Lakulas

PART - II
Blank Page
8 8
8 88
THE LKULAS :
NEW EVIDENCE OF A SYSTEM
INTERMEDIATE BETWEEN
PCRTHIKA PSUPATISM AND
GAMIC SAIVISM*
ALEXIS SANDERSON
The spark that ignited my first interest in Saivism was a
lecture on the Kashmirian poetician Mahimabhatta given by
your fellow-countrywoman Dr. S.S. Janaki in 1970 when I was
an undergraduate at Oxford preparing for a B.A. in Sanskrit
and she was completing her doctoral thesis. His objections
to nandavardhanas theory that the meanings that distinguish
the best poetry are conveyed to our minds by a power of suggestion
distinct from the semantic functions then admitted struck me as
prosaic and I was inspired to read the Dhvanyloka of his
opponent to see for myself what seemed to me much the more
compelling view. To my acquaintance with this work I owe my first
sense of getting to grips with Sanskrit as a medium of
refined thought and elegant expression. But it also introduced me
to the mind of Abhinavagupta as I looked to his commentary on
the Dhvanyloka for clarification of its meaning. Sensing the
greatness of his vision I went on to study his commentary on the
first six Adhyayas of the Bharatantyasstra. The difficulties I
144
INDIAN PHILOSOPHICAL ANNUAL
encountered in trying to read this text were great and I comfort
myself with hindsight that they were in some measure the
consequence of the very poor state in which the text has come
down to us. But what I could understand from this work was
enough to alert me to the idea that Abhinavaguptas view of
aesthetic experience was part of a doctrine of consciousness
itself for which I should look to his works in the tradition of
nondualistic Saiva soteriology. So in January 1972 I travelled to
Kashmir in the hope of being accepted as a student by the Saiva
Guru Swami Lakshman Raina. Under his guidance I read the
published Saiva literature of Kashmir, living near his Ashram for the
best part of six years. I think I did my best to understand
the system expounded in the Tantrloka but I cannot say that I
succeeded adequately even in the elementary sense. The problem,
I began to realize, was that Abhinavaguptas system could not be
understood on its own but only in relation to an independent
understanding of various traditions which he was drawing together
into his work. For it was clearly an exegesis that sought to harmonise
not just the statements within the Mlinvijayottaratantra, the
text on which the Tantrloka is a Slokavrtika, but the whole
body of Saiva scripture and practice then current in relation to a
view of that text determined by doctrines extraneous to it but
read into it in pursuance of a certain model of hierarchical
semantic dependence. I saw that it would be impossible to see this
hierarchy let alone perceive the purposes and contexts of its
elaboration if I remained within the corpus of texts preserved in
Kashmir, since the Kashmirians saw only the construction and
had no access to the realities on which it has been imposed.
And this was not only because so many texts important to
Abhinavagupta had become obsolete in the community of his
followers but also because the later development of Saivism in
Kashmir had been away from a system embracing the practice
of rites and ascetic observances towards one that saw in the
texts only what was of interest to the Janin. So thereafter I
concentrated my efforts on searching for manuscripts of the
works known to Abhinavagupta and, as I found them, seeking to
broaden my understanding from the purely doctrinal and
mystical into the more concrete realms of lived religion that
were surely its main substance in Kashmir as elsewhere. As I
did so I began to see the limitations inherent not merely in an
145
THE LKULAS : NEW EVIDENCE ...
approach to the study of the religion that centred on doctrine
but also in one centred on the relatively esoteric systems at the
centre of Abhinavaguptas vision: the Trika and the Krama.
These, I saw, drew their life and meaning from their relation to
the more broadly based and explicit religious systems among
which they flourished. In the first instance this led me to the study
of the Daksinasaiva tradition based on the Svacchandatantra, a
text which I then knew only as the beneficiary of a nondualistic
commentary by Abhinavaguptas successor Ksemaraja but
which I now know to be the basis of a large literature of ritual manuals
guiding a practice of Saiva initiation and worship that
had died out among the Kashmirians only shortly before my
time. But from this I went on at last to the study of the Kashmirian
Saiva Siddhanta expounded in the works of Bhatta Ramakantha and
others of his lineage, and have come in more recent years to
go beneath their constructions to the individualities of the
scriptural texts whose teachings they, like Abhinavagupta in the case
of the Trika, have tended to homogenize. My drive in all
this has been to appreciate the vision of the learned but by trying
to recover the context in which that vision developed and had
meaning. Their message is insufficient in itself when removed as
it must be with the passage of time from the context which gave it
meaning. For that context is barely expressed in the texts, their authors
assuming that future readers, like those of their own time, would be
thoroughly familiar with it. Reading outwards in this way I have
found that what at first I thought were localized Saiva traditions
were virtually pan-Indian, such early dichotomies as
that between Kashmirian Tantric Saivism and South Indian
Saiddhantika Saivism being found to be no more than consequences
of the places in which these traditions survive in recent times. The
Saiva Siddhanta is by no means a South Indian phenomenon,
though South Indians have massively expanded its literature and
institutions and added the splendid edifice of the Tamil Siddhanta,
at which, as yet, I can only gaze from afar, since I am ignorant of its
language.
The Saivism I have mentioned so far in this brief
journey through the curriculum of my studies covers virtually
the whole of what I shall call gamic Saivism. This falls in the
first place into two great divisions. On the one hand is the
146
Siddhanta centred on if not exclusively concerned with the
worship of Siva practised privately by all initiates for the benefit
of themselves alone and by those who inherit this right in Siva
temples for the public good. On the other hand are the non-
Saiddhantika systems concerned entirely with private worship
and observance that at least in the form in which they have
come down to us are post-Saiddhantika developments which
never question the validity of the Siddhanta as the means to
liberation and Siddhi but tend to see themselves as more esoteric,
more powerful methods of achieving the same ends. While Siva
is the deity of regular worship in the Siddhanta, most commonly
in the Lingam, in these non-Saiddhantika systems it is Bhairava or
various configurations of Goddesses culminating in cults of
solitary Kals, worshipped on Sthandilas or in small images. And
while the Siddhanta in almost all its variants through almost all
its history has been dualistic in its practice, that is to say, has
preserved in its observances and choice of offerings the distinc-
tions between the permitted and the forbidden propagated
by brahmanical Smrti, the non-Saiddhantika systems have
tended to advocate a nondualistic practice (advaitcrah) in
which these distinctions are held to be transcended. Within
these non-Saiddhantika systems we observe a further dichotomy
between the Saivism of the Mantraptha texts and that of the
Vidyaptha texts. The first is the Saivism of Bhairava worship
represented by the cult of Svacchandabhairava and his consort
Aghoresvar taught in the Svacchanda and the second is a more
diverse area in which the worship of goddesses predominates,
reducible in analysis to three main divisions based on texts that
were central in Abhinavaguptas time and for the most part still survive
in Nepalese manuscripts: the Yamala cult of Canda
Kapalin, consort of Kapalesvarabhairava, taught in the Picumata
alias Brahmaymala, the Trika, that is to say the cult of the three
goddesses Para, Parapara and Apara based on such texts as
the Siddhayogesvarmata, the Mlinvijayottara and the
Tantrasadbhva, and the Krama with an associated cluster of
Kal cults taught in the Jayadrathaymala also known as
the Tantrarjabhattraka.
But this is not the whole of the Sivasasana broadly
defined. For that includes the Pacarthika Pasupata system
147
which we know primarily through the Psupatastra and its
commentary by Kaundinya (Pcrthabhsya) and through the
Ganakrik and its commentary the Ratnatk. These two divisions
of the Sivasasana, the Pacarthika and the gamic differ in
fundamental respects. These are the principal according to the
authorities of each:
1. All initiates in the Pacarthika system are ascetics.
Initiates in the gamic Saiva system may be ascetics
or married.
2. All Pacarthika practice is for the sake of liberation
alone. gamic Saivas are either seekers of liberation or seekers
of Siddhis in this life and enjoyment of a paradise in the world
to come before they attain liberation.
3. Pacarthika practice has ritual elements but is predominantly
the cultivation of an enlightenment through Yoga and asceticism.
gamic practice after initiation may be of this kind but more
commonly it is principally the regular performance of obligatory
and other rites of worship.
4. Pacarthika practice is fireless. Homa, sacrifice into consecrated
fire (Sivagni, Bhairavagni or Devyagni), is central to gamic
practice.
5. All initiates in the Pacarthika system are actively engaged
after their initiation in Pasupata practice, either teaching and
initiating in the case of caryas, or practising the Pasupata
observance (vratam) in the case of Sadhakas. In gamic Saivism
early scriptural texts allow the benefits of initiation to be given
to deserving persons such as women, the elderly, young
children and rulers, who cannot be expected to undertake
postinitiatiory disciplines. There are therefore active and
inactive initiates, though the Kashmirian Saiddhantikas
were keen to limit this class of initiates, who were required
after initiation to do no more than express their devotion to
the religion by the means adopted by uninitiated Sivabhaktas.
6. Pacarthika initiation is a rite of passage into the practice
of the religion. That of the gamic Saivas is seen as bestowing
liberation: postinitiatory observance, where it is required,
serves merely to finish the task which initiation has almost
completed.
7. The Pacarthika system is open only to regenerate (upanta-)
brahmin men. The gamic Saiva systems are open to men and
148
women and to all four caste-classes, though in the Siddhanta
women were for the most part purely passive beneficiaries.
8. The Pacarthika system does not envisage its celibate
caryas accepting office as priests in temples. gamic Saivism
of the Saiddhantika variety did so, though this appears to be a
secondary development and though inscriptions reveal that
in course of time Pacarthikas too accepted this and other
compromises of their nature as world-transcending ascetics.
Not only are these two branches of the Sivasasana very different
in character. The dates at which they reveal themselves
to our observation are also widely separated.
The earliest indubitable evidence of the Pacarthika
tradition goes back to 380 AD, the date of the Mathura pilaster
inscription, and this testifies to a lineage of numbered generations
which might take the tradition back as early as the second
century. The earliest evidence for gamic Saivism is difficult to
interpret. For the most part we are reduced to cautious generaliza-
tions. Concerning the chronology of the early scriptural sources
of Tantric Saivism we can do little more than assert for most of
the texts known to us that they predate the citations that appear
in the works of the earliest datable commentators, that is to say, in
works of the tenth to early eleventh centuries from Kashmir or
Malava, and for a few of them, that they go back at least to the early
ninth century since they survive in Nepalese manuscripts of that
date, are recognizably paraphrased in the Haravijaya of Ratnakara
composed in Kashmir around 830, are listed in the text of the
Skandapurna preserved in a manuscript completed in 810, or are
mentioned as having been studied or practised during this period
in Saiva inscriptions from Cambodia.
Going back further than this we lose sight of titles and
can only establish that Tantric Saiva texts of certain familiar kinds
must have been present and that these or some of these were
probably works among those that were current later. Thus I
propose that a scriptural corpus of the kind we find later in the
Saiddhantika scriptures must have been in existence by the
beginning of the seventh century. There survive inscriptions
recording the Saiddhantika Saiva initiation of three major kings
during that century; and during the first half of the century the
Buddhist philosopher Dharmakrti (c. 600-660) goes to the trouble
149
of attacking the Tantric Saiva practice of initiation as the means to
liberation. These facts reveal that Tantric Saivism of this
relatively public and strongly soteriological variety was not
merely present in the seventh century but well established. And
this implies the existence of Tantric Saiva scriptures. For
while innovation in religious practice must have preceded its
scripturalization, it could not have survived without it, far less
reached such prominence. As is exceptionally clear in the case of
the Buddhist Yogintantras, the need to display a scripture as the
symbol of validity could be greater than the need that such
scripture be fully intelligible and coherent.
We also have some evidence from this period of the existence
of texts belonging to more esoteric, private and Siddhi-directed forms
of Tantric Saiva practice. In the first quarter of the eighth century
Bhavabhutis picture of Kapalika observances and doctrine in his
drama Mlatmdhava corresponds closely with what is seen in parts
of the Jayadrathaymala; his contemporary Vakpati makes an
incidental reference to the Kaulas; and in the first half of the seventh-
century Dharmakrti testifies that among holders of the soul-doctrine
there existed Mantrakalpas, texts of Mantra ritual, which taught
procedures involving the taking of life, theft and sexual congress.
He cites the Dakintantras and Bhagintantras as examples. His
contemporary, the poet Bana, mentions a Mantrakalpa manuscript
for the propitiation of a Mahakalahrdayamantra in his description
of a fictitious Tantric Sadhaka; and the terms Mantrakalpa and
Kalpa are used in exactly this sense in Tantric literature, to denote
the manuscript of a text setting out the procedure for the propitia-
tion of a Mantra. Finally, the jurist Bharuci, who may also belong to
the first half of the seventh century, refers to the Bhutatantras in
his commentary on Manu as sources teaching rites for the
mastering of Vetalas.
So much for the scanty evidence of the limits before
which Tantric Saiva literature can be seen to have existed.
For evidence of what it postdates, we must look to the texts
themselves. Since they have been composed as scripture, that is
to say, as transmissions of a timeless revelation, they are less
than generous in this respect. Their redactors seem to have been
careful to avoid references to historical persons and events that
would undermine faith by implying a terminus post quem, and
150
the device of prophecy, which would have allowed reference to
the past without this consequence, is rare in this predominantly
prescriptive literature. So one is reduced to trying to get the better
of the redactors by identifying elements in their texts which
they probably considered timeless facts but whose introduction
can nonetheless be dated, if only approximately. Thus there are
elements of Greek astrology and chronometry in some of the
early Saiva scriptures, elements such as the signs of the zodiac
(rsayah), their subdivisions into horh (hrai) and drekknh
(dekanoi), and the listing of the Grahas in the order of their
lordship of the weekdays. Texts with these elements can hardly
be earlier than the fourth century AD, though they may be three
or more centuries later. Some at least of the passages laying
down permissible temple-forms in early Tantras, and perhaps all
of them, postdate the treatment of this subject in Varahamihiras
sixth-century Brhatsamhit and this impression is supported by
what survives of early temple-architecture in India.
As for hard evidence of dependence on datable literary sources,
I have as yet little to offer. The Matangapramesvara paraphrases
the Smkhyakrik of svarakrsna (c. 350-400), and clearly echoes
the well-known definition of sense-perception formulated in the
Pramnasamuccaya of the Buddhist Dignaga (c. 480-540 AD) and
elaborated in the Nyyabindu of Dharmakrti (c. 600-660). The
*Jayadrathaymala echoes the Krik of Gaudapada (c. 500 AD).
And the *Brhatklottara, a rather late, eclectic text influenced by
non-dualism, echoes the Spandakrik of Bhatta Kallata, who
according to Kalhana flourished in Kashmir during the reign of
Avantivarman (855/6-883).
So by the beginning of the seventh century at the latest
there existed a Tantric Saivism of the kind known from the
early surviving literature, by which I mean a Saivism comprising
both of the two kinds of system whose scriptures the Saivas
distinguished as the ordinary (smnya-) and the extraordinary
(vaisesika-), that is to say both Saiddhantika Saivism and non-
Saiddhantika Saivism of some sort. Inscriptions indicate the
existence of the first and Dharmakrti and Bana that of the
second. It is quite possible that by the seventh century most of
the literature available to Saiva scholars in the tenth was already
in existence. But it is not until the beginning of the ninth that
151
we have firm evidence of specific texts. Our few witnesses
from that time show us the principal among the Saiddhantika
scriptures known later, and also texts of most of the major
divisions of the non-Saiddhantika tradition: the Vama, the Yamala
and, probably, the Trika. For our earliest evidence of certain divisions,
notably the Mantraptha and the Krama, and of the
majority of titles known to us in all categories, we have to wait
until the second half of the tenth century. But, of course, we cannot
make the absence of earlier evidence of these divisions and
texts the basis for an inference that they were not yet in existence.
Given the very different characters of the two kinds of Saivism
that are known to us through surviving texts and the evidence that
there might be as many as four or five centuries separating their
emergence, one is bound to wonder whether there were not
intermediate developments of which all evidence is lost or the
evidence for which has not yet been examined and correctly
evaluated. The purpose of these lectures, as their title indicates, is
to establish that there was indeed such a bridge between the
Pacarthika and gamic traditions and to establish what I can of
its character in the light of my knowledge of those two.
Conventional categorizations of Mahesvaras in non-
Saiva sources and some Saiva sources of a late date have
them divided into four groups. These, after the removal of
inconsistencies with which I shall not tax your patience, are
1) the Pasupatas,
2) the Lakulas, sometimes called Mahavratas or Kalamukhas,
3) the Saivas narrowly defined, and
4) the Kapalikas.
The Pasupatas of this list are the Pacarthika Pasupatas
and the Saivas are principally if not exclusively the Siva-
worshipping Saiddhantika Saivas. This leaves the Lakulas alias
Kalamukhas and the Kapalikas, whose seeming disappearance
from the corpus of Saiva literature gave David Lorenzen the title
of his 1972 monograph The Kplikas and Klmukhas. Two
Lost Saivite Sects, the term Kalamukha being merely a South
Indian version of the form Kalamukha seen in all northern and
some southern sources and confirmed by such commonly
152
encountered synonyms as Kalavaktra. Of these two the Kapalikas
are the followers of certain non-Saiddhantika systems, notably
those of the Vidyaptha, in which the practice of the kaplavratam
is emblematic. I shall not attempt to rehearse here the arguments
which have led me to this conclusion or to the identification
of their very substantial surviving literature. What concerns me
now is the identity and nature of other groups, the Lakulas or
Kalamukhas.
The principal source of the evidence I shall present is the
Nisvsasamhit. This work, which appears in all lists of the
Saiddhantika Saiva canon of scripture as one of the eighteen
Rudratantras, survives in a single palm-leaf manuscript (A) preserved
in the Nepalese National Archives in Kathmandu.
1
It is not dated
but it is written in a Nepalese Licchavi script which can be assigned
approximately to the period 850-900 AD. It is complete, but with
some loss of aksaras in the top and bottom lines at the right end
of the folios caused by physical deterioration. Fortunately some of
the aksaras now lost are preserved in an apograph (B) which was
prepared in 1912 for an agent of the Wellcome Institute for the History
of Medicine and can be consulted in its library. This difference
between the surviving codex unicus and its almost modern apograph
bears eloquent witness to the fragility of our evidence of the past
and to the urgency of the task of preserving the written evidence of
Indias unique cultural heritage by photographic reproduction and,
where possible, by conservation before more of it is lost forever.
The Samhit preserved in this manuscript comprises the
following sequence of five texts :
1. Nisvsamukha (nisvsamukhatattvasamhit): ff. 1v-18v6
2. Nisvsamla (nisvsatattvasamhitym mlastram):
ff. 18v6-23v1
3. Nisvsottara (samhitym uttarastram): ff. 23v1-29r5
4. Nisvsanaya (samhitym nayastram): ff. 29r5-42r5
5. Nisvsaguhya (samhitym guhyastram): ff. 42r5-114v.
The first work is seen as an introductory text, as its title suggests,
the four Sutras which following it forming the Samhita proper:
153
f. 41r4-5 (NiGu 1.1-3):
2
1 uparistac caturthan tu sutram arabhyate punah
tatra sutratrayam proktam bodhavyam anupurvasah
2 mula cottarasutra ca [nayasutran tathaiva] ca
guhyasutra caturthan tu procyamanan nibodha me
3 tenaiva saha samyukta samhitaika prapathyate
nisvaseti ca namena sampurnna tu tato bhavet
At the end of the work Siva invites Dev to ask him to go on to
teach the fifth Sutra, the Nisvsakrik:
B, f. 114v3 (18.15):
catvaro kathita sutra [samu]khadya varanane
pacaman tu param sutram karika nama namatah
sucita sutramatrena karikam [punah pr]cchatha
No manuscript of that work survives in Nepal to my knowledge.
But it has survived in the South. At its end we are told that the
whole Tantra in five Sutras is 12,000 slokas in extent (p. 1152: idam
pacastrojjvalam samudyena dvdasashasrikam samptam
nisvskhyam [conj. : nisvsakrikkhyam Cod.] tantram.) As we
have it, it is only about 1,500 short of this total, the text of the first
four Sutras together with the Nisvsamukha being 4500 verses in
the Nepalese manuscript and that of the fifth approximately 6000
in the South Indian.
It is probable that the Nisvsa is among the very earliest of the
Saiddhantika scriptures, belonging to a formative period well before
the ninth century, the time of the Nepalese manuscript. It shows a
greater awareness of pre-gamic Saivism than other texts of this
tradition; it contains a striking number of features that it shares with
that Saivism; and it shows elements of nondualistic practice that
suggest that the dichotomy between Saiddhantika and non-
Saiddhantika gamic Saivism, where the former is strictly dualistic
in this sense and the latter more or less nondualistic, had yet to
develop (Sanderson 2001: 29-31). Perhaps it was for these very
reasons that it soon fell into obscurity in India. By the time of the
first commentators in the late tenth century the text seems to have
been no longer in the front line. It was cited sparingly from time to
time but received no commentary and was not taken as the basis of
ritual, observance or doctrine. That it was not merely a pre-classical
text but a major authority in its time seems probable from various
154
evidences outside the corpus of later learned exegesis. (1) A passage
of the Varhapurna quoted by Apararka on Yjavalkyasmrti 1.7
equates the appearance of the Siddhanta in the Kali Age with the
revelation of the Nihsvsasamhit (sic).
3
(2) The Buddhist scholar
Padmavajra, probably of the ninth century, tells initiates into the
Buddhist Tantric system of the Guhyasamja (Guhyasiddhi 8.11-
16) to obtain a female consort by going to a family of untouchables
(Candalas) and pretending to be a Saiva guru. After initiating them
into the Siddhanta (Siddhantadharma) according to the Klottara
or else the Nisvsa he should ask for one of their girls as the
reward (daksin) for his services:
8.12 darsayec ca tatas tesam dharmam siddhantapurvakam
kalottaradisamsiddham no cen nihsvasasambhavam
12b dharmam siddhanta em. : dharmasiddhanta Ed.
12c samsiddham em. : samsuddham Ed.
Then he should reveal to them the Siddhantadharma established in
such [scriptures] as the Klottara, or else in that derived from the
Nisvsa.
(3) The Nisvsa was the Saiddhantika scripture that served as the
point of reference and source of much material for the important
non-Saiddhantika system of the Mantraptha taught in the
Svacchanda (See Sanderson 2001: 21-29). Finally (4), the Nisvsa
was in use among the Khmers for Saiva initiation ritual from at least
as early as the tenth century. An inscription (K. 532) of the reign of
the Khmer king Rajendravarman (944-968) tells us that a certain
Sivacarya, who, we are informed, was a Hotar not only of
Rajendravarman himself but also of his royal predecessors
sanavarman II (r. c. 923-c. 928), Jayavarman IV (r. c. 928-c. 941),
and Harsavarman II (r. c. 941-944), had become a celibate Saiva
officiant through initiation into the Mandala [of Siva] taught in
this scripture Nisvsa (v. 36: naisvsamandaln dksn naisthik-
cryyatrppanm / sivcryybhidhdhym yo bhisekavidhau
dadhau). It seems that the Nisvsa continued to be a pre-eminent
authority for Saiddhantika Saiva ritual in this part of Southeast
Asia until at least the end of the twelfth century and probably,
therefore, until the decline of Saiva culture in that region from the
end of the thirteenth century. That will follow if we conclude, as
I think we ought, that the Guhya mentioned in later inscriptions in
155
contexts that indicate that it must be a Saiddhantika work is the Guhya
that is the principal among the four Sutras of the Nisvsasamhit.
Thus in the Old Khmer section of the Sdok Kak Thom inscription
we are told concerning King Udayadityavarman II (r. 1050-1066)
(K. 235, D. 66-67):
vrah pada kamraten a thve vrah dksa damnepra gi
bhuvanadhva vrah vrahmayaja thve mahotsava
puja toy vrah guhya.
His Majesty accomplished the dks beginning with
the bhuvandhv and the brahmayaja. He
performed the Mahotsavas and Pujas in conformity
with the holy Guhya.
In the Phnom Sandak inscription of 1119 (K. 194) it is reported
that King Suryavarman II (r. 1113-c.1150) had studied all the
Siddhantas beginning with the holy Guhya (ll. 28-29: ryyan iss
siddhnta phon ta damnepra vrah guhya). And the Prasat Tor
inscription of 1189 or 1195 (K. 692, v. 5) mentions a locally
composed commentary on this text:
bhupendrapanditapadam munimastakali-
malanatam kajam iva pranamantu santah
sansarasindhubhuvanoddharanaya guhya-
tkapatha yad akarod yamasadma sunyam
Let the virtuous bow down to the foot of Bhupendrapandita
touched in reverential salutation by the heads of many sages
like a lotus by crowding bees, which emptied the abode of
Death through [his] commentary on the Guhya [composed]
to raise [souls] from the worlds within the ocean of
transmigration.
Finally, in the Prasat Khna inscription, probably of 1060 and in any
case from the reign of Udayadityavarman II (1050-1066), the author
Phalapriya describes himself as srsankarakaveh prptaguhyajnah
(K. 661, v. 108), which probably means that he received knowledge
of the Guhya from Sankarakavi, though it is possible that it means
only that he received secret (guhya-) knowledge of some kind from
him. That the Nisvsa continued in Cambodia to enjoy the central
position revealed by this evidence after its marginalization in India
156
may be seen as an example of the tendency of regions cut off from
metropolitan developments to preserve traditions in a relatively
archaic form.
The Nisvsamukha
The most important part of the evidence preserved in the
Nisvsa is found in the Nisvsamukha that precedes the Sutras
proper. There we are taught that Dharma, action prescribed by
valid scripture for the attainment of heaven and/or release, is of
five kinds, in an ascending order of excellence: the Mundane
(laukiko dharmah), the Vedic (vaidiko dharmah, vedadharmah),
the dhyatmika (dhytmiko dharmah), the Atimarga (atimrgah)
and the Mantramarga (mantramrgah).
f. 2r3 (1.21-22a):
nandikesvara uvaca
21 srnvantu rsayas sarve pacadha yat prakrtitam
laukikam vaidikam caiva tathadhyatmikam eva ca
22 atimargga ca mantrakhyam
22a atimarga ca mantrakhyam em. : atomarggas ca mantrakhya
[-8-] B : a[-7-] A
Each is said to have been revealed by Siva from one of
his five faces, the Mundane from the west-facing (Sadyojata), the
Vedic from the north-facing (Vamadeva), the dhyatmika from
the south-facing (Aghora), the Atimarga from the east-facing
(Tatpurusa) and the Way of Mantras from the upturned (sana).
Nisvsamukhatattvasamhit
f. 14v4-5 (3.191cd):
[pascime]naiva vaktrena laukikam gaditam sada
f. 16r6 (4.40c-42b):
vedadharmmo maya proktah svarganaisreyasah parah
41 uttarenaiva vaktrena vyakhyatas ca samasatah
adhyatmikam pravaksyami daksinasyena krttitam
samkhya caiva mahajanam yoga capi mahavrate
f. 18v3-4 (4.130):
130 atimarggam samakhyatam dvifprakaram varanane
purvenaiva tu vaktrena sarahasyam prakrttitam
130b dvifprakaram varanane em. : dvifpra[ + + + ]nare A :
dvisprakarava [+ ]nare B 130c purvenaiva tu vaktrena em. :
purvenaivaktrena AB (eyeskip)
157
f. 18v5 (4.133c-134b):
svarasya tu devasya mantramargam vyavasthitam
134 pacamenaiva vaktrena sanena dvijottamah
133c svarasya tu B : svarasya A 133d margam em. : marga AB
The same classification is seen elsewhere in gamic literature:
in the Mrgendra,
4
the Pauskarapramesvara,
5
the Svacchanda,
6
and the Jayadrathaymala.
7
The Nisvsamukhas description of Mundane Religion (laukiko
dharmah) is of the ordinary observances of the uninitiated but
regenerate (upanta-) householder devoted to Siva, comprising the
pj of Siva and other deities on the lunar days sacred to
them, donations to worthy recipients (dnam), pilgrimages to
Sivaksetras and so forth. The distinction between this and Vaidika
religion (vaidiko dharmah) is that the latter is the practice of the
celibate life-stages. It comes above the Mundane in the hierarchy
of paths because we are told that while the Mundane leads only
to heaven (svargah), this may go beyond that transient reward to
bestow [what it takes to be] liberation.
8
The third level, the
dhyatmika, which from the Saiva point of view is the highest of
the non-Saiva systems,
9
is the contemplation of the Samkhya
dualism of matter and spirit and its realization through the Yoga
system.
Passing for a moment over this level to the Mantramarga,
we are not surprised to see that this is gamic Saivism itself,
which is detailed not in the Nisvsamukha, but in the four Sutras
which follow it, and, of course, in the other Saiva gamas. The
Nisvsamukha, then, introduces the main text by expounding
the levels of religious observance which lead the soul by degrees
towards it and thereby to definitive liberation. It ends with the
identification of the fifth and a request that it be taught in detail.
f. 18v3-4 (4.130-131):
130 atimargam samakhyatam dvifprakaram varanane
purvenaivaktrena sarahasyam prakrttitam
ata urdhvam mahadevi kim vaksye paramesvari
devy uvaca
131 mantramargan tvaya deva sucitan na tu varnnitam
samsarocchittikaranan tam acaksva mahesvara
130b dvifprakaram varanane em. : dvifpra[ + + + ]nare A :
dvisprakarava [+ ]nare B
158
The Atimarga, which lies between the three non-Saiva systems
and gamic Saivism, is what concerns us here. As we
see from the verses I have just quoted the Nisvsamukha holds
it to be of two kinds (dviprakrah). It outlines the first, which
it calls the Observance of those beyond the Estates
(atysramavratam) in a rendering of the enigmatic prose
Psupatastra into verses that are clear (where they are not
lacunose through physical damage) and add a small amount of
information found neither in the Sutras nor in Kaundinyas
commentary. The first level of the Atimarga, then, is that of the
Pacarthikas. The rest of the section on the Atimarga introduces
us to a new form of devotion to Rudra, which it calls the
Kapalavrata (the observance of the skull), the Lokattavrata
(the observance of those beyond the world), and the
Mahapasupatavrata (the observance of the Greater Pasupatas).
11
It also refers to those who adopt this observance as the Mahavratas.
12
The questions before us are the identity of the followers of the
second Vrata, the nature of their practice and doctrine, and the
relations between their system and that of the Pasupatas on
the one hand and the gamic Saivas on the other.
However, before we move on to these, I wish first to
justify my use of this term Atimarga and to press for its adoption
in subsequent Indology to group these systems and certain other
satellites which need to be distinguished as a unity over and
against gamic Saivism.
Against its adoption it might be urged that the commentatorial
tradition of Kashmir understands this list and the Atimarga in it quite
differently. The usage of the Nisvsamukha might therefore be
thought eccentric. Thus when Ksemaraja comments on the same list
of five when it occurs at Svacchanda 11.43c-45b he does not see its
distinction between the Atimarga and the fifth as a distinction
between non-gamic and gamic Saivism. According to him - and
he is, after all, one of the most influential of gamic authorities - the
knowledge of the Atimarga mentioned in the text is knowledge
of the externals of gamic Saivism itself, while the fifth level is
knowledge of the core of the same system.
159
SvT 11.43c-45b:
laukikam devi vijanam sadyojatad vinirgatam
44 vaidikam vamadevat tu adhyatmikam aghoratah
purusac catimargakhyam nirgatam tu varanane
45 mantrakhyam tu mahajanam sanat tu vinirgatam
O goddess, the Mundane Knowledge came forth from Sadyojata, the
Vedic from Vamadeva, and the dhyatmika from Aghora. That of the
Atimarga emerged from [Tat]purusa, O you whose face is beautiful,
while the supreme knowledge, which is that of the [Way of] Mantras,
came forth from sana.
SvTU ad loc.:
l a uki ka m vartadandantyayurvedadhanurvedanatyavedadi
pratipadyakrsinayanayacikitsadivijanam.vaidikam nityanaimittika-
kamyayajadi<jana>svarupam. adhyatmikam samkhyayo-
gadipratipaditaprakrtipurusaviveka-janasarvavrttinirodha-
janadikam.atimargikam vedasamkhyayogadyuktopasatmaka-
prasiddhamargatikrantam samanyena paramesasastrapratipadita-
vividhamudramandala-kriyadyupayarupam vijanam ihabhipretam
na tu visistam catustayat. mantrakhyam iti tatraiva paramesesu
sastresu pacapranavadhi-karapratipaditantya mantresu a samantat
<khya> khyanam yasya tathabhutam yan mahajanam
mantravryadam janapadaprokta<m> kriyaditantratmakavijanad
vailaksanyenanubhavasaratam mantranam prathayati.
Mundane knowledge is that of farming, just and unjust judgement,
medicine, archery and the like, taught in [texts on] agriculture,
adjudicature, yurveda, Dhanurveda, Natyaveda and so forth. Vedic
knowledge is that of such matters as the regular, incidental, and
desiderative sacrifices. The dhyatmika is such as the knowledge
of the duality of matter and the soul and of the suppression of
all mental activity taught in the Samkhya and Yoga systems
[respectively]. The knowledge proper to the Atimarga, so called
because it is] beyond the well known paths of meditation taught
in the Veda, Samkhya, Yoga and the like, is that of the [practical]
means [of liberation] taught in the Saiva Tantras in general, such as
the various mudras, mandalas and ritual procedures, rather than
[some additional system] distinct from the [other] four [levels of
knowledge]. The supreme knowledge is that taught in the
Sections on Knowledge in those same Saiva Tantras. It is termed
mantrkhya-, which means [not relating to mantras (literally,
called [of] mantras) but] fully manifest in the mantras in the
manner taught in [this Tantras] chapter on the five pranavas,
because it bestows the power which animates the mantras. In a
manner quite different from that of the knowledge taught in the
160
Sections on ritual and the rest, it reveals that those mantras have pure
sentience (anubhavah) as their essence.
However, I propose that his explanation of the term Atimarga
is not that of the Svacchanda itself, and that on the contrary his
source exactly confirms the usage of the Nisvsamukha. This
conclusion rests on Svacchanda 11.179c-184. This classifies
the principal Indian soteriologies on the basis of the four
positive qualities of the faculty of understanding (buddhigunh),
namely virtuous action (dharmah), gnosis (jnam), detachment
(vairgyam) and power (aisvaryam). The passage, which I
translate without reference to Ksemarajas commentary, is as follows:
11.179c-184:
dharmenaikena devesi baddham janam hi laukikam
180 dharmajananibaddham tu pacaratram ca vaidikam
bauddham arahatam caiva vairagyenaiva suvrate
181 janavairagyasambaddham samkhyajanam hi parvati
janam vairagyam aisvaryam yogajane pratisthitam
182 attam buddhibhavanam atimargam prakrtitam
lokattam tu taj janam atimargam iti smrtam
183 lokas ca pasavah proktah srstisamharavartmani
tesam attas te jeya ye timarge vyavasthitah
184 kapalavratino ye ca tatha pasupatas ca ye
srstir na vidyate tesam svare ca dhruve sthitah
181d yogajane em : yogajana Ked
Mundane knowledge, O empress of the gods, is connected only
with virtuous action. The Pacaratra and the Vedic [knowledge] are
connected with virtuous action and gnosis. Buddhism and Jainism,
O faithful wife, have only detachment. The knowledge of the
Samkhyas, O Parvat, is associated with gnosis and detachment.
Gnosis, detachment and mastery are found in the knowledge of the
Yoga [system]. Transcending [all these] states of the understanding is
the [system] termed the Atimarga. That knowledge, the Atimarga, is
also known as the Lokatta [literally that which transcends the loka-
], where [the word] means [people rather than worlds,
specifically] those who are bound (pasavah) within the sphere of
emission and retraction [i.e. samsrah]. Know that those who
follow the Atimarga are [, like that Way itself, termed Lokatta,]
beyond those [souls in bondage]. Both those who have
followed the Kapalavrata and the Pasupatas are free of further
incarnation. They rest [after death] in svara: in Dhruva.
161
In his commentary on this passage Ksemaraja sticks to the
interpretation of the term Atimarga which he gave on 11.43c-45b.
For him the Atimarga mentioned after the Yoga system is not
that of the Nisvsamukhas exposition. It is still the Paramesvara,
that is to say, the Way of Mantras which follows the Atimarga in
the Nisvsamukha. But to carry this interpretation through he
has to abandon the obvious sense of the closing reference to
the Kapalavratins and Pasupatas. The text speaks of both
(ca) Kapalavratins and (tath ...... ca) Pasupatas (11.184ab:
kaplavratino ye ca tath psupats ca ye). Now, the meaning
and in this phrase is conveyed twice: by tath and by ca.
So, strictly speaking, one or other is redundant. Such redundancy
is entirely unexceptional in such writing, but is not, of course,
compulsory, and this gives Ksemaraja the opening he needs.
It enables him to propose a reading of the sentence which avoids
redundancy by taking the second conjunction to meaneven
rather than and. This move, which also requires the assertion that
the word is displaced (bhinnakramah)
13
, i.e. not where one would
normally expect it to find it if the meaning is as he claims, yields
the following meaning:
Know that those who follow the Atimarga are [like that Way
itself termed Lokatta,] beyond those [souls in bondage]. Both
those who have followed the Kapalavrata and the Pasupatas, are
free of further incarnation. They rest [after death] in svara: in Dhruva.
As he explains, the statement is now elliptical. It has been
made to mean -and I quote his gloss -:
SvTU 6 (11) 103,9-11:
ye kapalavratinah pasupatas ca tesam api srstir nasti. kim anga
sarvadhvottrnanam saivanam ity arthat.
Even the Kapalavratins and the Pasupatas are free of rebirth. How
much more the [gamic] Saivas [of the Atimarga] who transcend
[the followers of] all the other Ways !
It is clear, then, that the Nisvsamukhas use of the term Atimarga
is confirmed by the Svacchanda, and that Ksemarajas interpretation
is to be rejected. In both these gamas it refers to the systems of
Kapalavratins and Pasupatas.
Why Ksemaraja should have chosen to understand it differently
is not clear to me. I fail to see what was at stake. However, the
162
mere fact that the meaning of the term was open to such tendentious
interpretation suggests either that the traditions to which it
actually refers were little known in Ksemarajas circles, or, at least,
that they were not generally known under that name.
The same amnesia may also be suspected in the approximately
contemporary commentary of the Kashmirian Bhatta Narayanakantha
on the Mrgendra. That Siddhanta includes actions produced by each
of the same five levels of knowledge among the karma to be
eliminated by the officiants homa. It identifies Atimargic and Saiva
karma as actions produced by Yoga and Gnosis respectively.
Mrg KP 8.78-79:
lokamnayatimargabhisamdhisaivatmakany anoh
karmani.....
karmatatkrcchravairagyajanyani trisu dhamasu
yogavijanajanyani paratah paratah mune
The Mundane, Vedic, Internal, Atimargic and Saiva actions of the
individual ...... arising, O sage, from rites, from those and asceticism,
and from detachment in the [first] three levels, and from Yoga and
Gnosis respectively in the two beyond them.
Bhatta Narayanakantha notes ad loc. that the order of the text
(pthakramah), which gives the dhyatmika before the Atimarga,
is not the actual order (arthakramah).
This might suggest that the amnesia goes back to the
Tantra itself: if it meant by Yoga the Yoga of the Pasupatas
and Kapalavratins, why did it not say so? However, it is quite possible
that the author intended this narrower reference to be understood
from the context, which correlates the five levels with the five segments
(kal) of the universe of tattvas. The Atimarga, the penultimate level,
is therefore correlated with Santi, the penultimate kal; and that is
presided over by svara, the subordinate Siva-form which the Nisvsa
and the Svacchanda hold to be the point in which the followers of
the Atimarga attain their lesser liberation (apar muktih). It may
well be, then, that the equation of Yoga with the Yoga of the Pasupatas
and Kapalavratins was intended by implication. Such sastric brevity
is certainly a characteristic of this Siddhanta. Bhatta Narayana-
kantha, however, does not explain the implication. He simply
repeats the identification of Atimargic actions with actions
163
produced by Yoga and adds from the context their correlation with
Santikala.
14
One is therefore bound to doubt that he understood
anything more specific.
The term Atimarga, which I suggest we use for the non-gamic
Saivism of the Pasupatas and related systems, is extracted, then,
from a stage of the tradition which predates our famous commentators
and perhaps even some of the gamas themselves. But I make
no apology for putting it back to use: the dominion of these
commentaries over later tradition need not extend to us.
The second variety of the Nisvsamukhas Atimrga
While the Nisvsamukha is not the only source to use the
term Atimarga to refer to the systems of the Pasupatas and
Kapalavratins, it is, as far as I know, the only surviving source
to contain a systematic account of the practice and doctrine of
the second of these groups. After the verse prcis of the
Psupatastra in which it teaches the practice of the first group,
the text begins this account as follows:
f. 17v2-5 (4.87c-97):
atyasramavratam khyatam lokattam ca me srnu
88 alabdhah pacabhir guhyair ddksitas caiva so bhramet
khatvang ca kapal ca sa jat munda-m eva va
89 valayajopavt ca siromundais ca manditah
kaupnavaso bhasmang divyabharanabhusitah
90 jagad rudramayam matva rudrabhakto drdhavratah
sarvadas sarvacestas ca rudradhyanaparayanah
91 rudram muktva na canyo sti trata me devatam param
viditvaikadasadhvanam nirvisankah samacaret
92 prathame jalam etat tu dvitye murtisajakam
trtye pasur akhyatam pasas caiva caturthake
93 pacame vigrahah khyatah asuddhas te prakrttitah
asuddhamarggo vyakhyatah suddhamarga ca me srnu
94 yonir vagesvar dev pranavo yatra jayate
trtya caiva dhataram dhyana caiva caturthakam
95 tejsam pacaman khyatam dhruvam sastham prakrttitam
avcyadi dhruvanta ca etaj jatva vimucyate
96 krdarthasiddhaye caiva prakriyadhyanam asrita[h]
+ + vai prakriyadhvanam athasabdena dksayet
97 athasabdanipatena dksitas capasur bhavet
kriyavams ca duracaro mucyate natra samsayah
98 lokattam samakhyatam kim anyat pariprcchasi
164
Testimonia:
88c-89 valayajopavt ca siromundais ca manditah ...] Picumata f. 101r3-
4 (21.102b-104): adhuna samudayena vaksye vidyavratam subham /
bhasmoddhulitagatras tu jatamakutadharinah / candrarddhasekhara caiva
sirakapalamanditah / karnau sirasi bahubhyam asthikhandair vibhusitam
/ valayajopavt ca mekhalakatibhusitah; Nirmalamani ad
Aghorasivapaddhati (qu. SoSP III p. 680: *kanthika (em. : kundika Ed.)
kundalam caiva *rucakam (em. : uragam Ed.) ca sikhamanih /
kesayajopavtam ca mudra etah *mahavrate (em. : mahavratah Ed.); JY,
Satka 3, f. 201v3-5 (Yoginsamcraprakarana, 8.40c-42): dvityam tu
vratam vaksye ghorakapalarupina<m> / sire kapalamukutam
siramalavibhusitam / kare karnau tatha padau asthikhandair vibhusitam /
vame kapalam khatvangam tatha vai daksine kare / smasane vicaren maun
trisasti divasani tu; JY, Satka 3, f. 232r2-8 (38.156-166b):
kadyakhamdastakenaiva mumdamalam sucarcitam / pratikhamdam
karamkamkam krtva murddhni <-1-> dharayet / 157 khatvamgam
cchidrasampurnnam mahacamarasobhitam / crakimkanisamyuktam
sudamdam drdham avranam / 158 dharayed vamakaragam
nanasobhasamanvitam / karnnabhyam mudrike karye narasthighatite
subhe / 159 kamthamalam asthimayam latasaptakabhusitam /
pralambantam mahaharam tadvad eva *dharen (conj. : dharam Cod. )
narah / 160 kadyam daksinahastastham suramadyadipuritam /
mekhalasthimaya karya mundakhamdavibhusita / 161 kimkanbhih
samayukta narakesasamanvita / madhye naramanim tasyah karttavyam
kulisodari / 162 nupurav asthiracitau suksma*kimkanicitritau (conj :
kimkaricitrau Cod. ) / 163 vrabhasmasamalipto sopavtas samahitah /
madiranamdacaitanyo sugamdhakusumanvitah / 164 tambulavaktra-
sampamno madaghurnnitalocanah / hased vaharagam raudram
*paryateta (em. : paryatena Cod. ) samantatah / 165 kapaliko smi kamkal
rasmimelapalolupah / sarvabhakso pi pamcas vracakkresvaro hy aham /
166 evamvad bhaven nityam vicared vrarat sada; SvT 9.18-19a:
bhasmoddhulitadehas tu mudralankarabhusitah / kesayajopavt ca
digvasah samyatendriyah / sankharghapatrahastas tu
87d lokatta em. : [-1-]katta A : [-2-]ttam B 89d bhusitah conj :
divyabharanabhu[-2] B : divyabharana[-3-] A 93ab khyatah asuddhas
em. : khyata[-1-]ddhas B : khya[-2-]ddhas A
I have taught [you] the Atyasramavrata. Hear now the Lokatta. Touched
with the five Brahmamantras and initiated, he should wander. He should
carry a skull-topped staff (khatvngah) and [an alms-bowl fashioned
from] a human cranium (kaplam). He should have matted locks
(sajatah) or shave his head bald (mundah). He should wear a sacred
thread (yajopavtam) made from the hair [of the dead] and he should
165
adorn himself with a chaplet fashioned from human skullbones. He
may wear nothing but a strip of cloth to cover his private parts. He must
smear himself with ashes and decorate himself with celestial
ornaments. Seeing all things as Rudra in essence he should hold firmly to
his observance as Rudras devotee. He may eat and drink anything.
No action is forbidden to him. He should remain immersed in
contemplation of Rudra, thinking None but Rudra can save me. He is
the deity supreme. Provided that he has first understood the [Lakula]
cosmic hierarchy of eleven [levels] he should practice this observance,
remaining free of all inhibition (nirvisankah). On the first [level] is this
[lower universe which we call the] Net (jlam). On the second are the
Embodiments (mrtisamjakam) [= the Satarudrah, the Five Ogdoads
[pacstakni], the eight Devayonis, the eight Yogas, and the three
Lines of Gurus (gurupanktitrayam)]. On the third is the bound soul
(pasuh). On the fourth are the bonds (psh) (= Gahana up to Ananta)
and on the fifth are the Vigrahas. These are termed the impure [levels].
I have explained the impure cosmos (asuddhamrgah). Hear me now
as I teach the pure (suddhamrgah). [First is] the Womb (yonih),
Vagsvar, from which one is [re]born as Pranava [the second pure
level]. The third is [that of] Dhatr and the fourth is [that of] Dhyana.
The fifth is Tejsa[s] and the sixth is Dhruva[s]. When he has gained
knowledge of all this, from the lowest hell (Avci) [in the Net] up to
[the world of] Dhruva, he achieves liberation. In order to enable him
to accomplish his goal of sporting (krdrthasiddhaye) [in ever higher
levels of the universe the officiant] should first meditate on the
hierarchy of these levels. Then [when he has...] that hierarchy, he
should initiate [him] by means of the word atha. Initiated through the
descent of that word (athasabdaniptena) he will cease to be a soul in
bondage. Provided that [the initiate] maintains the observances he
attains liberation [at death], even if he is a sinner. Of this there is
no doubt. I have now explained the Lokatta. What else do you wish to
know? The goddess replied: I have learned these eleven levels
(tattvh) only as names. Explain this matter again in greater detail,
O Mahesvara.
The most obvious effect of this passage is that it shows
that these followers of the second form of the Atimarga stood
apart from the Pacarthikas by taking on, like the Kapalikas, the
visible attributes of the brahmin-slayer. As you know, the
Dharmasutras and other such orthodox sources rule that one who
is guilty of this crime may free himself from his sin only if he
removes himself from society for twelve years, living in a
cremation ground and begging for his food, carrying a skull-staff
166
and skull-bowl when he does so.
15
These ascetics stand between the
Pacarthikas and the Kapalikas, extending the range of the
Pasupatavrata into a more radical disregard for conventional
notions of ritual purity and intensifying the power of their
inauspiciousness, but without, it seems, transcending the conven-
tion of celibacy in the manner of the gamic Kapalikas.
A more elaborate listing of the levels of this Atimargic
cosmos then completes the chapter.
f. 17v5-18v6 (4.98-130b):
devy uvaca
98 ekadasaite tatvas tu namamatrena me srutah
punar vistaraso bruhi yatha vedmi mahesvara
99 [a]vc krminicayo vaitaran kutasalmal
giriyamala ucchvaso nirucchvaso hy athaparah
100 putimansadravas caiva trapus taptajatus tatha
pankalayo sthibhangas ca krakacacchedam eva ca
101 medosrkpuyahradas ca tksnayastundam eva ca
angararasibhuvanah sakunis cambar[sakah]
102 [+ + + + + + + + ]hy asitalavanas tatha
sucmukhah ksuradharah kalasutro tha parvatah
103 padmas caiva samakhyato mahapadmas tathaiva ca
apako[.]ara usnas ca sajvanasujvanau
104 statamondhatamasau maharaurarauravau
dvatrimsad ete naraka maya devi prakrttitah
105 [satastadhikasamyu]ktah + + + + + ]samyutah
catalsasatam hy etan narakanam prakrttitam
106 patalani pravaksyami nibodhaya yasasvini
adau mahatalan nama krsnabhaumam prakrttitam
107 rasatalan dvityan tu sphatikan tu prakrttitam
talatalan trtyan tu raityabhaumam prakrttitam
108 tamrabhauman tu nitala caturtham tu nigadyate
raityabhauman tu sutalam pacamam paripathyate
109 sastham vitalasajan tu ratnasarkkarasacitam
saptaman nitalan nama sauvarnnan tad udahrtam
110 kramena kathitas sapta pataladhipatim srnu
nagas ca garudas caiva tatha kimpurusantajah
111 agnir vayus ca varuno asuram patayas tatha
kathitas tu nivasinyo bhurlo[kam adhuna sr]nu
112 saptadvpasamudrantam varsavrksanagair yutam
vanopavanagudha ca nadbhis sagarair yutam
113 rsidevaganakrnam gandharvapsarasevitam
dharmmarthakamamoksan tu sarvam asmin pratisthitam
167
114 bhurlokah kathito hy esa bhuvarlokam ataf param
svarlokan tu tatordhvam tu maharlokam jana tapah
115 satya caiva tato[rdhvan tu brahma]lokan tatopari
visnos caiva niketan tu sivasya tu puran tatha
116 brahmanda esa vikhyatah kapalavaranair yutah
satarudras ca pacastau devayonyastakan tatah
117 yogastaka ca susivam gurupanktitrayan tatah
tatvasargam atordhvan tu kathyamana ca me srnu
118 pradhanabuddhyahankaratanmatranndriyani ca
bhutani ca tatha paca manas caivobhayatmakam
119 caturvimsati tattv[ani pu]rusah pacavimsakah
pacavimsakam etat tu satkausikasamudbhavam
120 matrjaih pitrjais caiva annapanavivarddhitam
gahana ca tatordhvan tu vigrahesam tatordhvatah
121 sivasankaram asadhyam harirudra dasesakam
pacasisyas tathacarya mahadevatrayam tatah
122 gopater granthir urdhvan tu murdhnabhibhavapacakam
ananta caiva pasas ca jalam etat prakrttitam
123 karyam duhkham tatha janam sadhanan tatvam eva ca
sastham sadhyam tathaisvaryam karana ca tathastamam
124 proktam visayam ajanam karanordhv ca kathyate
asuddhadhva samakhyatah suddhadhvana ca me srnu
125 mukta rsikulebhyas tu samsarac ca duratyayat
yonya capy atha vagesyam jatah pranava ucyate
126 dhataran damana caiva svaram dhyanam eva ca
bhasmsa ca samakhyatam pramanastakam eva ca
127 vidyastakam ca murtyastau tejsas ca dhruvas tatha
iti sankhya samasena suddhadhvanah prakrttita
128 kapalavratam asritya dhruvam gacchanti tat padam
lokattam samakhyatam mahapasupatam vratam
129 prakriyacaryasamyukto dhruvam gacchati tat padam
vipluto narakam yati prakriyacaryavarjitah
130 atimarggam samakhyatam dvifprakaram varanane
Testimonia:
108c raityabhauman tu] NiGu ff. 54v6-55r3 (on these Patalas): ayas
prathama bhumih ...... dvitya sphatika bhumih ...... raityabhumi trtye tu
...... tamrabhauma caturthan tu ...... pacamam raityabhaumam tu 115b
brahmalokan tatopari / visnos caiva niketan tu sivasya tu puran tatha]
Nisvsamla f. 20v6 (5.7): tapolokam tatah prapya satyalokan nayet punah
/ brahmavisnupuran ntva punah sivapuran nayet 116b kapalavaranair
yutah] NiGu f. 44r3-4 (1.114-115b): brahmavisnupura caiva etesam upari
sthitam / sivasya tu puran divyam kapalavarane sthi[tam] / [bra]hmanda
esa vikhyatah satarudras ca bahyatah; NiGu f. 63v3 (7.81-82b): dasa paca
168
ca kotbhih purad urdhvam vinirddiset / *yavadandakapalas (em.:
yavaddandakapalas A) tu sa saptavara-norddhvatah / brahmanda esa
vikhyato satarudrani me srnu 122c anantam caiva pasas ca] NiGu f. 67v2
(7.242ab) (=SvT 10.1132b): ata urdhvam anantesa pasas caiva tu samsthitah
123] NiGu f. 67r1-2: [athatah] sampravaksyami vigrahan tu yathasthitam
/ karyan duhkhan tatha janam sadhanan tatvam eva ca / sadhya caiva
tathaisvarya karana ca tathastamam; SvT 10.1089: karyam ca karanam
caiva sukhaduhkhakaram tatha / janam sadhyam ca vikhyatam sadhanam
karanam tatha 125a mukta rsikulebhyas tu] Nisvsamla f. 21r2 (5.12c-
13b): yojya brahmapatau sive / madamohesu samyojya tatha risikulesu
ca; NiGu f. 67v2 (7.242c-243b): purvokta eva ye pasa urdhvam rsikulam
smrtam / yonir vagesvar caiva yatra jato na jayate 127c murtyastau]
Nisvsamla f. 21r2 (5.15ab): pramanastaka vidyastau murtisv astasu
yojayet; NiGu ff. 67v2-68r1 (7.243-258) (>SvT 10.1132c-1162,
Tantrasadbhva ff. 86v1-87r4): 243 yonir vagesvar caiva yatra jato na
jayate / onkaram sadhya dhataram damanesan tatha smrtam / 244 dhyanad
urdhvan tu bhasmesah kramena kathitani tu / pramanani tatordhvan tu
procyamanani me srnu ... 246ab atah param bhaven maya (bhaven maya
SvT : mahamaya Tantrasadbhva) sarvajantuvimohan ... 250 tatopari
mahavidya sarvavidyasamavrta ... 252 astabhedavibhinna tu vidya sa eva
pathyate / vamadyais saktibhir bhinna punas ca paripathyate / 253 mayopari
mahamay sarvakaranakaranah / astavidyesvarair yukto vtarago
nirajanah / 257cd anantesa mahatmanah suksmas sivottamottamah /
258 ekanetraikarudras ca [trinetras ca][68r]naladyutih / srkanthas ca
sikhand ca astau murttidharah smrtah
101d sakunis cambarsakah em. : sakunisvascar[...] A :sakunisvas-
carpakah B 104b maharaurarauravau em. : maharaurava[-1-]navau B :
mahara[-1-]vara[-2-] A 104d devi em. : dev AB 105ab satastadhika-
samyuktah conj. : satastadhikasamyu[] B : [] A 105c catalsasatam (Aisa,
from MIA cattalsa forty) em. : sacatalam satam AB 108c raityabhauman
em. : [-3-]man AB 112b vrksa em. : vrrksa A : vrrkka B 113b apsarasevitam
em. : apsarassevitam AB 116b kapalavaranair em. : kapala[...]ranair AB
118ab buddhyahankara em. : budhyahankara AB 121b mahadeva-
trayam tatah conj. : mahadevatraya[-2-] AB 122c ananta caiva pasas ca
em.: ananta caiva [-1-]as ca A : ananta caiva [-1-]as ca B 123d karana
em. : karana AB 125a mukta rsikulebhyas conj. : mu[-1-]rsi A : mu[-1-]
si B 127a vidyastakam ca murtyastau conj. : vidyastakam ca [murtya]
stau B : vidyastaka[m ca murty]astau A 127a prakrttita conj. : prakrttitah
AB (taking suddhdhvnah as gen. singular with its penultimate syllable
lengthened for fit the metre) 128a asritya corr. : asrtya AB 128d
mahapasupatam vratam em. : mahapasutascratam A : mahapasutam
vratam B 130b dvifprakaram varanane em. : dvifpra[-4-]nare A:
dvisprakarava[-1-]nare B.
169
My purpose now is to develop my identification of these
celibate skull-bearers by collating this account with certain
other remarks scattered in the Saiva gamic literature.
The Eight Pramas
Firstly, it now appears that certain passages in the corpus
of Saiva gamas are speaking of the Nisvsamukhas second
level of the Atimarga when they refer to a system of salvation
called the Pramana (i.e. pramnasstram). Thus Svacchanda 11.71-73b:
71 vrate pasupate proktam aisvaram paramam padam
mausule karuke caiva mayatattvam prakrtitam
72 ksemeso brahmanahsvam tesam tat paramam padam
tejeso vaimalanam ca pramane ca dhruvam padam
73 kapalavratam asthaya svam svam gacchati tat padam
In the observance of the Pasupatas the highest attainable level is said
to be that of svara [, the thirty-third tattva, and the second in ascent
within the pure segment of the universe]. In the Mausula and Karuka
[systems] that limit is within mytattvam [, the highest of the levels
of the impure cosmos]; for the ultimate goal of their followers is
Ksemesa and Brahmanahsvam respectively. For the Vaimalas,
however, the goal is Tejsa; and for the Pramana system it is Dhruva.
Purified by initiation and by knowledge, and by practising the
Kapalavrata, following their discipline until they die, these reach their
respective goals. Devoting themselves to the practice of the mantras
(japah) and to the ash-rites (bhasmakriy) they go [at death] to the
level of svara.
The identity of the Nisvsamukhas Kapalavratins and
the followers of the Pramana system described in this passage is
evident here from the fact that Tejsa and Dhruva, identified as
the ultimates of the Vaimala and Pramana systems respectively, appear
in the Nisvsamukha as the two highest worlds in the pure segment
of the Kapalavratins cosmos.
The same identity is established by a similar passage in
the Hierarchy of the Worlds contained in the Nisvsaguhya, the most
detailed of the four Sutra texts that the Nisvsamukha precedes and
introduces:
f. 68r1-2 (7.261-262b):
261 tejsas ca dhruvas caiva pramanadhvana krttitam
kapalavratam asthaya pramanagamasiddhaye
262 gata dhruvapadam ye tu dksajanavisodhitah
170
261d pramanagamasiddhaye conj. : pramanagamasi[ + + ] AB
[Then] come Tejsa and Dhruva, taught in the world-hierarchy of the
Pramana system. Purified by initiation and gnosis and having
practised the Kapalavrata in order to fulfil [the command of] the
Pramana scriptures those [ascetics] go to the realm of Dhruva.
Likewise Svacchanda 10.1174ab:
tejesas ca dhruvesas ca pramananam param padam
Tejesa and Dhruvesa, the ultimate goal of the Pramana[ scripture]s.
Here the Svacchanda appears to contradict its assertion in
11.72cd quoted above that Tejesa is the goal of the Vaimalasastra
and Dhruva (/Dhruvesa) alone the goal of the Pramanasastra.
But a remark in the Nisvsakrik shows us how the two
statements may be reconciled:
Dksottara A p. 901, B p. 82 (7.78):
tejsas ca dhruvsas ca pramane paramam padam
dvau rudrau tu samakhyatau yogamoksaphalaprada
78a dhruvsas em. : bruvsas B : bravsas A
Tejsa and Dhruvsa are the ultimate goal in the Pramana [system].
These two Rudras are held to bestow the rewards of Yoga and Moksa.
Evidently, the intention of this passage is to explain that Tejsa
and Dhruvsa bestow Yoga and Moksa respectively. Tejesa, the
Vaimala domain of final liberation has been capped in the
Pramana system by the world of Dhruvesa. It is now redefined as the
domain of the state of Yoga that is the immediate antecedent of final
liberation.
A further passage in the Hierarchy of the Worlds of the Nisvsa
and the Svacchanda - the text is almost identical in the two
Tantras - gives us the names of these Pramana scriptures, personifying
them as a set of eight Rudras with worlds within
the my level of the gamic cosmos:
SvT 10.1134-1135b NisvSam f. 67v2 (NiGu 7.243-246b)
yonir vagsvar caiva 243 yonir vagesvar caiva
yasyam jato na jayate yatra jato na jayate
omkarasadhyadhataro omkaram sadhya dhataram
damanesas tatah param damanesan tatha smrtam
dhyanam bhasmesam evahuh 244 dhyanad urdhvan tu bhasmesam
kramena kathitani tu
171
pramanani tatdurdhvatah pramanani tatordhvan tu
procyamanani me srnu
pacartham guhyam evahu 245 [paca]rtham sivaguhyan tu
rudrankusam atah param rudrankusam ataf param
hrdayam laksanam caiva hrdayam laksana caiva
vyuham akarsam eva ca vyuham akarsakan tatha
adarsa ca tathaiveha adarsam ca tathaiveha
astamam parikrttitam astamam parikrttitam
245a pacartham em. : [-1-]rtham AB
245d akarsakan conj. : akarsakan A
246a adarsa em. : akarsa AB
Thus:
1. Pacrthapramna
2. Guhya (SvT) or Sivaguhya
3. Rudrnkusa
4. Hrdaya
5. Vyha
6. Laksana
7. Akarsa
8. Adarsa
Ksemaraja comments:
ete rudra etannamakapasupatasastravatarakah
These Rudras are the propagators (avatrakh) of the Pasupatasastras
which bear these names.
and adds that the fourth of these works, the Hrdayapramna,
contains six subsidiary Pramanas which unlike the main Pramanas
are principally concerned with ritual rather than knowledge. He
associates these with the inferior division of the Pasupatas called
Mausula, saying that its originator Musulendra, a disciple of
Lakulsa, removed these six Pramanas from their proper context
in the gnostic Hrdaya and propagated them independently for less
advanced practitioners. This, he says, is why they have not been
included in the Svacchandas list:
tatra ca hrdayakhyam yat pramanam uktam tasyantarbhutani yani
purakalpakanakasalaniruttaravisvaprapacakhyani sat kriya-
pradhanani pramanani proktajanapradhanapramanastaka-
vilaksanani hrdayakhyat pramanal lakulesasisyena musulendre-
noddhrtya aruruksunam prathamam pradarsitani na tanha prthag
ganitani.
172
This tradition that there were really fourteen Pramanasastras,
divided into eight primary, gnostic texts and six subsidiary texts
concerned with ritual is also seen in the Jayadrathaymala.
The bhuvandhv of its first satka contains in its Maya level
fourteen Mahadevas described as masters of the Pramanasastras,
divided into eight with names identical with the eight Pramanas
already listed and said to be devoted to reasoning and gnosis
(hajnapravrttisthh) and six others, who are described as expert
in the rites of the various gamas (nngamakriyvidah) :
f. 66v6-8 (9.543-546b):
543 remus tatraiva te vrah pramanarthavicarakah
pacarthankusahrdguhya-uhalaksanakadayah
544 adarsakarsasamyuktah sarvavidvedaparagah
vidyadhipativresatarodayaviramagah
545 caturdasa mahadevah prabuddha nirahankrtah
uhajanapravrttistha hy adyastau ye prakrtitah
546 mahavisaradah sat ca nanagamakriyacitah
543b pramanartha conj. : pramanapu Cod.
16
544a adarsakarsa em. :
adarsakarsa Cod. samyuktah em. : samyukta Cod. 544b sarvavid
conj. : sarvacid Cod. 545b nirahankrtah conj. : nirahankrtah Cod.
545c uhajana em. : uhajana Cod.
The names of the Rudras and those of the six ritualistic Pramanas
do not tally. But that is not an objection to the assumption that the
former are the revealers of the latter, since when lists of scripture-
revealing Rudras appear elsewhere in our gamic hierarchies of worlds
they sometimes do and sometimes do not share their scriptures names.
As for the titles of the six works, unfortunately Ksemaraja has
given them in a Dvandva compound with seven members:
purakalpakanakaslniruttaravisvaprapackhyni. One
title must therefore be a compound of two members, but which
it is cannot, of course, by determined without additional infor-
mation. Help is at hand from the South Indian Dptgama,
which in its chapter on the characteristics of proper caryas
(Acryalaksanapatala) gives lists of the scriptures of the main
divisions of the Saivas. These lists are garbled and chaotic; the
scriptures of the Lakulas are wrongly assigned to the Kapalikas;
and the verses in question are highly corrupt in the witnesses
I have consulted. Nonetheless, the scriptures in question are
said to be fourteen in number:
173
A p. 741; B p. 134 ; C pp. 133-134:
pacartham ca purakalpam sivaguhyapramanakam
11 rudramsam hrdayam caiva tallaksanapramanakam
vyuham adarsakam caiva padmahatakasalakam
12 niruktam caiva prapacam siddhartha ca caturdasa
kapalam eva vyakhyatam lakulam srnu canala
10c pacartham em. : pacardham ABC purakalpam BC [= SvTU ad
10.1134-5b] : purakulah A 10d sivaguhya em. : sivaguhyam A : sivagu[...]
B : sivaguru C pramanakam conj. : pramapakam A : lacuna BC
11a rudramsam ABC, for unmetrical rudrankusam 11b pramanakam
AB : pramanatah C 11c vyuham adarsakam conj. : vyomadarsanam B
vyomaddarsanam A : vyoma ca darsanam C 11d visva conj. : padma AC :
ya ... B hataka conj. : kotaka A : khetaka C : [-3-] B salakam conj. :
sailakam AC : [-3-] B : 12a niruktam caiva AB : trisulam caiva C
prapacam B : prapaca AC 12b siddhartham ca caturdasa conj. :
siddhartham caturdasa BC : siddha[-1-] caturdasa A
and by collating this list with that of Ksemaraja we can deduce
that his compound should be resolved to give the following six Pramanas:
9. Purakalpa(Dpta: Purakalpa);
10. Kanaka(Dpta: Hataka);
11. Sala(Dpta: Salaka);
12. Niruttara(Dpta: Nirukta);
13. Visva (Dpta: Padma);
14. Prapaca
The names do not occur in either of the Nisvsamukhas
accounts of the Atimargas Hierarchy of the Worlds. But the
longer account of the two does refer to eight Pramanas. This
is how it describes the levels which make up the pure segment
of the universe (suddhdhv):
18v2-3 (4.125c-127b):
yonya capy atha vagesyam jatah pranava ucyate
126 dhataran damana caiva svaram dhyanam eva ca
bhasmsa ca samakhyatam pramanastakam eva ca
127 vidyastakam ca murtyastau tejsas ca dhruvas tatha
127a vidyastakam ca murtyastau conj. : vidyastakam ca [murtya]stau
B : vidyastaka[m ca murty]astau A 127a prakrttita conj. : prakrttitah
AB
174
When it has been regenerated in the womb which is the goddess
Vagsvar, [the soul] is known as a Pranava. [The levels above that
are as follows:] Dhatr, Damanesvara, Dhyana, Bhasmsa, the Eight
Pramanas, the Eight Vidyas, the Eight Murtis [, who are the Lords of
those Vidyas], Tejsa and Dhruva.
The possibility that these unidentified Pramanas are entities
other than the eight Atimargic texts listed in the Nisvsa, the
Svacchanda, and the Jayadrathaymala is all but removed by the
fact that the sources which name the Pramanas frame them in an
almost identical set of worlds, as can easily be appreciated from
Fig. 1.
Fig. 1.
The Atimrgas pure cosmos and the corresponding
section of the cosmos of the Nisrsa and the Sracchanda
ATIMRGA Nisvsaguhya and Svacchanda
Pure Universe:
...............
Pure Universe: Ekaksa, Pingala, Hamsa
Dhruva Dhruva
Tejsa Tejsa
svara [the paramam padam accessible to
the Pasupatas]
Eight [Vidyesvaras] Eight Vidyesvaras: Ananta, Suksma,
Sivottama, Ekanetra, Ekarudra, Trinetra,
Srkantha, Sikhandin
Mahavidya, Vagsvar (II)
Eight Vidyas Eight Vidyas: Vama, Jyestha, Raudr, Kal,
Vikaran, Balapramathan, Bhutadaman,
Manonman
Impure Universe:
MY (Yoni II)
Eight Pramanas Pramanas: Pacartha, Sivaguhya etc.
Bhasmsa Bhasmesa
Dhyana Dhyana
Damanesvara Damanesvara
Dhatr Dhatr
Sadhya
Pranava Omkara (Pranava)
175
Yoni, Vagsvar Yoni (I), Vagsvar (I)
Impure Universe
Rsikula Rsikula
............ ............
To my knowledge all that remains of this canon of texts is
a passage of seven verses attributed to the Pacrthapramna
in Ksemarajas commentary on the Svacchanda (ad 1.41-43). Short
though it is, it does serve to confirm that the Pramana system is the
basis of the system described by the Nisvsamukha. It is an analysis
of the Aghoramantra, one of the five Brahmamantras which are the
mantras of the Pasupatas. The form of that mantra is the offering
of obeisance (namaskrah), or, as the Pasupatas understand that
action, the donation of oneself (tmasamarpanam, paritygah),
to the three classes of Rudras, or ectypes of Rudra: the Benevolent
(aghorah), the Terrible (ghorah), and the Utterly Terrible (ghora-
ghoratarah): aghorebhyo tha ghorebhyo ghoraghoratarebhyas
ca sarvatah sarva sarvebhyah namas te rudra rpebhyah. The
Pacrthapramna correlates these three groups with the Rudras
who govern the levels of its Hierarchy of Worlds. The Aghora
Rudras of the Mantra are equated with a series of Rudras
beginning with Vamesvara which reside above the Net of Bonds.
The Ghora Rudras are said to be those that begin with Gopati
and end with Gahana, while the Ghoraghorataras are identified
as the Mahamahesvaras from Vidyesvara to Ananta who
occupy worlds below them. Now the world-hierarchy text of
the Nisvsamukha is incomplete in its information on the Rudras
of the various levels. A complete correlation is therefore impossible.
However, the series from Gopati to Gahana which is identified
as the intermediate Rudras (the Ghoras) is a well-defined group in
the Nisvsamukhas more detailed version, and one that is excellently
positioned to be considered intermediate. For it is
just above the Tattvasarga, which contains the raw materials
of embodiment as a bound soul, and just below the pure universe
of the liberated.
Beyond this single quotation by Ksemaraja the only specific
evidence of this canon that I have detected consists of two
references to the names of individual texts within it. The first
occurs in the Paramoksanirsakrikvrtti composed c. AD
176
1000 by the Kashmirian Saiva scholar Bhatta Ramakantha. He
refers to such texts as the Hrdayapramna while refuting the
view of liberation which, as we shall see, was held by this
tradition.
17
The second reference is in Sayana-Madhavas fourteenth-
century account of the Lakulsvara-Pasupatadarsana. He speaks of
the Trthakaras, the propagators of the Pasupata system, as the
authors of the Adarsa and the rest.
18
The quotation and the two
references do at least prevent us from supposing that we know so
little about this area of the Saiva tradition because it was already
virtually extinct in the period of the commentators. Inscriptional
evidence from South India also proves that it flourished well into
this period, as we shall see presently.
Lkulasstra and the Lkulas
In his commentary on the passage of the Svacchanda
quoted earlier, in which it was said that in the observance of the
Pasupatas the attainable goal is svara, Ksemaraja points out
that this applies only to the Pasupata system taught by Lakulsa,
not to that taught by his disciple Musulendra. For this division
within Pasupatism he quotes the following without attribution (SvTU
6 (11) 52):
lakulam mausulam caiva dvidha tantram prakrtitam
The [Pasupata] canon is held to have two divisions: the Lakula and the
Mausula.
That the first comprises the eight Pramanas is evident from
his commentary on the passage of the Svacchanda which lists
them (10.1134-35b):
ete rudra etannamapasupatasastravatarakah. tatra ca hrdayakhyam
yat pramanam uktam tasyantarbhutani yani purakalpakanaka-
salaniruttaravisvaprapacakhyani sat kriyapradhanani pramanani
proktajanapradhana-pramanastakavilaksanani hrdayakhyat
pramanal lakulesasisyena musulendrenoddhrtyaruruksunam
prathamam pradarsitani na tanha prthag ganitani. evam lakulam
mausulam caiva dvidha tantram prakrtitam <ity [uktam ?]>.
These Rudras (Pacartha, Guhya, Rudrankusa, Hrdaya, Laksana,
Vyuha, karsa and darsa) are the propagators [on earth] of the
Pasupata scriptures which bear their names. Of these the Pramana
called Hrdaya contained six further Pramanas: the Purakalpa-, the
Kanaka-, the Sl-, the Niruttara-, the Visva- and the Prapaca-.
These were mainly concerned with ritual and were therefore quite
177
different from the eight Pramanas listed, since the emphasis of the
latter was on gnosis. [These six lesser Pramanas] were extracted
from the Hrdayapramna by Musulendra, one of Lakulsas disciples,
who propagated them as a preliminary teaching for beginners. Of the
Pasupata [corpus] therefore we read: The canon is held to have
two divisions: the Lakula and the Mausula.
The term Lakula is also used to describe the followers
of the Lakula text-corpus.
19
Lkula and Psupata
The passage just quoted divides Pasupatism into two
branches, the Lakula and the Mausula, and takes these terms
to mean taught by Lakulsa and taught by Musulendra.
Presumably, if Ksemaraja had been required to do so, he would
have had to justify this analysis by applying the rule that one may
use the first member of a compound name in place of the full form,
the standard example being the use of Bhma in place of Bhmasena.
This implies that he took the correct full form to be Lakulesa
or Lakulesvara rather than Lakulsa or Lakulsvara, since only
the former pair can give the shortened name Lakula necessary
to derive the word Lakula in the meaning taught by Lakulesa.
Now the Psupatastra, as we have seen, was also taught
by Lakulesa. It might be thought therefore that Ksemaraja
must have considered that work to be part of the Lakula, non-
Mausula branch of the corpus. In that case, does this Lakula/
Mausula dichotomy obliterate the Pasupata/Lakula dichotomy?
The answer seems to be that Ksemaraja did see the Lakulas as
following Lakulsas Psupatastras as their fundamental authority,
but that he still recognized the Pasupatas as a group outside
the Lakulas. Evidence for the first is seen in the commentary
introducing Svacchanda 11.74ab:
20
ye tu kapaladyasthivratadharinah purvoktalakulamnaya bhasmani
saytetyadipasupatasastracodanatah japabhasmakriyanisthas te
vrajanty aisvaram padam
lakulamnaya em : lakulamnayad Ed.
Those who adopt the observance of human bones, in which they
carry a skull and so forth, that is to say, those who follow the
178
Lakula tradition (Lakulamnaya) attain the domain of svara,
devoting themselves to the practice of the mantras and to the
ash-rites in response to such injunctions of the Pasupata scripture
as He should lie on ashes.
The injunction cited here is PaSu 1.3 (bhasmani sayta).
Evidence for the second is his commentary on Svacchanda
4.391c-392, in which he distinguishes Saivas and Pasupatas from
the Lakulas.
21
Each of these groups, he says, adheres to the view
that there exists an irreducible plurality (nntvam) of souls.
Mahvratas
Among the terms used by the Nisvsamukha to refer to
these skull-bearing, Dhruva-seeking ascetics of the Atimarga is
Mahavrata. In principle this could be no more than a description,
and might have either a more general or a more specific reference.
In the first case it would mean one whose observance is great, in
the second one who has adopted the Great Observance. In
the latter meaning the term would be referring specifically to
the Kapalavrata, the terms mahvratam and kaplavratam
being synonymous in this sense, and therefore to a feature which
was not peculiar to them as followers of the Pramanas, since it is
also characteristic of various kinds of gamic Saiva asceticism
and its Buddhist parallels.
However, when the Nisvsamukhas Atimarga-text is
read in the light of certain other remarks in the Saiva literature it
becomes apparent that the term is being used by the Nisvsamukha
to identify them by name. Thus, in his commentary on Sadyojyotis
Naresvaraparks Bhatta Ramakantha writes the following
(p. 248):
ye pi mahavratah saivabhidhanena prthivyadidhruvantan
arthan sarvam iti kathayanti te sarvajam prstah santah
sarvasabdartham nirapavadam vaktum na saknuvanty eva.
As for the Mahavratas, who though they call themselves Saivas
apply the word universe (lit. the all) to the [series of] realities
from Earth to Dhruva, they are quite unable to give a stable
definition of all, when asked what they mean by the word
all-knowing (sarvajah).
179
He also attacks these Mahavratas in his commentary on
the Matangapramesvara for believing that the material cause of
the impure universe, which they call the knot (granthitattvam),
is internally differentiated:
MatParVr VP, p. 261, lines 15-17:
tac ca suksmam nyagrodhabjakanikavat karanatvena
prak pratipaditatvat. na tu sthulam puraprakaradimad yatha
mahavratair granthitattvam abhyupagatam. tathatve hi karyatvat
paramakaranatvanupapattih.
And this [mytattvam] is subtle, like the minute seed out of
which the banyan tree develops, because it has already
been taught as the cause [of the impure universe]. It is not a gross
entity comprising worlds, barriers and so forth, like the knot
(granthitattvam) in which the Mahavratas believe. For if it were,
it would have to be an effect, and could not therefore be the
first cause [of that universe].
The Klamukhas, Klmukhas, Klavaktras
Having established this identity between our sect, Bhatta
Ramakanthas Mahavratas and the Lakulas we can go on to
demonstrate its identity with a tradition of Raudra asceticism
known up till now almost exclusively from epigraphs. These
record the preceptorial lineages of gurus and the temples and
mathas over which they presided, but little of their practices and
almost nothing of their beliefs. The ascetics to whom I refer are
the Kalamukhas studied by David M. Lorenzen (1972).
The name Kalamukha appears in North Indian and Kashmirian
sources in the form Kalamukha or its synonym Kalavaktra; but the
epigraphs which are the basis of Lorenzens study come exclusively
from south India. They show orders of these ascetics (parisat, val,
mnyah) flourishing in Karnataka, Andhra and Mysore from the
ninth to the thirteenth centuries.
a. The synonymity of the terms Mahvrata and Klamukha
In his Saivaparibhs Sivagrayogndra Janasivacarya, writing
in the Far South in the sixteenth century, describes and challenges
four Saiva views concerning liberation. In each the end result is
180
defined as a state of equality with Siva (smyam). They differ as to
how that state comes about. According to the Mahavratins,
we are told, it is something produced in the individual (the
smyotpattivdah). According to the Pasupatas it is transferred
to him from Siva, as one lights one lamp from another
(smyasamkrntivdah). According to the Kapalikas it is
comparable to a state of possession (smyasamvesavdah).
According to the Saivas, or rather to some of them (saivaikadesinah),
since the author represents a breakaway tradition influenced
by nondualistic Vedanta, it already exists in the individual and
is simply made manifest (abhivyaktivdah). This schema of the
four views goes back to our earliest philosophical exegesis on
the gamic tradition of the Siddhantas. For it is already seen in
the late seventh or early eighth century in Sadyojyotis
Paramoksanirsakrik:
7 samata samutpattisankrantyavesapaksatah
abhivyaktih para gta buddhivacam agocara
7c abhivyaktih para em. : abhivyaktipara Ed.
Sadyojyotis does not identify the schools that follow these views;
but the Kashmirian Bhatta Ramakantha does. In his commentary
on this verse he tells us that those who believe that liberation is
an extension of gods qualities to the individual are the Pasupatas:
p. 12, l. 17:
sikhasamkrantivadinah pasupatah
Those who hold that there is an extending [of the qualities
of Rudra, in the manner] of a flame [spreading] are the Pasupatas.
and he conveys the same in his commentary on Sadyojyotis
Naresvaraparks 1.62:
p. 10, l. 19:
pasupatas tu muktasyesvaragunasamkrantes tatsamatvam ahuh
As for the Pasupatas, they say that [liberation is] a state of equality [with
Rudra which comes about] through extension of the qualities of God to
the liberated.
In his commentary on the Matanga KP 8.10c-12b he adds
that it is the Kalamukhas who hold the view that the equality is
produced ex nihilo.
181
MatVr KP p. 162, ll. 710:
purvasiddhapasutvanivrttya ya pumsah sivasamanasya dharmasya-
bhivyaktih sa muktir ucyate. na tv asata eva tasyotpattir yatha kalamukhah
prahur asadutpatteh prak satkaryasiddhivicarena nirastatvat. napi sata
eva tasya sivat tatra samkrantir yatha pasupatah prahuh. prak
satkaryasiddhivicarena sato pi vinasasya nisiddhatvat.
3 kalamukhah em. (following the Kashmirian Mss) : kalamukhah cett.
BHATT
By liberation we mean the manifestation of the qualities of the soul [as it
really is, that is to say, ] equal to Siva, through the cessation of the
bondage that we have established above. It is not the coming into
existence of what was [previously] non-existent, as the Kalamukhas claim,
because our analysis of the proof that effects [pre-exist in the cause]
refutes the possibility that something non-existent could come to
exist. Nor is it the extension to the [soul] of [qualities] existing [in
Siva], as the Pasupatas hold, for the same analysis has refuted the
possibility that the qualities that already exist [in the bound soul] could
cease to exist [when this extension occurs].
Sivagrayogndra Janasivacarya, then, is simply reproducing
traditional doctrine when he gives this scheme. Since he had
certainly studied the works of Bhatta Ramakantha - not to have
done so is inconceivable in a Saiva of his background - the fact he
attributes the production ex nihilo theory to the Mahavratins, a term
evidently a variant on Mahavrata, is evidence that the terms
Mahavrata and Kalamukha were considered to be synonyms.
There is also the evidence of the Saiddhantika scripture
Brhatklottara. It twice pairs Pacartha and Kalavaktra in
contexts in which it must mean to refer to the two branches of the
Atimarga. The first term, as we have seen, is a name of the
Psupatastra and evidently refers here to a follower of the
earlier system based on this text seen in the Pacrthabhsya
and the Ganakrik.
f. 129r5-v1 (Pryascittapatala II):
kaulakapalikanam tu samparkkam parivarjjayet
pacarthakalavaktranam vivadam naiva karayet
1a kaula em. : kauli Cod.
182
He should avoid contact with Kaula and Kapalikas. He should not dispute
with Pacarthikas or Kalavaktras.
f. 187r3-4 (Lingoddhrapatala, 1):
limgoddharam pravaksyami naisthikanam tu sanmukha
pamcarthakalavaktranam sarvapasandikesu ca
Skanda, I shall teach [you] how to remove the [previous] religious
obligations of ascetics, Pacarthikas, Kalavaktras, indeed of any
religious outsiders [who seeks initiation in the Siddhanta].
22
b. The synonymity of the terms Klamukha and Lkula
We can now see the significance of the references to the
Lakula scriptures or system (lkulasiddhntah, lkulasamayah,
lkulgamah, lakulsvarasiddhntah and lkulsvarasiddhntah)
that abound in the south Indian inscriptions that speak of the
Kalamukhas, as when the ascetic Rudrabharana is described
as an ornament of the Lakulasamaya (Lorenzen 1972: 102), the
Kodiyamatha as a place where commentaries are composed on the
Lakulasiddhanta (ibid. 104). Lorenzen, who was unaware of the
existence of the Lakulas as distinct from the Pasupatas,
assumed that these expressions referred to the basic Pasupata
tradition then known. The earliest Kalamukha inscription, whose
date falls in AD 810, is reported by Lorenzen (ibid. 161) as speaking
of Brahmanas versed in ...... the eighteen Pramanas and
Siddhantas. I have not been able to consult the inscription
itself, and so I cannot yet test my suspicion that this is a translation
of an ambiguous compound -astdasapramnasiddhnta-, which
I should be inclined to translate the eighteen Pramana scriptures
or the Siddhanta of the 18 Pramanas. The use of the term
Siddhanta as a name of the Lakula system is attested in the
Sarvcrahrdaya, a lost Kaula work quoted by Abhinavagupta
on Tantrloka 13.305. It speaks of the superiority of its esoteric
tradition over the other Saiva systems in the following terms:
saivavaimalasiddhanta arhatah karukas ca ye
sarve te pasavo jeya bhairave matrmandale
The Saivas, the Vaimalas and the Siddhanta-followers, the
rhatas (Mausulas) and the Karukas - all of them are unliberated
in relation to the cult of the circle of goddesses taught in the
Tantras of Bhairava.
Since the Saivas here can only be the followers of the texts
usually called Siddhantas in the gamic literature, the Siddhanta
183
of this verse must be something else. The list will be satisfactorily
complete only if that is the Pramana school.
Klamukhas and the Kaplavrata
Lorenzen observed references to the Kalamukhas in the
inscriptions which describe them as Mahavratins and Mahapasupatas.
But he could not accept the implication of these terms, namely that
the Kalamukhas were Pasupatas who had adopted the Great
Observance (mahvratam) otherwise known as the Kapalavrata.
This was because he associated this practice with the grim rites of
the Kapalikas to which he devoted the other part of his study; and
because he assumed that such a practice could not be shared by the
Kalamukhas. He assumed this because he could see they were a
kind of Pasupata from the references to the Lakulagama and to
Lakulsvara in the inscriptions, and because the Pasupata tradition
accessible to him in commonly known published works was that
of the Psupatastras discussed in the last lecture. The Pasupata
ascetic, as we saw, ends his life in a cremation ground; but he certainly
does not take up the skull and khatvnga of the brahmin-killer.
There was, in fact, one piece of evidence of which he was
aware, apart from the epithets Mahavratin and Mahapasupata,
which contradicted this picture of things. That was the descrip-
tion of the Kalamukhas that is given by Yamunacarya in his
Agamaprmnya. It is as follows (p. 47):
evam kalamukha api samastasastrapratisiddhakapalapatrabhojana-
savabhasmasnanat at pr asanal agudadhar anasur akumbha-
sthapanatatsthadevatarcanader eva drstadrstabhstasiddhim
abhidadhanah srutibahiskrta eva.
The Kalamukhas too are outside the Veda; [for] they claim
to be able to obtain miraculously all that they desire whether visible
or invisible simply by eating from a bowl fashioned from a human
skull, bathing in the ashes of the dead, eating them [mixed with
their food ?], carrying a club, installing a pot containing alcoholic
liquor and worshipping their deity in it, practices which all the
Sastras condemn.
However Lorenzen will not let this report override his
assumption of the Kalamukhas purity. Observing that eating
from a skull-bowl and worshipping the deity in a pot of liquor are
184
practices especially associated with the Kapalikas he suggests
the following explanation for what he takes to be the accounts
confusion:
23
One might suggest a more sinister explanation. At the time of Yamuna
and Ramanuja the Kalamukhas were rapidly gaining popular and
even royal support in South India. The two Vaisnava priests may
have purposely confused the two Saivite sects in order to discredit their
more important rivals.
The Lkula system compared with the Pcrthika and
gamic Saivism
Initiatian
a. Pcrthika
Nowhere in our Pacarthika texts is Dksa, initiation, treated
in any detail. The author of the commentary of the Ganakrik
refers the reader for such information to the Samskrakrik, a
work now apparently lost. All that we can learn from our texts
is the following. The brahmin candidate is first examined to
determine his fitness for initiation.
24
Since it is said to be Rudra
himself that impels him to seek this rite of entry, it is likely that
the examination of the candidate stressed in our texts was to
identify signs of this divine intention. In this respect, then,
Pasupata Saivism would be comparable to gamic Saivism.
For the gamic systems hold that the officiant is to carry out an
initiation only if he can recognize in the supplicant signs of the
descent of [Sivas] power (saktiptah).
If accepted, the candidate fasts in preparation for the
ceremony.
25
According to Kaundinya the ceremony itself consists
of three parts. (i) First the candidate is placed facing the right face
of an idol of Mahadeva (daksinasym mrtau) and empowered
with ash empowered by means of the five Brahmamantras.
(ii) Then the marks of his original socio-religious identity
(utpattilingam) are removed; (iii) and finally, he is given the
Brahmamantras, i.e. the five Brahmamantras of the cult
(mantrasrvanam).
PaBha p.8, ll.7-9:
mahadevasya daksinasyam murtau sadyojatadisamskrtena
bhasmana samskaroti utpattilingavyavrttim krtva mantrasravanam
ca karoti ......
185
The commentary on the Ganakrik takes us a little further.
One of the eight sets (ganh) of five taught in the Ganakrik is
that of the Requisites of Initiation (dkskrni). These are
(i) substance (dravyam), (ii) time (klah), (iii) ritual action (kriy),
(iv) the idol (mrtih), and (iv) the guru. According to the commentary
the first, substance, comprises knowledge in the candidate and the
guru (vidy), the materials of ritual (kal), and a bound soul (pasuh),
i.e. the brahmin candidate himself. The time is the forenoon.
The ritual is that of the empowerment (samskrakarma) of the
idol of the Deity (kranamrtih) and the candidate (sisyah), for the
details of which we are referred to the Samskrakrik. The
idol, we are told, means not the idol proper, in which Mahadeva
receives his regular worship, but an area of ground on its right
side concealed from view by some such means as the building
of a hut (kut). The guru is the officiant (cryah), but also
Mahesvara, since it is he that empowers his action (tadadhistht).
26
As for the materials of worship (kal), he quotes the following verse
on the matter:
GanKar p.8, ll.24-27:
kala darbhadya dksangam. uktam hi - purvam darbhah punar bhasma
candanam sutram eva ca / puspani ca punar dhupam mantra esa kramah
smrtah
The materials as an auxiliary of initiation are those beginning
with blades of darbha grass. First darbha blades, then ashes,
sandal-paste, the thread (stram), flowers, incense and the mantras.
This is the order according to tradition.
From this slight and scholastically ordered information
we can at least perceive or hypothesize certain continuities with
the gamic practice and theory of Dksa. In the first place the
ritual of the candidates empowerment (sisyasamskrakarma)
corresponds closely enough to the first phase of gamic Dksa
to justify the hypothesis that the latter has developed from the
Pasupata by the process of extension and elaboration found
elsewhere in the relations between the two systems. The use of
blades of darbha grass and the empowerment with ashes are
prescribed as the first two elements of the Pasupata initiation.
The first rites involving the candidate in an gamic Saiva initiation
are likewise (i) the officiants tapping him on the head with
186
ashes, an action which is framed or preceded by aspersion with
water, and (ii) the officiants touching him above and below
the navel with blades of darbha grass. These actions are merely
preliminaries to preliminaries in the final structure of gamic
initiation; but their original importance is probably preserved in
the fact that the function of these touchings (labhanam) with
darbha grass is said to be to loosen the bonds that bind the soul
to samsra. This explicit meaning is implicitly confirmed by
the next action of the gamic Dksa, which is to install the mantras
of the deity in the candidate and then worship him as the deity.
For the sequence then exemplifies a model basic to all gamic
ritual: an action negating individuality or bondage is followed
by the construction of a divine identity in its place.
How are we to interpret the remaining items on the list, the
sandal-paste, the thread, the flowers, the incense and the mantras?
The best way forward, I suggest, is to assume that this list
beginning with darbha grass and ending with the mantras is
coterminous with the three phases into which Kaundinya
has divided the Dksa, namely (i) empowering with ashes,
(ii) removing original marks of identity (utpattilingavyvrttih)
and (iii) communicating the mantras (mantrasrvanam). This
is probable in principle, given the very close relation between
Kaundinyas bhsya and the system of the commentator on
the Ganakrik; and it is further suggested by the fact that both
the list of three and the list of eight end with the mantras. Which
of the eight then corresponds to the removal of the utpattilinga ?
It seems likely that it is the thread, item five. For in the gamic
Dksa the candidate receives a sacred thread (yajastram)
marking his enrolment as a true brahmin after the symbolic
removal of the caste of his birth (jtyuddhrah). The remaining
items in the list of eight are probably the elements of a pj
offered by the officiant to the initiate.
Fig 2. The two Pcrthika Dky lists correlated
The eight material require- The three phases of the
ments for Dksa according Dksa according to
to Ratnatk p. 8,26-27 PaBh p. 8, 4-9
blades of darbha grass
ashes empowerment with ashes
187
sandal-paste
the thread removal of marks of pre-initiatory
identity
flowers
incense
the mantras communicating the mantras
In gamic Dksa too the imparting of the mantras is naturally
enough the final phase of the ritual.
The commentator on the Ganakrik refers to the empowerment
both of an enclosed area to the south of an idol (temple) of the deity
and of the candidate. This too is parallel to the gamic case. There
one empowers the candidate only after first installing and
worshipping the deity of the cult on the sacrificial ground
(sthandilam) within a pavilion (mandapah).
Also thoroughly in keeping with the doctrine of gamic Saivism
is the commentators explanation of the guru as both the human
officiant and Mahadeva, since the latter empowers him. It is a
fundamental tenet of the gamic Saivas of all traditions that Siva
is the true agent of the ritual. The initiating officiant is to see
himself as nothing more than the medium through which the deity
is acting.
However, while there is a certain continuity between the Pasupata
and gamic Saiva Dksas, there is perhaps a major difference. It is
this, that the former may have been seen as no more than a rite of
qualification, a samskrah in the precise meaning of the term, namely
an action which qualifies a person or thing for some further action.
For the gamic Saivas Dksa is a samskrah in this sense but only
for some, namely for those who are required to pursue Saiva practice
after it. For others, considered incompetent to take on such duties, it
is not a samskrah. What it is for all is a ritual which guarantees
liberation by materially destroying the soul s bonds and by bestowing
knowledge in the form of a latent realization of inherent equality
with or identity with the deity (sivatvbhivyaktih), a dual function
asserted throughout the gamic scriptures through a creative
etymology (nirvacanam) of the word Dksa out of d to give and
ksi to be destroyed. No such claim is made for the Pacarthika
Dksa by Kaundinya or the author of the Ratnatk. There we
find only the claim that it empowers one to move into the first
188
stage of post-initiatory practice (the first avasth), just as the first
produces certain capacities which enable one to proceed to
the second, and so on.
GanKar p.8, ll.17-19:
anadhikarino vasthapraptir ayukta. tatra yatha dvitya-
dyavasthapraptau janakalusatvadayo dhikaritvapadakas tatha
prathamavasthapraptau ko hetur ity ucyate. atrapi karanavisesotpadita
dksa hetuh.
1 tatra yatha em : tatra tatha Ed
b. Lkula Initiation
The Nisvsamukha speaks twice of the Lakulas initiation.
On the first occasion, at the beginning of the account of the
observance, the text states simply that the practice of the
observance is conditional on initiation and that Dksa is preceded
by or begun by the candidates being touched with the five
Brahmamantras:
f. 17v2 (4.88ab):
alabdhaf pacabhir guhyair ddksitas caiva so bhramet
When he has been touched by the five Brahmamantras and received
initiation he should wander [as a mendicant].
This statement is followed by the description of the Vrata and
by the summary account of the worldlevels knowledge of
which is the means to liberation. Then follow two more verses
telling us what the initiation that comes after the touching with
the Mantras comprises.
f. 17v4-5 (4.96-97):
96 krdarthasiddhaye caiva prakriyadhyanam asrita[h]
+ + vai prakriyadhvanam athasabdena dksayet
97 athasabdanipatena dksitas capasur bhavet
kriyavams ca duracaro mucyate natra samsayah
The officiant, we are told here, must meditate on the cosmic
hierarchy (prakriy), make that hierarchy
27
the object of an action
unfortunately lost in a lacuna caused by damage to the manuscript,
and then initiate the candidate through what the text calls the
descent of the word atha (athasabdaniptah). As a result of its
descent the initiand will cease to be a Pasu (apasur bhavet), he will
be liberated (vimucyate).
189
At first sight this might appear to be contradictory. For
we are told that it is the Dksa that is the cause of liberation. Yet
between the two statements on initiation we have been told that
one is liberated through gnosis of the eleven world-levels from
the Avci hell to Dhruva (avcydi dhruvnta ca etaj jtv
vimucyate [f.17v4 (4.95cd)]. The gamic Saivas would face the
same objection; and they answered it by saying that though the
purpose of Dksa was to sever the souls bonds (psacchedah) of
Impurity (malah), Materiality (my) and Karma and to make
manifest its latent Sivahood (sivatvbhivyaktis ca) it could not do so
completely and immediately without ending the life of the initiand.
It therefore leaves intact some part of these bonds; and it is the
function of the daily practice of the Saiva religion between
initiation and death gradually to eliminate this residue so that
liberation will occur when the initiate dies. This is how Bhatta
Ramakantha expresses the matter in his commentary on the Matanga
ad VP 26.63d:
28
kevalam asadyonirvanadksayarabdhakaryanam karmanam
anupaksayat tadbhogoparodhena malamayyayor api pasayoh
sarvathanapadhvamsitatvena sivatvasyanabhivyaktatvad dksottara-
kalam janadibhis tathapacayakramena pratyaham tesam pasanam
vicchittih sivatvasya cabhivyaktih kriyate yatharabdhakaryanam
karmanam na bhogavirodho jayate. napi ca pasanam punahprarohah
vyaparantarena va nivrttih sivatvasya vyaparantarena vabhivyaktih.
iti bhavaty eva catustayad dksajanader yathasvam vyaparabhedan
muktih. iti suktam mokso vatha catustayad iti.
If Sivahood is not made manifest through initiation alone,
that is because unless it is the form of initiation that bestows immediate
release [in death] (asadyonirvnadksay), it does not destroy those past
actions whose effects are already active, and so also desists from
eliminating entirely the bonds of Impurity and Maya [, leaving
something of them in place] in order to allow the experience of
the fruition [of those actions]. So (anabhivyaktatvt) it falls to
knowledge [, meditation, the discipline] and [regular worship]
after initiation to remove [the remnant of] those bonds by gradually
diminishing them, day by day, and bring about the revelation of
[the souls] Sivahood, so that the experience [of the fruition]
of those past actions whose effects are already active is not
impeded. Once destroyed the bonds cannot arise again. No other
operation [but that of initiation completed by this postinitiatory
discipline] can destroy them; and no other operation [than this]
can reveal [ones] Sivahood. So it is that liberation comes about
190
from the four factors: initiation, knowledge [, meditation] and [the
discipline], as the result of the specific functions of each. Thus our text
has rightly declared: Liberation is the result of the four, or [enjoyment].
Much the same was intended, I suggest, by the Lakulas, that is
to say, that gnosis of the prakriydhv completes by the
time of death the task that initiation had almost but not
completely accomplished. And there is evidence for this interpretation
not merely in the fact that it is the most obvious
way of accounting for the concurrence of the two statements of
the cause of liberation but also in the Nisvsaguhya. For that
describes the followers of the Pramanagama as purified by
initiation and gnosis:
f. 68r1-2 (7.261-262b):
261 tejsas ca dhruvas caiva pramanadhvana krttitam
kapalavratam asthaya pramanagamasiddhaye
262 gata dhruvapadam ye tu dksajanavisodhitah
29
261d pramanagamasiddhaye conj. : pramanagamasi[ + + ] AB
Evidently, then, the Lakula initiation departs from the
Pacarthika both in procedure and in purpose and in both
these respects stands with the gamic Saiva tradition over and
against the Pacarthika. The rite, which the Nisvsamukhas
account of the Lakula initiation covers even more elliptically
than Kaundinya and the commentator on the Ganakrik in the
phrase labdhah pacabhir guhyaih (f.17v2 [4.88a]) 'touched by
the five Secret [Mantra]s corresponds, I suggest, to the preliminary
initiation of the gamic system known as the samayadks, and
the Lakulas initiation proper, involving as it does the cosmic
hierarchy, corresponds to the full initiation, the nirvnadks, in
which it is precisely this adhvasuddhih the elimination of the
hold of that hierarchy on the initiands soul that is the essential
process. A memory of the connection of the samayadks with the
Pasupata tradition from which the Saiva evolved by extension via
the Lakula is preserved, I believe, in the gamic doctrine of the
destiny of those who receive only the samayadks in their
lifetimes. They are said to achieve at death the lesser liberation
(apar muktih), which is to enter the level of svara between
Suddhavidya and Sadasiva in the Pure Universe.
30
This is, as we
have seen, the point of liberation assigned by the Nisvsa and the
Svacchanda to the Pasupatas.
31
Elsewhere they are said to become
191
Rudresas
32
or to attain the rudrapadam.
33
And the rites which
supplement the samayadks in certain gamas
34
have as their goal
the transformation of the initiate into a Rudramsa.
The Lakula system added to this proto-dks an early version of
the nirvnadks and in so doing departed from the Pacarthikas
not only in the structure of the rite but also in its purpose. For
whereas the Pacarthika rite was essentially a rite of passage,
here it has become, as in gamic Saivism, an tmasamskrah,
a rite that bestows or prepares the soul for liberation.
As for the descent of the word atha that is specified by
the Nisvsamukha as the means of dks and thereby liberation
(athasabdena dksayet / athasabdaniptena dksitas cpasur
bhavet [f. 18r5 (4.96d-97ab)]), its nature is not immediately obvious.
I propose the following solution. In the first place atha is the
opening word of the Psupatastra, the root-text of the Atimarga,
which according to Kaundinya was taught by Rudra incarnate as a
brahmin to Kusika, the first disciple, when the latter had come to
him and asked whether there exists a means of attaining the
complete and definitive cessation of all suffering (duhkhntah). Atha,
the opening word of his answer, that is, the first word of the first
Sutra, is interpreted by Kaundinya to mean Yes: athasabdah
prstaprativacanrthah. asti sa duhkhnta iti (PaBha p. 4). This
affirmation that there is indeed a means of escape from suffering,
coming as it does at the beginning of the sacred teaching, has, I
suggest, been transferred to the parallel context of the ceremony of
initiation. But for that transference to be meaningful the word atha
would have had to be endowed with new and unprecedented
meaning. It cannot be the word itself that descends and liberates but
some esoterically constructed referent of the term. I propose that
for this purpose the word atha was understood as a symbol of
Sivas Power (sivasaktih). The expression athasabdaniptena
would, then, be synonymous with sivasaktiniptena by the
descent of Sivas Power. We find just this expression in the same
context in the same corpus, in Nisvsanaya: sivasaktiniptena dks
jnam prayacchati (f. 31v2 [1.88cd]) Dksa bestows knowledge
through the descent of Sivas Power. That atha should have been
understood in this way may seem far-fetched. But it has more
support than this parallel and the gamic Saiva doctrine it
192
exemplies, namely, that the descent of this power (sivasaktiniptah,
saktiptah) is the necessary precondition of initiation and thence of
liberation. For in his comments on atha when it occurs at the
beginning of the Saiva scripture Partrimsik Abhinavagupta quotes
a verse that interprets the word in just this sense:
akarah siva ity uktas thakarah saktir ucyate
The sound a means Siva and the sound tha means Sakti.
Abhinavagupta tells us that his predecessor Somananda (fl. c. 900-
950) had quoted this verse with approval in his own (lost) commentary
on the Partrimsik. He adds that although he too has reproduced it
here he has not presumed to analyze it at length
because he has no independent knowledge of the scripture from
which Somananda drew it:
Partrimsikvivarana, p, 99, ll. 8-9 and p. 100, ll. 3-4
yat tu srsomanandapadah akarah siva ity uktas thakarah saktir ucyate
ityagamapradarsanena atha ity etavad evanuttaram iti vyacaksire
... tan nasmabhir vitatya vivecitam tadrsasyagamasya yato na saksad
vayam abhijah
Perhaps this unfamiliar source was indeed a work of the Lakula
Atimarga. Given the deep and comprehensive knowledge of the
gamic Saiva corpus that is evident in Abhinavaguptas works, it
would be would be most surprising if this source were gamic.
Lakula initiation, then, differs from Pacarthika by adding
the purification of the cosmic hierarchy, so transforming it into a
means of liberation; and in this most fundamental respect it is one
with the gamic Dksa. Details apart, about which we know almost
nothing in the Lakula case, it differs from the gamic in only two
respects. The terminus of its cosmic hierarchy is lower than that
of the gamic systems, because those have extended their own
beyond it in their bid for supremacy within the greater religion;
and they appear not to have developed the hautr dks, the Dksa
through the placing of offerings in a consecrated fire, that is the
principal formal characteristic of the gamic rite. In this respect
the Pacarthika and Lakula systems stand together against the
gamic. Just as they are systems for ascetics outside society so
193
they are without fire. And just as the gamic systems incorporate
the householder so they have their initiates install fire in order
that they may gratify their Mantra deities therein.
Doctrine
In the field of ontological and soteriological doctrine
we find both similarities and profound differences between
the Lakulas and the other two divisions of the Saivas that concern
us. They differ from the Pacarthikas and agree with the gamic
Saivas in conceiving of the universe as a hierarchy of worlds
divided into two parts, a lower, impure universe (asuddho mrgah)
and a higher, pure universe (suddho mrgah), the two separated
by a vast barrier termed Granthi which only initiation and the
cultivation of gnosis during postinitiatory observance can enable
one to go beyond. For the Pacarthikas the only ontological
hierarchy is that of the twenty-five principles (tattvni) inherited from
the Samkhyas with Rudra above the individual soul as
twenty-sixth. There are only two major differences between
the Lakula and gamic conceptions of the world-hierarchy.
The first is that in the process of formulating their superiority
over the Lakulas the Saivas added further world-levels beyond
the summit of the Lakulas pure universe and moved the position of
the barrier between the impure and the pure upwards so that certain
worlds which the Lakulas had located above it in the
pure now fell below it in the impure. The second bears on the
means of penetrating the barrier. For the Lakulas this is
initiation followed by asceticism and gnosis whereas for the
Saivas it is initiation followed by a much more accommodating range
of religious practice, from the devotions of the exonerated initiate,
which differ in no way from those of the lay Sivabhakta, to
specifically Saiva practice in which Yoga, Gnosis or ritual may
predominate. In the case of the Saiva Yogin or Janin, in other
words, there is little difference in the means of liberation after
initiation. For him as for all Lakulas the path is one of visionary
ascent through a gnosis-driven Yoga up a ladder of principles or
worlds. This tradition, which was particularly associated with
ascetic initiates, preserves the general character of the Lakula
past. But, of course, it is possible that for the great majority of gamic
194
initiates the essence of postinitiatory observance was not Yoga and
asceticism but the practice of regular and incidental rituals in which
Yoga and Gnosis were subordinate even in the ideal and in common
reality were thoroughly ritualized.
In their theory of the nature of liberation, however, the Lakulas
occupied a position that set them apart radically from
both the Pacarthikas and the Saivas, so that in this respect at
least the Pacarthikas and the Saivas agreed against them. For
the Pacarthikas liberation came about through the passing
into the soul of Rudras properties of omniscience and omni-
potence, his vibhusaktih and his prabhusaktih. The Siddha,
therefore, is an agent both in bondage and release. The Saivas
did not agree that omniscience and omnipotence were transmitted
to the soul but they held that both knowledge and action were
natural to the soul and that the only difference between
their operation in the states of bondage and release is that in
the former it is hindered by the force of Impurity (malah) and that
in the latter through the removal of that force through initiation
and observance they shine forth unfettered and absolute. The
liberated then are equal to Siva (sivasamnah), both omniscient
and omnipotent. The Lakulas, however, held, like the Samkhyas,
that the soul does not act, that liberation, therefore, is the
attainment of omniscience without omnipotence, and that,
consequently, all souls that act are bound (pasavah). The Lakula
category of the Pasu, then, includes not only the uninitiated, but
also the initiated while they yet live and, what is more striking,
the various Rudra-souls that govern at the various levels of cosmic
hierarchy (prakriydhv). This second Lakula doctrine that the
Rudras are not Siddha and they and all men until they die are
Pasus is reported by Abhinavagupta in his svarapratyabhijvivrti-
vimarsin:
Vol. 2, p. 405, ll.18-25:
siddhah samayyadigurvantah suddhatmano niskalah vidyesvara
anantabhattarakadya mantramahesvara mantresvara angustha-
matraprabhrtayah. adigrahanan mantrah... na tu lakuladidrsva
pasurupah. tadartham eva siddhasabda esam visesanam.
Siddhas the perfected are (1) all initiates from Samayins to Gurus
(2) the Pure, those who are free of the products of Maya,
195
(3) Anantabhattaraka and the other Vidyesvaras, (4) Angusthamatra
and the other Mantresvaras, and (5), indicated by the word etc., the
Mantra-souls. They are not Pasus as they are held to be in the Lakula
and related doctrines. It is precisely to make this point that
Utpala has used the word Siddha to describe them.
The first Lakula doctrine, that the soul in liberation is inactive
and so merely omniscient, is attested by Bhatta Ramakantha
in his commentary on the Paramoksanirsakrik of Sadyojyotis:
Paramoksanirsakrikvrtti, pp. 6, l.17-p.7, l.13:
With the words omniscient [but] inactive [3b] [Sadyojyotis refers
to] the [final] fruit of another doctrine. [That doctrine is as follows.]
That Paramesvara is the agent of acts is established by the fact that
otherwise such effects as bodies, faculties, and the worlds would be
impossible. This being established, the explanation of the existence
of such effects does not require us to posit agency in those who are
liberated. The Saiddhantika may hold that his agency is established
by the fact that he is liberated and that since God is liberated the
liberated person must like him be an agent. But this is incorrect,
for it is not proved that God is liberated; and so the inference
lacks a valid example. Only one who has been in bondage can be
liberated and in this respect there is a complete difference
between God and the Siddha. As the Holy Avadhuta has
said in his Vysksin:
Though Paramesvara no longer exercises authority over
the Siddha he continues to do so over others. The Siddha, however,
ceases altogether to exercise authority of any kind. This is the
difference.
In fact those [who follow this school] deny that the individual
possesses agency even before [liberation]. That for them, as for
the Samkhyas, belongs exclusively to Prakrti. So because there
is no evidence of its being an agent it is inactive in the state of
liberation. Consciousness, on the other hand, is innate in individuals.
Experience attests its existence prior [to liberation]; but in the
state of liberation itself all the factors that limited it are absent and
so there is nothing which is not its object. Thus those versed in
these scriptures maintain that the liberated individual is omniscient
[but] inactive. It is these [authorities] that the author of [Raurava]
stra [on which the Paramoksanirsakrik is a commentary]
indicates here by the expression pramna[...]kartrtvam
(pramngneyakartrtvavisikhmalakrakh). For they are the
authors of such texts as the Hrdayapramna.
196
Bhatta Ramakantha leaves it to us to infer that the holders of
this doctrine of release are the Lakulas from the statement
that they are the authors of the Hrdaya and other Pramanas. But
the same doctrine is attacked by Sadyojyotis in his Moksakrik
(118-); and in this case Bhatta Ramakantha is more explicit in
his commentary, identifying the holders of the view as the
Mahavratas, whom we have seen to be the Lakulas under
another name:
Now, to analyse this [liberation] he rejects the liberation
[theories] of others. Among these the Mahavrata [argues
that]:
118 IF SIDDHAS WERE EQUAL [IN ALL RESPECTS] TO SIVA
THERE WOULD BE A PLURALITY OF [FIRST] CAUSES.
BECAUSE THEY AND SIVA WOULD HAVE CONFLICTING IDEAS
THE EFFECT (I.E. THE UNIVERSE) COULD NOT ARISE.
If the liberated were omnipotent like god they would
disagree as to the order of creation and the result would be that nothing
would be produced. So only omniscience belongs to
the liberated and omnipotence belongs only to Paramesvara.
such is the inevitable conclusion. This [says Sadyojyotis] is not a
defect [in our doctrine. For]
119 MEN MAY DISAGREE, FOR THEY ARE SUBJECT TO
CRAVING AND HATRED, AND SINCE THEY DISAGREE IT IS
ONLY TO BE EXPECTED THAT THEY SHOULD HAVE
CONFLICTING DESIRES.
It is natural that the bound should have different intentions
because being subject to craving and the like they would take
up opposing positions. But
120-121 SIDDHAS BEING FREE OF SUCH DEFECTS AS CRAVING
AND BEING OMNISCIENT DO NOT HAVE CONFLICTING
DESIRES. OUR AUTHORITIES DECLARE THAT THEIR SPHERES
OF POWER ARE NOT SEPARATE FROM SIVAS. SO THERE
IS NOTHING IN THIS [VIEW] WHICH WOULD CONTRADICT
THE PRODUCTION OF THE VARIOUS EFFECTS WHICH
CONSTITUTE OUR UNIVERSE.
Since they, being perfect, are free of craving and the rest, and since
they are omniscient, their authority and Sivas do not
clash. Their intentions are one, because they lack partiality. So
there can be no question of the defect of the non-origination of
197
effects if our view is accepted. If on the other hand the view of
the Mahavratas is accepted there could be no liberation. For
122 IT IS NOT TRUE THAT THE LIBERATED COULD BE
OMNISCIENT. FOR HE IS NOT OMNIPOTENT [IN YOUR VIEW]
AND THEREFORE INFERIOR. NOR COULD HE BE OMNISCIENT,
BECAUSE [FOR YOU] HE IS AS INACTIVE AS THE BOUND.
For liberation is absolute fullness. So it would follow that
[your liberated person] is not omniscient. Moreover, since he
lacks omniscience he would be no different from the bound.
The doctrine that the liberated soul is inactive indicates an
intensely renunciationist perspective such as is found nowhere
else in Saivism. In all other Saiva systems the liberated soul
takes on the qualities of God; and God, though beyond the
universe and encompassing it with his omniscience, must also
have the power to activate it in order to enable the souls under
his control to be incarnated, to act, and to experience the results
of their actions. But here the ultimate state is without this
essential quality of Godhood. Either God was considered to be
of a profoundly different nature from souls in that while they
have only the power of knowledge (jnasaktih) he has both
that and the power of action (kriysaktih); or they looked upon
God as a soul who had taken on a limiting office, the most
exalted, least limiting office there is, but one that its holder will
eventually renounce to a successor in order to enter the final
state of perfect transcendence. In favour of the first view we
have the testimony of the Sivaratnkara of Basvaraja, king of Keladi
(9.6.92c-95c):
mahavrate tv anavaditrimalanam sarrinam
93 janasaktir ivaikaiva na kriyasaktisambhavah
dksayukta vidhanena sarvadaivasthidharinah
94 sastroktacaryanirata muktim asnuvate khilah
te muktah kevalajanasaktimanto bhavanti hi
95 janakriyakhyasakt tu vartete paramesvare
evam mahavrate rtih.
93d sarvadaiva conj : sarvathaiva Ed
In the Mahavrata, on the other hand, incarnate souls are
subject to the three Impurities: nava [, Mayya] and [Karma] but
possess the power of knowledge alone. There is no possibility
198
of their also having the power of action. When they have been
initiated according to the prescribed procedure all practice the
ascetic observance ordained in their scriptures, wearing [ornaments
of human] bone at all times, and so attain liberation. When liberated
they [still] possess only the power of knowledge. God, on the
other hand, has both the power of knowledge and that of action.
Such is the doctrine of the Mahavrata.
Perhaps we should accept this testimony without further
thought. But there is something unrounded and unsatisfying
about a doctrine which does not merely differentiate between
God and other souls but sets God outside the definition of the
soul as a being of an entirely different nature. It is, at least, unusual
enough in the context of Saiva, Vaisnava and Vaidika doctrine,
to allow one to speculate that the Sivatattvaratnkaras account may
be inaccurate in this one particular and that the Lakulas saw
the liberated as superior even to that Being whose office was
to govern the universe as the highest among the Rudra-souls
that are manifest at its various levels. In favour of this
hypothesis I would urge that it adds to the conclusion emerging
from other details that gamic Saivism is a development from
a Lakula base. For if the Saiva doctrine of liberation is a revision
of such a Lakula position, this will enable us to explain rather
neatly what has always seemed to me a suspiciously awkward
structural feature of the gamic picture, namely that all the
other deities that govern the levels of the Saiva universe are
office-holders (adhikrinah) who move on when their drive
for transcendence through Sivas grace awakens distaste for
the level they occupy, whereas God himself at the top of the
hierarchy remains forever unsucceeded. What then happens to
souls that have ascended to the point at which they are ready
for Godhood? They cannot bypass the now permanent holder
of that position in order to go beyond him to their goal. Instead
they fall in beside him experiencing the manifestation in their
souls of all the qualities of Godhood, including the power to
create and destroy the universe, to bind and free souls. Thus
where in the Lakula system there were the holder of the highest
office and liberated souls inactive beyond him, now the devout
were asked to accept two kinds of Sivas, a Siva who has
always been and always will be [a] Siva (Anadisiva) and others who
are Sivas now that they are released (Muktasivas). So far
so good. But if, as the Siddhanta maintains, these released
199
Sivas have exactly the same powers as the Siva in office,
either they will exercise them, in which case, since there are
no other universes but this, they will get in each others way,
or they will not exercise them, in which case their omnipotence
is redundant. The Siddhanta, forced into this awkward position
by its decision to pull souls back to the level of God and to make
his office permanent, can only answer lamely, as we have seen,
that the released are free of pettiness and so do not wish to
create difficulties by putting their powers into action.
Continuity between the Atimrga and the Mantramrga
We have seen evidence of structural continuity between
the Lakula division of the Atimarga and the gamic Saivism of
the Mantramarga. The bhuvandhv of the Nisvsa and
Svacchanda incorporated and extended that of Lakulas,
preserving the division of the universe into two segments, pure
and impure, and simply shifting the boundary between the two
so as to concede no more than a lower liberation to the Lakulas
and Vaimalas. But even this strategy was inherited from
Atimargas side of the divide. For in going beyond the Lakulas
the Saivas were simply repeating the procedure by means of
which the Lakulas had asserted their superiority over the
Vaimalas and the Vaimalas or the Lakulas themselves their superiority
over the Mausulas and Karukas. In the first case my conclusion rests
only on the relative positions of the Vaimalas
and Lakulas ultimate worlds. But in the case of the relation
between the higher systems of the Atimarga and the lower known as
the Mausula and Karuka it is confirmed by what we might
call archaeological evidence visible in the ordering of the strata
that make up the Lakulas hierarchy of worlds. The worlds that
are reported by the Svacchanda to have been the goals of the
Mausulas and Karukas are those of the Rudras Ksemsa
and Brahmanahsvam (SvT 11.71c-72b). In the Lakula hierarchy
reported by the Nisvsamukha these are located among the five
Abhibhavarudras, and the latter are placed immediately above
the world of the Rudra Gopati. Now, our text refers to that world as
Granthi, a term that we know denoted the barrier between the
impure and pure universes in both Lakula and gamic systems.
200
It is evident, then, that the line between the pure and impure universe
lay here in the earlier systems. In the Lakula system,
and probably in the Vaimala, though we have no independent
evidence of that, a number of levels have been added above the
Abhibhavas, and the boundary between the two segments has
been pushed further up so that the goals of the Mausulas and
Karukas are now within the universe of the unliberated.
Fig. 3
The Pure Universe of the Lkulas (saddhdhr):
Dhruva
Tejsa
Eight Murtis [Vidyesvaras ]
Eight Vidyas
Eight Pramanas [The Lakula scriptures]
Bhasmsa
Dhyana
Damanesvara
Dhatr
Pranava
Yoni, Vagsvar
The Impure Universe of the Lkulas (asaddhdhr):
Rsikula
8 Vigrahas
Pasas
Ananta
5 Abhibhavas ( Kyemsa, Brahmaodhipati etc.)
Granthi
Gopati
3 Mahadevas
5 Sisyas and 5 caryas
Sivasankara, Asadhya, Harirudra, Dasesa
201
Vigrahesvara
Gahana
.........
We have already seen the grounds on which Ksemaraja justifies
this discrimination between the two leagues of Pasupatas, between
those that do and those that do not cross the barrier to liberation.
35
There may, of course, have been stages of extension and
subordination between the Mausulas-and-Karukas and the Vaimalas-
and-Lakulas, and between the latter and the gamic systems of the
Nisvsamhit and other texts. I know of no hard evidence of this;
but it should perhaps be pointed out that the gamic accounts
contain two further Dhruvas at intervals above the Dhruva who is
the lord of the Atimargas highest world, suggesting the possibility
that the bhuvandhv went through further stages of extension
before the emergence of the Mantramarga system of the Nisvsa.
The relevant segment, as it is seen in the Nisvsa (Guhya and
Dksottara) and the Svacchanda is as follows:
36
Fig. 4
Nisvsa, Svacchanda Svacchanda
Sadasiva (with the promulgators = Susivavarana
of the 28 Siddhantas)
Rudronkara
Ghana = Samayavarana
Prabhava etc.
8 Rudras: Prta, etc. = Prabuddhavarana
Nirajana
Dhruva III = Dhruvavarana
11 Rudras: Bramodarka etc. = Moksavarana
Dhruva II = Sivavarana
Ekaksa, Pingala and Hamsa = Susuddhavarana
Brahma, Rudra, Pratoda, Ananta = Manavarana
Dhruva I
Tejsa
.........
The Saivas have conventionally divided the means of
liberation taught in the gamas, that is to say their subject matter,
202
into the four categories, ritual (kriy), doctrine or gnosis (jnam,
vidy), meditation (yogah), and ascetic observance and other rules
of governing the conduct of the various classes and kinds of
initiate (cary). Continuities between the Lakulas and the Saivas
have now been shown in the areas of the ritual of initiation and in
the doctrine of the path to liberation, though not of that of the nature
of the goal itself. Insufficient evidence exists to permit much of a
comparison in the domain of meditation. We have seen, however,
that when Yoga is not subordinate to ritual in gamic Saivism its
purpose is to enable a visionary ascent to Siva up the hierarchy of
worlds and principles and that in this respect it conforms to what we
know of the path to liberation to be followed after initiation by the
Lakulas. This leaves only the domain of ascetic observance
(vratacary). Of course, as with Yoga, the big difference here with
the move to gamic Saivism is in the limitation of the scope of this
method. What were once the only ways to liberation after initiation
now survive as full-time disciplines only among what were
probably relatively small elites within the community of Saiva
initiates. But here, and in the ascetic observances of limited
duration incumbent on Gurus before or after their consecration
to office, there is much that remains unchanged from the systems
of the Atimarga.
But in this case the inheritance is from both divisions of
the Atimarga, both the Pacarthika and the Lakula. A striking instance
of the preservation of Pacarthika vratacary in Saiddhantika
Saivism appears in the Rudravrata taught in the ninth Patala of the
Carypda of the Matanga:
1 atha rudravratam punyam idanm munisattama
varnayisyami samksepad yena lokas caracarah
2 vasam ayanty asandehac caratah sadhakatmanah
muktis ca sasvat vipra sivasastrakramagata
3 vyajyate vyaktatapasas caryapadavidhanatah
ragat paranmukhasyettham dvesac ca kalusatmakat
4 buddhir utpadyate kasmat karanecchapracodita
samay bhavitatma syad rudraradhanatatparah
5 anujatas ca guruna bhasmanistho pratigrahah
jat maun prasannatma trisul bhaiksabhuk sada
6 ekalingam smasanam va vrksamulam athapi va
ratrav asritya yuktatma samadhisthas cared vratam
203
7 ekavasa jitakrodho digvasa vasahayavan
vivecanena satatam bhumim apuya panditah
8 sanakair niviset tatra bhasma dattva tatopari
dvesam utpadayan yuktya lokasyaparamarthatah
9 vimalena hrda nityam sampurnena ca cetasa
hrtpadmakarnikasamstham taditpujacayopamam
10 svangusthaparvana tulyam vyaktangavayavam subham
pacavaktram sasankardhasekharam tam trilocanam
11 dhyatva samyag viset tatra yo yam so ham na samsayah
dhyayams cared diva ratrau vyomavyapimanum smaran
12 sangam brahmasamayuktam alpabhuk sattvavan sada
statapasaho nityam maitr karunikah sada
13 bhavayan munisardula jagad atmany avasthitam
sthito vaham jagaty asmims cittajid yogato khile
.....................................................................................
18 paramesarkakiranaih samsprstasya mahatmanah
carato sya sivajanaprakatkrtacetasah
19 parair akrusyamanah sams tadyamanas ca sadhakah
hrstim ayati paramam caryatejopabrmhitah
20 matto nanyah kvacic chrestho jaghanyo vatha vidyate
sarvam evabhimanyeta jagad bandhur iva sthitam
21 abhyantare prahrstatma bahyavrttisamasrayat
evam vai caratas tasya gramad gramantaresv atha
22 samvatsaraj jayet sarvan dvandvan dehaprabadhakan
samvatsaradvayenasau trikalajanam apnuyat
23 agnivat sarvadah syad virajaskah sa jayate
tribhih pasyati siddhams ca caturbhis ca divaukasah
24 pacabhis ca naravyaghra rudran pasyaty asamsayah
taih sardham ramate sadbhis tattulyabalavikramah
25 svecchavidhay bhavati caryavryabalan mune
evam padat padam yati utkrsto vapusa sada
26 krdatas ca pare vyomni bhagavan sa sadasivah
dadati padam atmyam aisvaram sadhakatmanah
27 nirsvaram va paramam sasvatam padam uttamam
etad vidhanam paramam vratasyasya prakrtitam
28 dvityam aparam bhadram samasat kathayami te
aluptaniyamotthabhir vasanabhir yada mune
29 viyuktakhilakalusyo dhvastasangena cetasa
tada caryasu visrantah sarvarthaksaparisraman
30 bahiskrtya jagat sarvam vratastho pastasamsayah
bahiskrtas ca lokena gatvaranyam atandritah
31 smasanam va jitapranah pranan samyamya sadhakah
tyajed deham sukhenaiva virakto tva duhsahat
32 samsararamagahanad anityat sumahabhayat
204
Testimonia:
3a vyaktatapasas] vyaktalingapurvakatvad avyaktadikramasya
PaBha ad 4.10 3d dvesac ca kalusatmakat] atrakalusa yasya matih so
yam akalusamatih, bahuvrhisamasah, atrakarah kalusapratisedhe.
bhavakalusyam evatra kalusyam. katham gamyate. praksiddhatvat.
iha ca purastad uktam. naveksed nabhibhased ity ukte arthad
apannam drste cabhibhasite ca dvesecchakrodha utpadyante
PaBha ad 1.18 7ab ekavasah ...digvasah] ekavasa avasa va PaSu
1.10-11 7c-8b vivecanena satatam bhumim apuya panditah / sanakair
niviset tatra bhasma dattva tatopari] vastrasikyabhasmadhara-
bhaiksabhajanadni muhur muhur vivecayitavyani. ... muhur muhur
vi<vecya> ...... bhupradese *bhasma prastarati (conj. : bhavati Ed)
PaBha p. 17; bhupradesam vastrantadimrdupavitrena vivecya
bhasmanaiva sucim kuryat. tad anu tatropavisya vidhyabhini-
vistas tavat tished yavad atinidrabhibhutah srantas ca bhavati.
tatah punar utthaya vivecayet. tad anu mantraih samskrtya bhasma
prabhutam prastaret GanKaT p. 19 8cd dvesam utpadayan yuktya
lokasya] avamatah sarvabhutesu paribhuyamanas caret PaSu
3.3-5; api tat kuryad api tat bhased yena paribhavam gacchet.
paribhuyamano hi vidvan krtsnatapa bhavati PaSu 3.16-19;
vipartani karmani kurval loke jugupsitah aparamarthatah]
anenanrtabhiyogena-sya yat tesam sukrtam tad agacchati. asyapi ca
yat papam tat tan gacchati PaBha ad 3.12, 3.13 9a vimalena hrda]
akalusamateh PaSu 1.18; akalusyena bhavena jantum pasyeta
sarvatah NiMu 4.94 9c-11a htpadmakarnikasamstham ... ... dhyatva]
onkaram adhidhyayta. hrdi kurvta dharanam PaSu 5.24-25 12c
statapasaho] godharma mrgadharma va PaSu 5.18; tayos tu sati
dharmabahutve samano dharmo grhyate: adhyatmikadi-
dvandvasahisnutvam PaBha ad loc; statapapariklesair ... japa-
dhyanaparo nityam sarvadvandvasahisnukah NiMu 4.96 maitr]
yog ... maitro bhijayate PS 5.2, 5,5-6 19ab parair akrusyamanah
sams tadyamanas ca sadhakah] paribhuyamanas caret PaSu 3.5;
yastimustyadibhih samyojanam paribhavah PaBha ad loc. ;
akrusyamano na vadami kim cit ksamamy aham tadyamanas ca
nityam Mahbhrata (Pune ed. ) 12.288.12ab; akrusyamano
nakrosen manyur eva titiksatah 12.288.16ab (quoted by Kaundinya
ad PaSu 4.12). 31a smasanam va] smasanavas dharmatma
yathalabdhopajvikah / labhate rudrasayujyam sada rudram
anusmaret PaSu 5.30-34
3a vyajyate vyaktatapasas conj. : vyajyam vyaktitayapasvarya C
[perhaps from vyajyate (Aisa form) vyaktatapasas carya] : vya-jyate
klistatapasas carya Bhatt 4b pracodita Bhatt : pravedina C 5c jat C
: japen Bhatt maun Bhattt : man C 7d apuya conj. : apurya
205
Codd. Bhatt 8a niviset KhChJ : nivaset N : nivaset C Bhatt 8d
lokasyaparamarthatah C : lokasya paramarthatah cett. Bhatt 10b
vyaktangavayavam C : vyaktavayavam Kh : yaktavayavayam
ChJ : suvyaktavayavam Bhatt 11c cared C Bhatt : caran N : caram
KhChJ diva ratrau C : divaratram cett. Bhatt 11d vyomavyapi-
manum smaran Bhatt : vyapimanusmaran KhNChJ :
vyomavyapinam anusmaran C 12a sangam C : sanga cett. Bhatt 12b
sattvavan NChJ Bhatt : satyavam Kh : tatvavan C 13d yogatokhile
C : yogatokhilam cett. Bhatt 19a akrusyamanah em. : akrsyamanah
Codd. Bhatt 25cd yati utkrsto vapusa C : bhoktur aklistavapusah
NChJ Bhatt : bhoktur aklistavapusah Kh 29d parisraman C :
parisramat J : paribhramat KhNCh Bhatt 30b pasta conj. : patta
KhChJ Bhatt : panga N samsayah CKh : samgamah Bhatt :
samgamam ChJ
1-3b. Now, O best of sages, I shall describe in brief the holy observance
of Rudra (rudravratam), as a result of which the wandering adept,
whose ascetic practice is openly displayed (vyaktatapasah) [at this
stage], wins without doubt full control of the world moving and
immobile, and by following the prescriptions of the Section of
Observance (carypdah), experiences manifestation of that
eternal liberation which, O brahmin, has come down through the
tradition of the Sivasastras.
3c-4. When a man has turned away in this manner from impure
desire and loathing wisdom suddenly arises [in him] roused by the
will of God. He will then become a fervent initiate devoted to
the propitiation of Rudra (rudrrdhanatatparah).
5-8. With the permission of his Guru he should take a trident and do
Japa in silence and tranquility of mind, seated on ashes, detached
from possessions, eating what he may obtain as alms. He should go
at night to a Linga which is the only one within a radius of
five leagues (ekalingam), or to a cremation ground (smasnam),
or to the foot of a tree, and there he should commence the practice
of his observance with his awareness inwardly focused, in a state
of undistracted concentration, wearing a single garment or none,
free of anger and unaccompanied by an assistant (asahyavn).
The wise [initiate], should repeatedly (satatam) purify the site
by examining it carefully [to remove any insects and the like that
he might harm]. He should then lay down a covering of ash [on
the ground] and then seat himself gently upon it.
9-13. He should contrive to rouse the hatred of the world [against
himself] by feigning [misconduct, disabilities and the like]
(aparamrthatah). As he does so he should meditate on [Rudra]
206
with a heart free of taint and complete awareness, visualizing him
as a radiant figure the size of ones thumb, shining liking a ball
of lightning in the the calix of the lotus of his heart, with each of
the parts of his body clearly distinct, five-faced, three-eyed, his
crest adorned with the sickle moon. When he has visualized [him
in this way] he should immerse himself in him entirely, [thinking]
Without a doubt this is who I am. Meditating [thus] he should
wander day and night, ever mindful of the Vyomavyapi mantra with
its auxiliaries and Brahmas, eating little and full of resolution,
indifferent to [extremes such as] heat and cold, ever benevolent and
compassionate, contemplating, O best of sages, that the universe
rests on his identity or that his identity pervades this universe,
conquering his mind through yoga. ... ... ... ...
19-20. Nourished by the radiant power of his observance (cary-
tejopabrmhitah) the adept experiences the height of joy even
as others verbally abuse or strike him. He should look on the
whole world as his kin, thinking that nobody is either better or
worse than himself.
21-27b. By manifesting this behaviour he experiences an inner
joy. If he wanders thus from village to village, then after a year
he will overcome all the contrasting extremes that torture the
body. After two years he will attain knowledge of the past,
present and future. Like fire he can burn whatever he wishes.
He becomes free of all impurity [of mind]. After three years he
beholds the Siddhas, and after four years the Gods. After five,
O best of men, he beholds without doubt the Rudras [themselves].
After six he sports in their company, with strength and power
equal [to theirs], accomplishing whatever he desires, O sage, all
by the power of his observance (caryvryabalt). Thus he
constantly rises from level to level (padt padam) [of the
universe], excelling [all others] in the beauty of his form (utkrsto
vapus). When [finally] he moves at will (krdatah) within the
highest Void the lord Sadasiva bestows on the Sadhaka his own
rank as God, or else the highest and eternal state beyond God
himself (nirsvaram).
27c-32b. I have explained to you the supreme procedure of this
ascetic observance. Now I shall briefly tell you another. When the
ascetic (vratasthah) has been freed of all impurities through the
impressions laid down by his unceasing austerities, then with his
awareness void of attachment, at peace in his observances, rejecting
(bahiskrtya) activities concerned with any sense-object or sense
(sarvrthksaparisramn), rejecting the whole world and rejected
by all (bahiskrtas ca lokena), casting aside all doubt (ap-
207
stasamsayah), he resolutely retreats to the wild or to a cremation
ground. As a master of his vital forces [through his practice of
Yoga] the adept can then throw off his body without effort,
by restraining the course of the vital energies, detached from the
ever-changing, horrifying, and utterly intolerable darkness of the
grove of Samsara.
Without a doubt this Vrata is closely related to that of
the Pacarthikas described in the Psupatastra. Particularly
striking in this regard is the fact that the ascetic is said to move
from a sedentary practice to one of wandering in society
provoking contempt (dvesam utpdayan). For this combination
is the essential and distinctive characteristic of that observance,
constituting its first two stages, termed the manifest (vyaktvasth)
and the unmanifest (avyaktvasth).
There is nothing here which corresponds to the Pacarthikas
third stage, the stage of Victory (jayah), in which he retreats to a
cave or deserted dwelling; but the fourth and fifth stages of his
observance, the stage of cutting off (chedah), in which he
moves to a cremation ground, and the stage of completion
(nisth), in which he leaves his body through yogic suicide, are
evidently the basis of what our text describes as its second
observance.
Even more striking than the similarity in the form and
progress of the Vratas is the fact that the Matangas account
seems to preserve elements of Pacarthika doctrine or devotiona-
lism, as though the passage were an incompletely agamicized
recension of an Atimargic document rather than an account of an
gamic Saiva practice which happens to have developed from a
Pacarthika prototype. The Yogin is described as a devotee of
Rudra (rudrrdhanatatparah), an anomalous description in the
gamic Saiva context. Also evocative of the Atimarga - though this
time it is perhaps the Lakula form that is evoked - is the description
of the successful ascetic coming to behold the Rudras and consort
with them as their equal, ascending from level to level of the cosmos.
Gradual ascent through the levels of the cosmos is the essence of
the Nisvsamukhas account of the Lakulas path to liberation;
37
and Ksemaraja describes the systems of the Mausulas and Karukas
in similar terms as methods for reaching (prptih) the Rudras of
their highest world-levels.
38
It might be argued that since such
208
ascents from level to level of the cosmos are characteristic of
gamic Yoga in general, this particular instance is no more
Lakula than any other, and so may be seen rather as an gamic
overcoding of what is merely a Pacarthika exemplar. However,
the emphasis that is placed on the Rudras in the present context
suggests otherwise. For while ascent through Rudra-worlds is
preserved in the gamic Saiva bhuvandhv texts, it is applied
only in the rituals of initiation, where it persists, as it were, in a
fossilized form. In the gamic texts, Yogic ascents to be experienced
directly rather than enacted in ritual, such as we have before us
here in this account of the Rudravrata, have moved away from
the hierarchy of Rudras as lords of worlds (bhuvanesah) to the
hierarchy of the principles or reality-levels (Tattvas) and to a series
of visualizations which are either completely impersonal or else
of deities other than the Rudras, and female as well as male.
39
Lakula asceticism, with its characteristic kaplavratam
or mahvratam, is, one would have thought, less a part of
Saiddhantika asceticism than of that advocated in the non-
Saiddhantika traditions of the Mantraptha and Vidyaptha.
This is largely true. One may say that the primary dichotomy
within the Mantramarga between the Saiddhantika and non-
Saiddhantika systems reproduces and reflects that between the
two branches of the Atimarga, the milder Pacarthika tradition
being matched by the Saiddhantika Saiva in this domain and
the more shocking Lakula tradition by the Kapalika cults of
Bhairava and the Goddesses seen in the Picumata, the Siddha-
yogesvarmata and the Jayadrathaymala and, to a lesser extent,
in the Svacchanda. But just as the Nisvsa shows particularly
intimate links with the antecedent Lakula tradition in other
respects so here, in spite of its place among the canonical works
of the Saiddhantikas, we find a range of observances that includes
all the elements attributed to the Lakulas by the Nisvsamukha and,
in addition, a number that match Yamunacaryas description of
Kalamukha practices in the Agamaprmnya. The Nisvsaguhyas
third Patala, called sarvasiddhisamdohah and, as this title indicates,
devoted to detailing methods (prayogh) by means of which
Sadhakas may attain Siddhis of every sort, prescribes the following
observances in this context:
209
f. 49r1-2 (3.30-34b):
30 siddhavidyavratastho hi jape ca vratam arabhet
go mata ca pita bhrata atithir [v]iprabrahmanah
31 hato me papa[ka]rena caren mithyavratam vrat
+ + + + kapalena khatvang bhasmagunthitah
32 smasane carate ratrau smasanavrata ucyate
nrtyate gayate caiva unmatto hasate bruvan
33 bhasmang cravasas ca ganavratam idam smrtam
japayukto bhaiksabhujo lostusay jitendriyah
34 dhyanasamyamayuktas ca lostukavratam acaret
31a papakarena conj. : [ + ]rena B : [ + + ]na A 31b vrat conj. :
vrata AB 31c kapalena B : lena A
The first of these Vratas, in which a person accuses himself
of the murder of a cow, his mother, his father, his brother or a
Brahman guest, is evidently in the tradition of provoking unmerited
condemnation through feigning sin that characterizes the
Pacarthika in the second stage of practice, in which he conceals
his identity from the world. The third, in which one smears
oneself with ashes, wears rags, dances, sings, laughs and babbles
like a madman, could also be said to go back to the same origin,
since the Psupatastra instructs the Pacarthika to provoke abuse
by acting like a madman (4.6: unmattavad vicareta). In the
Lakula system there was an independent Vrata of this name, an
unmattavratam. This, according to Abhinavaguptas commentary
on Bharatantyasstra, was the practice of Lakulas in the
advanced Paramayogin stage of their practice:
ad 12.85:
paramayogyavasthayam lakuladarsanapratipannanam unmatta-
vratam apy asti.
1 lakula em : nakula Ed
But the second Vrata, in which one goes about in a cremation
ground at night smeared with ashes, carrying a skull-bowl and a
skull-staff, has no parallel in the Pacarthika system, but is matched
only by the kaplavratam of the Lakula/Mahavratas.
210
NOTES AND REFERENCES
* This article is the text of my Ramalinga Reddy Memorial Lectures
written in December 1996 and delivered in the University of Madras
in January 1997, unchanged beyond the correction of minor errors and
the removal of some infelicities of expression. I have not attempted
substantial revision in the light of subsequent discoveries and
reflection, since though I now disagree with certain positions that
I have expressed here these are few and have no direct bearing on
the principal thesis. They are as follows. I have said here that the
earliest evidence of the Pacarthika tradition is the date of the
Mathura pilaster inscription (AD 380). In fact the long-held
assumption that this records a specifically Pacarthika tradition is
invalid, and the earliest evidence of the Pasupatas is in an inscription
of AD 376 (see Sanderson 2004, p. 444). Following Frauwallner I
have given the approximate date of Dharmakrti as 600-660. In fact
we can safely deduce about the prior limit only that it is after c. 550.
Finally, I have expressed the view here (as in Sanderson 1988, pp.
664-679 [132-147]) that of the four divisions of the Mahesvaras the
Kapalikas were the followers of certain non-Saiddhantika gamic
Saiva systems, notably those of the Vidyaptha, in which Kapalika
practice was conspicuous. The Kapalikas, also called Somasiddhantins,
were in fact, as a number of Saiva sources assert, a division of the
Atimarga, and a seventh-century copper-plate inscription recently
discovered in Chattisgarh confirms this tradition, revealing enough of
their doctrine to establish that they were a variant pre-Mantramargic
Saivism closely related to that of the Lakulas.
I am very grateful to my former pupil Dr. Dominic Goodall for his
kindness in helping me with the work of preparing the1997 manuscript
for publication and for overseeing in Madras the incorporation of the
final corrections. I also wish to thank my pupil Miss Isabelle Rati
for taking the time to go very carefully through my manuscript and
its corrections at the last hour. Her kindness has saved me from a
number of misprints that I had overlooked. Finally, I express my most
sincere thanks to Dr. Godabarisha Mishra of the University of Madras
for ensuring that I was given time to make these corrections before
publication.
1. All the evidence of the chronology of gamic Saivism given in the
following has been set out in detail with text-references in Sanderson
2001, pp. 2-19.
2. This and the subsequent passages that I shall cite from the
Nisvsatattvasamhit are presented here in the orthography of the
211
manuscript. Thus generally consonants are geminated after -r-, -tt-
is degeminated before -v- (e.g. tatvam rather than tattvam), the
homorganic nasals are preferred to final -m (e.g. tvan tu and rather
than tvam tu) and the homorganic fricatives (smnah) are preferred
to final -h. (e.g. tatas sivah rather than tatah sivah. ). The scribe has
also preferred to use the labial fricative termed upadhmnyam before -
p/ph-. I have transcribed this as f (e.g. ataf param rather than atah
param). In these citations letters enclosed within square brackets are
those that are seen only in the apograph of AD 1912 (B), having been
lost in the ninth-century exemplar (A) through further friation of the
edges of the leaves of the codex after that date.
3. Vol. 1, p. 12, ll. 7-16. The verses quoted are found in Varhapurna
70.42-43 and 71.52-54 with inferior readings.
4. Mrg KP 8.78-79.
5. See MrgVr on KP 8.78-79.
6. SvT 11.43c-45b.
7. JY, Satka 1, f. 166v6 (35.71c-72): evam caturvidham sstram lauki-
kdyam ca pacadh / 72 laukikam vaidikdhytmam ati(em. : avi
Cod. ) mrgam athnavam / phalabhedavibhinna ca sstram evam
tu pacadh. The term navam here means pertaining to Mantras
(anuh) and so denotes the Mantramarga.
8. NisMu f. 2v5 (1.47): laukikam sampravaksymi yena svargam
vrajanti te /; 15r5 (4.1): vedadharmam katham deva kartavyo gatim
icchat / svargpavargahetos ca pras<dd> vaktum arhasi //; 16r6
(4.40cd): vedadharmo may proktah svarganai<h>sreyasaparah /.
9. See the verse quoted without attribution at PaBha p.5, ll.15-16:
smkhyayogena ye mukth smkhyayogesvars ca ye / brahmdayas
tiryaganth sarve te pasavah smrth //, [Up to and including] those
liberated by Samkhya and Yoga and the [omniscient] lords of that
system, all [creatures] from Brahma down to animals are termed bound
souls (pasuh). See also SvT 11.68c-74, which has the order Buddhism,
Jainism, Vedism, Samkhya, Yoga, Atimarga, Saiva.
11. NiMu f. 17v2 (4.87cd): atysramavratam khytam lokttam ca
me srnu; f. 17v5 (4.97v): lokttam samkhytam; f. 18v3 (4.128):
kaplavratam sritya (corr. : srtya AB)dhruvam gacchanti tat
padam / lokttam samkhytam mahpsupatam vratam.
12. NiMu f. 3r1 (1.51b): loktt mahvrath /.
13. SvTU 6 (11) p.103, l.9: co pyarthe bhinnakramah.
14. MrgVr Kp p. 146, ll. 16-17: *timrgni (em. : atimrgni Ed.)karmni
yogajanyni tatpurusdhidaivatasntikalvyptyantni sodhyni
212
15. See Lorenzen 1972, pp.74-76.
16. An alternative to the emendation I propose is pramnptavicrakh.
This is closer graphically to the transmitted reading. It would mean Expert
scholars of the Pramanas. However, the compound ptavicraka- is
awkward. My emendation pramnrthavicrakh is prompted by the
definition of Mmamsa as vedrthavicrah, examination of the
meaning of the Veda.
17. Paramoksanirsakrikvrtti p. 7, ll. 11-12: ta eva hrdayapramn-
digranthakartrtventra strakrt [pramnaJ-kartrtvapadeno-
paksipth.
18. GanKa, Appendix II, p.33, ll.4-5: ity darsakrdibhis trthakarair
nirpitam.
19. SvTU 6 (11) p.54, ll.14-15: bauddhebhyo lkulnts tattvdhva-
bhjo na mukth, [All these religious] from the Buddhists up to the
Lakulas remain within the hierarchy of the tattvas. They are not fully
liberated.; SvTU 2 (4), p.247, ll.8-9: saivapsupatalkuldibhir
nntmavdibhih, Believers in a plurality of selves, such as the
[dualistic] Saivas, the Pasupatas and the Lakulas.
20. SvTU 6 (11), p. 54, ll.3-7.
21. SvT 2 (4), p.247, ll.8-9: saivapsupatalkuldibhir nntmavdibhih
......
22. This verse is quoted in the Dksdarsa; see SoSP
III, pp.553, 559.
23. Lorenzen 1972, pp.4-6.
24. PaBha p.8, l.6: prvam atahsabdt parksitam brhmanam ......
25. PaBha p.8, l.7: ...... <krta>vratopavsdyam ......
26. GanKar p.8, l.22 - p.9, l.10.
27. For prakriy in the sense of system, meaning a hierarchy of worlds,
tattvas and the like, see, e.g., Nisvsottara, f. 23v6 (1.14ab):
prakriyvypino mantr yathdhvne vyavasthith.
28. MatVr on VP 26.63d (mokso vtha catustayt).
29. Cf. SvT 11.73: dksjnavisuddhtm dehntam yva caryay /
kaplavratam sthya svam svam gacchati tat padam. Ksemaraja
introduces this as applying to the followers of the Vaimala and Pramana
systems: vaimalapramnasstranistho hi ...; and he comments on the
expression dksjnavisuddhtm to the effect that it is their reliance
on dks and jnam that differentiates them from the Mausulas and
Karukas, who rely on nothing but their ritualistic ascetic observance:
213
dksjnavisuddhtmetipadena proktakriypradhnavratamtra-
nisthamausulakrukebhyo pi viseso darsitah.
30. SvT4.97ab: carydhynavisuddhtm labhate padam aisvaram.
31. SvT10.1170ab: bhasmanisthjapadhyns te vrajanty aisvaram
padam; 11.71ab: vrate psupate proktam aisvaram paramam padam;
11.74ab: japabhasmakriynisths te vrajanty aisvaram padam.
32. SoSP III, samayadksvidhih 107.
33. Aghorasiva. See SoSP III, p. 100.
34. Brunners stage 3. See introd. to SoSP III, p. xxxiii.
35. SvTU 6 (11), p.52, l.14 - p.53, l.2: psupatabheda eva tu kriypradhne
mausule kruke caiva mytattvam prakrtitam srlakulesasisyena
musulendrena krohanasthvatrnena cparena mytattvaga-
ksemesabrahmasvmiprptihetukriybahulh sve sve sstre
vratavises ukt iti mytattvam eva tatra paramam padam; ibid.
p.53, ll.17-19: dksjnavisuddhtmetipadena proktakriypradhna-
vrata-mtranisthamausulakrukebhyo pi viseso darsitah.
36. NiGu f. 68r1-v1 (7.261-279b): 261 tejsas ca dhruvas caiva
pramanadhvana krttitam / kapalavratam asthaya pramanagama*
siddhaye (em. :si [ + + ] AB) / 262 gata dhruvapadam ye tu
dksajanavisodhitah / atha bhagavatonkaram vidhumam suksmavya-
pinam / dhruva caivoparistat tu bhuvanatrayasamhatim / rudra-
pratoda*manavas (em. : manavat Cod. ) trayaite bhuvanesvarah / paca-
laksam smrtam tesam ekaikasya mahatmanah / *tan (em. : tam Cod. )
atikramya-m ekaksah pingalo hansa-m eva ca / saumyas te tu trayo rudra
ity ete parikrti[tah] / dasakotiparvaraf parato nyo dhruvocyate /
yasmin na mrtyur na jara na samsaro vidhyate /dasakotisahasrais tu
krdante bhuvanesvarah / atoparistad rudranam dasaika bhuvana
smrtah / *brahmodarkko (em. : brahmadarkko Cod. ) dindimundis
saurabheyas ca pavanah / janmamrtyuharo rudrah prantas sukhaduh-
khitah / vijrmbhitas ca vikhyato rudra hy ekadas[aiva tu] / suksma ca
avyaya nitya galake sampratisthitah / trtyan tu dhruvam vaksye
nirajanam ataf param / saptakotisahasrani bhuvananam dhruvasya
tu / nirajanasya devasya parisamkhya na vidyate / atisayabuddhiyukta
pradhanaracitani tu / satasimhoparistat tu vidyarajesvaresvarah / saktis
caivordhvasamskaram khyati pramuditas tatha / [prala]mba ca tatha
visnum mandaram gahanan tatha / pravitata ca-m uddistam astau te
paramesvarah / prabhava samayah ksudrah vimalam sivapacamam /
trayodasa mahavrya bhuvanesaf prakrttitah / samkhyata sahasrani
parivaro bhuvanani vai / paratas tu ghano nama durbhedyas sarva-
vadinam / nlotpaladalasyam yatha *samvarttakanalah (conj. : samvartta
[+ + + ] A : samvarttaka[ + + ] B) / ghurmmate jvaladhumena ghano janena
214
samyutah / *amrto mrtasamspars (conj. : amrtamrtasamspars AB)
trilaksabhuvanesvarah / atto devanikayair ghana ity abhidhyate /
dhyanaharo mahateja rudrais tu bahubhir vrtah / amrtamrtaprado hy esa
rudronkaram atah param / taluke devadevas tu rudronkaraf pratisthitah /
sadasivo stabhedena bhru*mandale tu drsyate (conj. : manda[ + + + + + ]
A : ma[ + + ] drsyata B); Dksottara A, p. 901-903, B, pp. 82-84 (7.78-90);
A pp. 1149-1151 (19.43-56); SvT 10.1174-1265b.
37. NiMu f. 17v4 (4.95cd): avcydi dhruvntam ca etaj jtv vimucyate .
38. SvTU 6 (11), p.52, l.14 - p.53, l.2: psupatabheda eva tu kriypradhne
mausule kruke caiva mytattvam prakrtitam (11.71cd).
srlakulesasisyena musulendrena krohanasthnvatrnena cparena
mytattvagaksemesabrahmasvmiprptihetukriybahulh sve sve
sstre vratavises ukt iti mytattvam eva tatra paramam padam.
39. See, e.g., SvT 12.83c-168.
ABBREVIATIONS
Guhya = Nisvsaguhya
EFEO = Ecole franaise dExtrme-Orient
GanKa, GanKar = Ganakrik, -tk (Ratnatk)
IFI = Institut franais dIndologie, Pondicherry
JY = Jayadrathaymala
K = Khmer inscription, numbered as in Coeds 1966
KP = Kriyapada
KSTS = Kashmir Series of Texts and Studies
Mat, MatVr = Matanga and -vrtti.
Mrg, MrgVr = Mrgendra and -vrtti
NAK= National Archives, Kathmandu
NGMPP = Nepal-German Manuscript Preservation Project
NiGu = Nisvsaguhya
NiMu = Nisvsamukha
PaBha = Pacrthabhsya
PaSu = Psupatastra
SoSP = Somasambhupaddhati
SvT, SvTU = Svacchanda[tantra} and -uddyota
T = Devanagar transcript prepared for the IFI
VP = Vidypda
215
BIBLIOGRAPHY
Primary Sources
Astaprakarana, ed. Krsnasastr. Devakottai: Sivagamasiddhantaparipalana-
sangha. 2 vols. 1923 and 1925.
svarapratyabhijvivrtivimarsin of Abhinavagupta, ed. Madhusudan
Kaul Sastr. 3 vols. KSTS 60, 62, 65. Bombay, 1938, 1941, 1943.
Agamaprmnya of Yamunacarya, ed. Rama Misra Sastr. Reprint from
The Pandit. Benares, 1937.
Ganakrik with the commentary Ratnatk, ed. C.D. Dalal. Gaekwads
Oriental Series, No. 15. Baroda: Central Library, 1920.
Guhyasiddhi. In Guhydi-astasiddhi-sangraha, ed. Samdhong Rinpoche
and Vrajvallabh Dwivedi. Sarnath: Rare Buddhist Text Project. 1987. Rare
Buddhist Text Series, 1.
Jayadrathaymala, Satka 1. NAK MS 5-4650, NGMPP Reel B 122/7.
Paper; Newari script.
Jayadrathaymala, Satka 3. NAK MS 5-1975, NGMPP Reel No. A 152/9.
Paper; Newari script.
Tantrasadbhva. NAK 5-445, NGMPP Reel A 44/2. Palm-leaf; early Newari
script; A.D. 1096/97.
Tantrloka of Abhinavagupta with the commentary (-viveka) of Rajanaka
Jayaratha, ed. Mukund Ram Sastr. KSTS 23, 28, 30, 35, 29, 41,47, 59, 52, 57,
58. Bombay and Srinagar, 1918-38.
Dksottara. See Nisvsakrik.
Dptgama. IFI T. 507 (A); IFI T. 147 (B); IFI T. 145 (C).
Naresvaraparks of Sadyojyotis with the commentary (-praksa) of Bhatta
Ramakantha, ed. Madhusudan Kaul Sastri. KSTS 45. Srinagar, 1926.
Ntyasstra ascribed to Bharatamuni with the commentary (Abhinavabhrat
of Abhinavagupta. 4 vols. Baroda: Central library, 1926-64. Gaekwads
Oriental Series, 36, 68, 124, 145. Vols 1-3 ed. Manavalli Ramakrishna Kavi;
vol. 4, ed. M. Ramakrishna Kavi and J. S. Pade.
Nisvsakrik. IFI MS T. 17, comprising 61 Patalas [(pp. 1-794), followed by
the Dksottara under the same heading, in 19 Patalas (pp. 795-1152) (A);
IFI MS T. 150, incomplete but containing the whole of the Dksottara
(pp. 32-202) (B).
Nisvsaguhya. = Nisvsatattvasamhit, ff. 42r5-114v.
Nisvsatattvasamhit. NAK MS 1-227, NGMPP Reel No. A 41/14. Palm-leaf;
Nepalese Licchavi script; c. A.D. 900 (A); Apograph of AD 1912: London.
Wellcome Institute for the History of Medicine, Sanskrit MS I.33 (B).
216
Nisvsanaya. = Nisvsatattvasamhit, ff. 29r5-42r5.
Nisvsamukha. = Nisvsatattvasamhit, ff. 1v-18v6. Full title in Patala
colophons: Nisvsamukhatattvasamhit
Nisvsamla. = Nisvsatattvasamhit, ff. 18v6-23v1.
Nisvsottara. = Nisvsatattvasamhit, ff. 23v1-29r5.
Pacrthabhsya. See Psupatastra.
Paramoksanirsakrik of Sadyojyotis with the commentary (-vrtti) of
Bhatta Ramakantha. In Astaprakarana, vol. 2.
Partrimsik with the commentary (-vivarana) of Abhinavagupta, ed.
Mukunda Rama Shastr. KSTS 18. Bombay, 1918.
Psupatastra (=Pacrtha) with the commentary (Pacrthabhsya)
of Kaundinya, ed. R. Ananthakrishna Sastri, Trivandrum Sanskrit Series,
no. 118. Trivandrum: The Oriental Manuscript Library of the University of
Travancore, 1940.
Picumata (/Brahmaymala). NAK MS 3-370, NGMPP Reel No. A 42/6.
Palm-leaf; early Newari script; dated in A.D. 1052.
Matanga: Matangapramesvargama, Vidypda, with the commentary
(Matangavrtti) of Bhatta Ramakantha, ed. N.R. Bhatt. Publications de lIFI,
Pondicherry, No. 66. Pondicherry: IFI, 1977; Matangapramesvargama,
Kriypda, Yogapda and Carypda, with the commentary (Matangavrtti)
of Bhatta Ramakantha up to Kriypda 11.12b, ed. N.R. Bhatt. Publications
de lIFI, Pondicherry, No. 65. Pondicherry: IFI, 1982.
In the abbreviated apparatus of my edition here of Carypda p. 9:1-
32b C is a Nepalese palm-leaf manuscript. The other witnesses are
Grantha manuscripts from Tamilnadu. No Kashmirian manuscript
survives for this part of the text.
Brhatklottara. NAK MS 1-89, NGMPP Reel B 24/59 (Klottaratantram).
Palm-leaf; early Newari script.
Mrgendra: Mrgendratantra, Vidypda and Yogapda, with the commentary
(-vrtti) of Bhatta Narayanakantha, ed. Madhusudan Kaul Sastr. KSTS 50.
Srinagar, 1930; Mrgendra: Mrgendrgama (= Mrgendratantra), Kriypda
and incomplete Carypada, with the commentary (-vrtti) of Bhatta
Narayanakantha, ed. N.R. Bhatt. Publications de lIFI, No. 23. Pondicherry:
IFI, 1962.
Moksakrik of Sadyojyotis with the commentary (-vrtti) of Bhatta
Ramakantha. In Astaprakarana, vol. 2.
Yjavalkyasmrtitk of Apararka, ed. the Pandits of the nandasrama.
Pune, 1903-4. 2 vols. nandasramasamskrtagranthaval, 46.
Varhapurna, ed. S. Gupta. Varanasi: All-India Kashiraj Trust, 1981.
217
Sivatattvaratnkara of Basavaraja, ed. Narayanasvamisastr. 3 vols. Mysore:
Pracyavidyasamsodhanalayah, Maisuruvisvavidyanilayah, 1964-1984.
University of Mysore. Oriental Research Institute Series, 108, 112, 148.
Saivaparibhs of Sivagrayogndra Jansivacarya, ed. H.R. Rangaswamy
Iyengar and R. Ramasastri, Mysore, 1950. Oriental Research Institute
Publications, Sanskrit Series 90.
Somasambhupaddhati. See Brunner-Lachaux 1963-1998.
Svacchandatantra with the commentary (Svacchandoddyota) of Rajanaka
Ksemaraja, ed. Madhusudan Kaul Sastr. KSTS 31, 38, 44, 48, 51, 53, 56.
Bombay, 1921-35.
Secondary Sources
BRUNNER-LACHAUX, Hlne. Ed. and trans. 1963, 1968, 1977, 1998.
Somasambhupaddhati. 4 vols. Premire Partie. Le rituel quotidien dans la
tradition sivate de lInde du Sud selon Somasambhu; Deuxime Partie. Rituels
occasionels dans la tradition sivate de lInde du Sud selon Somasambhu I.
Pavitrrohana, Damanapj et Pryascitta; Troisime Partie. Rituels
occasionels dans la tradition sivate de lInde du Sud selon Somasambhu II.
dks, abhiseka, vratoddhra, antyesti, srddha; Quatrime Partie. rituels
optionnels. pratisth. Publications de lIFI No. 25. Pondicherry: IFI.
COEDS, George. 1966. Inscriptions du Cambodge. Volume 8. Paris: EFEO.
LORENZEN, David. 1972. The Kplikas and Klmukhas. Two Lost Saivite
Sects. New Delhi: Thomson Press. Australian National University Centre of
Oriental Studies Oriental Monograph Series, Vol. XII.
SANDERSON, Alexis. 1988. Saivism and the Tantric Traditions. In The Worlds
Religions, edited by S. Sutherland, L. Houlden, P. Clarke and F. Hardy, pp.
660-704. London: Routledge. Reprinted in The Worlds Religions. The
Religions of Asia, edited by F. Hardy, pp. 128-72, London: Routledge, 1990.
SANDERSON, Alexis. 2001. History Through Textual Criticism in the Study of
Saivism, the Pacaratra and the Buddhist Yogintantras. In Les sources et le
temps. Sources and Time. A Colloquium, Pondicherry, 11-13 January 1997,
edited by Franois Grimal, pp. 1-47. Publications du dpartement
dIndologie 91. Pondicherry: IFP / EFEO.
SANDERSON, Alexis. 2004. The Saiva Religion Among the Khmers, Part 1.
Bulletin de lEcole franaise dExtrme-Orient, 90-91 (2003-2004), pp.
349-462.
***

Lakulas

Uploaded by

Copyright:

Available Formats

Lakulas

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lakulas

Uploaded by

Copyright:

Available Formats

What is the overall content about?

What are some of the main topics discussed?

PART - II

You might also like