratio n (dharana) ,
contemplative inquir y (tarka), an d samadhi.1 It is taught in this Upanisad as a method
for achievin g unio n wit h th e suprem e Sel f (paramdtman). I f we accep t tha t th e
MaitrdyanTya Upanisad predate s Patanjali , w e ca n assum e that thi s six-phase d yoga
also predates the eight-phase d ;yog a (astanga-^oga) o f the classica l Yoga system. The
fact tha t Patanjali never make s any reference to a six-phased yoga and tha t his Yo -
gasutra never mentions contemplativ e inquir y (tarka) is not sufficien t evidenc e to re-
gard the six-phase d ;yoga as a later revision of the eight-phase d ;yoga, as Giinter Gron-
bold suggests. 2 Eve n i f th e sixt h chapte r o f th e Maitrdyaruya Upanisad, whic h
incorporates a six-phased ;yoga, is a later interpolation, a s Mircea Eliade speculates,3
the antecedenc e o f the sixth-phase d yoga to th e yoga of Patanjali is still quite plau-
sible. The phras e "for it is said elsewhere," which often occurs at the beginning of the
verses of the sixt h chapter, indicate s that the Maitrayaruya Upanisad draw s its yogic
elements from the earlier yogic sources. Even though we are unable to determine th e
exact source s of the yogi c elements i n th e Maitrdyanya Upanisad, i t i s obvious that
different form s of its six-phased ;yoga have very early origins in India. The six-phase d
yoga was later modifie d int o diverse forms o f yoga with varying numbers of phases.

25
26 The Inner Kalacakratantra

For example, in one of the earlies t Puranas, the Vdyu Purdna, Ch. 10, v. 76,4 one en -
counters a five-phased yoga, whose fifth phase i s recollection (smarana), correspond -
ing in name to the fifth phase of the Kalacakratantra's six-phased yoga. In this Purana
as i n th e Kalacakratantra, contemplative inquir y (tarka) i s replaced b y recollection .
Considering tha t the Purana s underwent man y revisions afte r th e majorit y of their
material was composed during the Gupta reign (c. 32o-c. 500 CE), it is extremely dif-
ficult to establish whethe r th e recollectio n phas e of yoga was established first in th e
Puranic tradition or in the Buddhis t tradition, specifically , i n the Guhyasamdjatantra,
which som e scholars dat e a s early as the fourt h centur y C E an d som e as late a s th e
eighth centur y CE.
Within later Hindu sources , a six-phased ;yog a is also mentioned in a number of
texts belongin g to the Upanisad s of the Yog a class—specifically, i n the Amrtabindu
Upanisad—and i n the Saiva Agamas, Saiva tantras, and some Dharma Sutras , wher e
there is a slightly different order of phases than that found in the six-phase d yoga in
the Maitrdyaniya Upanisad. Fo r example, i n th e Amrtabindu Upanisad, v . 6, th e si x
phases of yoga are retraction (pratydharo), meditativ e stabilization (dhydna), breath -
control (prdnaydma), concentratio n (dhdrana), contemplativ e inquir y (tarka), an d
samadhi. This particular sequenc e o f the phase s o f yoga is almost identical t o that of
the Kalacakratantra. The differenc e between th e tw o lies in the designatio n of the
fifth phas e of yoga as contemplative inquir y (tarka) instea d o f recollection (anusmrti).5
Even though contemplativ e inquir y is not explicitly mentioned amon g the six phases
of the Kalacakratantra's six-phase d yoga as a separate member , i t i s not absen t fro m
there. Rather, i t i s included within the phas e of meditative stabilizatio n (dhydna),
along with wisdom (prajnd), analysi s (vicdra), jo y (rati) , and immutabl e bliss (acala-
sukha).6 Contemplativ e inquir y as a constituent of the phas e of meditative stabiliza-
tion i s explained i n th e Vimalaprabhd a s the apprehensio n o f the phenomeno n o f
empty form that is being observed or meditated upon during this phase.7 As such, i t
is an indispensabl e elemen t i n the practic e o f the Kalacakratantra's six-phase d [yoga .
Nevertheless, i t is not give n superiority over all other phases of the six-phase d yoga
and their elements as it is in KasmTr Saivism—specifically, i n the Saivagama s and i n
the work s o f Abhinavagupt a an d Jayaratha . Abhinavagupt a (975-1025 ) i n hi s
Paratrisikavivarana asserts that "among al l the light s of the componen t parts of [yoga,"
contemplative inquiry (tarka) ha s already been determined in the earlier Mdlimvijaya
"to be the brilliant sun by which one gets liberated and liberates others."8 When com-
menting o n Abhinavagupta' s Tantroloka , Jayarath a (thirteent h century ) i n hi s
Tantralokaviveka mention s th e six-phase d yoga that ha s breath contro l (prdnaydma)
as its first member and contemplative inquir y (tarka) a s its fifth member and exalts it
as the highes t (uttama) phase. 9 Moreover, just as contemplative inquir y is included
in the six-phase d yoga of the Kalacakratantra, even though i t is not regarde d as a sep-
arate phase, so too ar e meditative posture (asana ) an d restrain t (niyama) implicitl y
included in this yoga. The vo/'ra-postur e (vajrdsana) i s often referred to as the postur e
in which an adept of the Kalacakratantra does his meditative practice , whereas niyama
is included in the observanc e of the Kalacakratantra's ethica l discipline, in the form
of restraint fro m indulgin g in the five objects of desire and keepin g the twenty-fiv e
tan trie precepts (vrata), whic h ar e deemed prerequisites for the successfu l outcome
of the practic e of the six-phase d ;yoga.10 The Vimalaprabhd define s niyama as a Bud-
 A History of the sad-ahga-yoga of the Kalacakratantra 27

dha's command (buddhanujna) wit h regard to the twenty-five precepts.11 Since these
two prerequisites to th e Kalacakratantra's six-phase d yoga are present i n each phase
of the yoga as qualifying conditions , the y ar e not considere d t o b e separate phases.
Within later Hindu source s there are also those who speak of a six-phased yoga
that doe s not includ e th e phas e o f contemplative inquir y but include s meditativ e
posture (asana ) a s the first phase. Fo r example, some Yoga Upanisads—specifically ,
the Dhyanabindu {Jpanisad, v. 41 and th e Yogacuddmaru Upanisad, v . 2—several texts
of the GoraJcs a corpus (c. twelfth century), an d the Netratantra, cited in Ksemaraja's
Vimarsini (elevent h century ) commentary on the Siv a Sutra 6, contain the following
list of the six phases: posture (asana), breath-control (pramyama), retractio n (pratyahara),
meditative stabilizatio n (dhydna), concentratio n (dhdrana) , and samadhi. This for m
of th e six-phase d yoga seems to b e later tha n that found in th e Guhyasamajatantra
and later incorporated int o the Kalacakratantra. Thus, i t is most likely that the Bud-
dhist six-phased yoga chronologically succeed s the six-phase d yogas containing con -
templative inquir y (tarka) a s the fift h phase , whic h continue d t o be i n practic e i n
later times as well. However, it is more difficult t o determine with certainty whether
the Buddhis t six-phase d yoga precedes th e six-phase d yoga o f Kasmir Saivis m that
contains meditative posture (asana) as its first phase o r whether i t was contempora-
neous with it. If one were to rely only on the extant Saiva texts that refer to the sixth-
phased yoga having meditative posture as its first member, it would seem that the Bud-
dhist sixth-phased yoga preceded that particular yoga of KasmTr Saivism. Considering
the incompletenes s o f textual and historica l information , it i s impossible to recon -
struct an accurate an d precise history of the six-phase d yoga in India. Therefore, I of-
fer her e onl y a limited comparative tabl e o f the differen t type s of six-phased yogas
that wer e cited i n specifi c Saiva , Vaisnava, an d Buddhis t texts. A s table 2.1 . indi -
cates, not onl y teacher s o f different religiou s traditions but als o various teachers of
different school s withi n th e sam e tradition taugh t divers e forms o f the six-phase d
yoga, according to their intended goals. Even though these diverse types of the six-
phased yoga were couched withi n the differen t theoretica l an d practical framework s
of disparate traditions, they all share some commonalities. The mos t salient point of
commonality is that each form of the six-phase d yoga is viewed within its own tradi-
tion as inducive t o the accomplishmen t o f both limited, or mundane, an d supreme
siddhis. There are also certain commonalitie s in th e mor e general interpretations of
some phases o f the divers e types of six-phased yoga, despite the clea r divergence i n
the manne r i n which particular phases are structured and practiced withi n th e dif-
ferent traditions . Fo r example, in both KasmTr Saivism and Buddhism, the phas e of
breath-control (pranayama) involve s bringing the prdnas int o th e centra l nodi; th e
phase of retraction (pratyahara) involve s the withdrawa l of the sense s from externa l
objects; and meditativ e stabilizatio n (dhydna) implie s meditation o n a divine form ,
and so on. Their interpretations also coincide to a certain degree with Patanjali's def-
initions i n the Yog a Sutras. For the variant listing s of the si x members of the saA-ahga-
yoga within th e differen t school s o f the Hind u an d Buddhist traditions see table 2.1.
Within th e India n Buddhis t tradition , teaching s o n th e six-phase d yoga ar e
found withi n tw o Buddhis t tantric systems—th e Guhyasamajatantr a an d th e Kala -
cakratantra. Th e Hevajratantra (Ch . 8 , vs. 21-22) als o mentions a six-phased yoga,
but it does not list its members nor does it elaborate on it. Even though the six-phased
28 The Inner Kalacakratantra

TABLE 2.1 Th e Variant s of the Sadanga-yog a

The Tark a class of the Sadanga-yog a

Maitrdyamya Yog a Upanisads Visrtu Samhitd, Tantralokaviveka, 3
Upanisad, 16:1 8 Amrtandda, 6, etc. 30:57-58
prdndydma pratydhdra prdndydma prdndydma
pratydhdra dhyana pratydhdra dhyana
dhyana. prdndydma dhdrand pratydhdra
dhdrand dhdrand tarka dhdrand
tarka tarka samddhi tarka
samddhi samddhi dhyana samddhi

The Anusmrt i class of the Sadanga-y o

Guhyasamdjatantra, 18:14 0 Kalo£akratantra, 4:116
pratydhdra pratydhdra
dhyana dhyana
prdndydma prdnaydma
dhdrand dhdrand
anusmrti anusmrti
samddhi samddhi
The Asan a class of the Sadanga-yog a
Netratantra (Mrtyujit) Goraksa Texts, Goraksasataka, 7, etc.
dsana dsana
prdndydma prdnaydma
pratydhdra pratydhdra
dhyana dhdrand
dhdrand dhydna
samddhi samddhi

yogas of the Kalacakr a and Guhyasamaja systems accord in the name s and in the se-
quences o f thei r phases , the y diffe r i n thei r conten t an d practica l implications .
Among the Indian sources of these two traditions, the majorit y of treatises and com'
mentaries on the six-phased yoga belong to the Kalacakra corpus. According t o th e
Blue Annals, the six-phase d yoga of the Kalacakratantra wa s initially taught b y Vaj-
radhara i n th e for m o f Avadhutip a t o Anupamaraksit a (c . eleventh-twelft h cen -
turies), who passed it on t o his friend Sridhara. 12 Two works on the six-phase d yoga
are traditionally attributed to Anupamaraksita: the Sadangayoga an d the Sadahgayo-
gandma. The later Indian author Ravisrijnana (eleventh-twelft h centuries)—i n the
introductions t o hi s Gunabhararn, a commentar y o n th e Sadangayoga an d t o hi s
Sadahgayogatika, a commentary o n th e Sadahgayoganama—gives a brief account o f
Anupamaraksita's revelatory experience.13 According t o th e account s recorde d i n
the Gunabharara an d th e Sadangayogatikd, Anupamaraksit a studie d Buddhism and
other Indian systems of thought. Under th e guidance of Srlkhasarpana, he practice d
for twelv e years a meditation o n reality without an object and free o f conceptualiza-
tions, but was unable to gain a special insight. Depressed, he fell asleep, during which
VajrayoginT appeared to him, instructing him to go to Vikramapura, where he would
attain tha t specia l insight . Afte r arrivin g at midnigh t i n Vikramapura—accompa -
nied by his disciple, the grea t pandita Sridhara—Anupamaraksita received instruc -
tion o n th e six-phase d yoga directly from the Buddh a i n the for m o f Avadhuta. By
 A Histor y of the sad-angO'yoga of the Kalacakratantra 2 9

merely receiving the instruction tha t confirmed, "This is reality," he entered samadhi;
and upo n emergin g from hi s samadhi i n the earl y morning, Anupamaraksit a taugh t
this knowledge t o Sridhara.
With some variations, this story is repeated severa l times in later Tibetan chron -
icles o f Buddhism an d th e lineag e o f the Kalacakratantra's six-phase d ^oga. 14 Apart
from Padm a dkar po, who mistook Vikramapura for Vikramaslla monastery in Bihar,
none of the source s specif y th e locatio n o f Vikramapura no r th e plac e fro m whic h
Anupamaraksita wen t t o Vikramapura . I t i s likely that th e Vikramapur a to whic h
RavisrTjnana refers is VajrayoginT village in contemporary Dacca, located in east-cen-
tral Bengal, which is also thought t o be the birthplace o f AtTsa.15 This is perhaps th e
same Vikramapur a mentioned i n th e inscription s foun d i n nort h India . In th e in -
scriptions relate d t o the ruler s of the Varma n an d Vikramaditya dynasties of north-
ern India, Vikramapura is mentioned as their capital during the eleventh and twelfth
centuries. Th e Varma n dynast y ruled easter n Benga l i n th e secon d quarte r o f th e
eleventh century , an d thei r Vikramapur a was eventually overtake n b y Vijayasena ,
the greates t kin g o f the Sen a dynasty , in th e middl e o f the twelft h century . Thus ,
Ravisnjnana, who , accordin g t o Taranatha' s History of Buddhism in India, lived dur-
ing the reign of the Sena dynasty, coul d have been referring to that Vikramapura.16
Some inscriptions mention Vikramapura as a capital founded by Vikramaditya VI (c.
1076-1126). His father, Somesvara I, reigned i n Magadha and easter n Bengal , an d
he himsel f conquere d centra l Benga l shortly befor e 106 8 CE, afte r defeatin g Vigra-
hapala III . According t o Taranatha' s History of Buddhism in India, Anupamaraksit a
lived during the perio d o f the Bhayapal a and Nayapala kings of the Pal a dynasty. 17
Nayapala, th e fathe r o f the mentione d Vigrahapala III , ascended th e thron e in th e
early eleventh centur y and ruled the kingdom that extended o n the west up to Bihar
and to the east to central Bengal. Taranatha's informatio n coincide s wit h 'Gos l o tsa
ba gzhon nu dpal's assertion in the Blue Annals that Anupamaraksita could not hav e
been later thanNaro (956-1040 CE), since Naro cites Anupamaraksita's teachin g i n
his Sekkodesatika.16 Thus , whether Ravisnjnan a wa s referring to the Vikramapura of
the Palas , Varmans , o r Vikaramadityas , accordin g t o Buddhis t tradition th e Kola-
cakratantra's six-phased yoga was first disseminated in Bengal.
Anupamaraksita's name coul d have been easily related to the well-known Anu -
pama monaster y (vihara) i n KasmTr , whic h produce d BuddhasrTjnana , Sarvajna -
sriraksita, an d Sakyasrfbhadra , th e grea t early eleventh-century KasmT r scholars of
the Kalacakratantra. His name also could have been related to Anupamapura, the seat
of the tw o greatest Buddhis t center s o f learning i n KasmT r during the elevent h an d
twelfth centuries—th e monasterie s of Ratnagupta an d Ratnarasmi . I n eithe r case ,
Anupamaraksita coul d have come t o Bengal from KasmTr .
It is clear from the extant Indian an d Tibetan sources that there were several lin-
eages of the Kalacakratantra's six-phase d yoga in India. As these source s indicate, th e
most importan t amon g thos e lineage s was that of Anupamaraksita. I n the Gunabha-
ran, RavisrTjnan a give s th e followin g lineage : Anupamaraksita 19—Srfdhara—
Bhaskara—RavisrTjnana. Th e sam e lineage, bu t i n an extende d form , i s also give n
in th e Blue Annals, th e fifteenth-centur y Tibetan chronicl e o f Buddhism in Tibet,
which als o mentions th e famou s lineage s of Indian Buddhis t masters. According t o
the Blue Annals, the mos t famous lineag e of the Kalacakratantra's six-phase d ;yoga in
30 Th e Inner Kalacakratantra

India begins with Anupamaraksita and ends with the Bengali mahd-pandita, Vanaratna
(1384-1468). Vanaratna received the transmissio n of the six-phase d yoga from th e
mahd-siddha Savaripa , one o f the eighty-fou r legendary mahd-siddhas o f India, and h e
taught i t extensively i n Tibet during the first half of the fifteent h century. 20 The ex-
tended lineag e i s given a s follows: Anupamaraksita—Srldharanandan a (Sadhupu -
tra)21—Bhaskaradeva—Ravisrljnana (Suryasn)—Dharmakarasanti—Ratnaraksita —
Narendrabodhi—Muktipaksa—Sakyaraksita—Sujata—Buddhaghosa—Va-
naratna. The exac t same lineage of Indian masters i s also mentioned in Padma dkar
po's (sixteent h century ) Dpe med 'tsho'i lugs kyi rnal 'byor yan lag drug pa'i khrid rdo
rje'i tshig 'byed.22
Earlier Tibeta n historian s o f Buddhism in Indi a an d Tibe t recorde d a shorte r
branch of Anupamaraksita's lineage i n India . In hi s Dpe med 'tsho'i sbyor drug gi br
gyudpa, include d i n the Gsang sngags rgyud sde bzhi'i gangs'bum,23 Bu ston offers th e
following list for the India n masters following the lineag e of Anupamaraksita: Anu -
pamaraksita—Sndhara—Bhaskaradeva—Dharmakarasanti—Ravisnjnana—
Ratnaraksita—Vibhuticandra. Thi s lin e o f India n Buddhis t master s end s wit h
Vibhuticandra (twelfth-thirteent h centuries) . Accordin g t o Padm a ga r dbang, 24
Vibhuticandra received hi s Kalacakratantra initiatio n an d teaching s fro m thre e In -
dian scholars : Sakyasrlbhadra, th e maha-pandita o f Kasmir, 25 wh o was his principa l
spiritual mentor, Vikhyatadeva , and Dharmadasa . In Nepal, he mastere d th e Kdla-
cakratantra under the guidance of Ratnaraksita, the Newari mahd-pandita, from whom
he receive d the teaching s of the six-phase d yoga of the Kalacakratantra i n the tradi-
tion o f Anupamaraksita. During his stay in Nepal, Vibhuticandra becam e an exper t
in the Kalacakratantra and in the practice of the six-phased yoga. According t o Padma
gar dbang, 26 h e wrot e annotation s t o th e Kalacakratantra an d th e Vimalaprabhd,
which influenced later Tibetan translators and commentators on the Kalacakratantra.
As one of the India n mahd-panditas, Vibhuticandra visite d Tibet thre e times and be-
came fluen t i n th e Tibeta n language . He himsel f translate d hi s Sadangayogandma
(Rnal 'byor yan lagdrugpa)27 int o Tibetan. According to the Tibetan six-phased yoga
tradition, the Sadangayogandma i s the direc t transmission of the six-phase d yoga prac-
tice tha t Vibhuticandr a receive d fro m Savarip a durin g hi s sta y a t Stha m Biha r
monastery in Kathmandu, upon which he attained dhdrond, th e fourt h phase of this
yoga. In subsequent centuries, this text became one o f the mos t importan t an d au-
thoritative text s fo r the direc t transmission of the Kdlacakratantra's six-phase d yoga
in Tibet, especially in the Jonangpa tradition. According t o Taranatha,28 the teach -
ings on the six-phase d yoga that Savaripa revealed to Vibhuticandra wer e based on
the dohas o f Saraha, an d Saraha' s yogi c practice itself wa s based on th e six-phase d
yoga.
In th e Sbyor ba yan lag drug gi rdzogs rim gyi gnad bsdus pa, Tshong kha pa 29 (four -
teenth-fifteenth centuries) , followin g his teacher B u ston, cite s the India n lineag e
of Anupamaraksit a i n thi s way : Anupamaraksita—Sndhara—Bhaskaradeva —
Dharmakarasanti—RavisrTjnana—Ratnaraksita—Vibhuticandra.