saiva officiants
DESCRIPTION
saiva OfficiantsTRANSCRIPT
ALEXIS SANDERSON
RELIGION AND THE STATE: SAIVA OFFICIANTSIN THE TERRITORY OF THE KING’S
BRAHMANICAL CHAPLAIN
1. INTRODUCTION
The literature of the Saiva Mantram�arga is dominated by the pre-scription of the rituals through which the Saivas initiated candi-dates into their religious discipline (d�ıks
_�a), consecrated successors
to office (abhis_ekah
_), installed images and other substrates of wor-
ship (pratis_t_h�a), and performed the repeated services of worship
(y�agah_) and propitiation (mantras�adhanam) required of all or cer-
tain classes among them.1 By studying this literature, which extends
1 I adopt Mantram�arga (‘the Path of Mantras’) as the Saivas’ term for what
Indologists have commonly called Tantric or �Agamic Saivism. It serves to differen-tiate this Saivism from that of the Atim�arga, ‘‘the Path Beyond [the brahmanicalsocio-religious order]’’, the earlier and contemporaneous Saivism of the P�asu-pata divisions, principally the P�a~nc�arthikas, L�akulas/K�alamukhas and Soma-siddh�antins/K�ap�alikas. The Mantram�arga comprises a number of related systems.The principal among them and their principal surviving scriptures are (1) the Sid-
dh�anta taught in the Nisv�asa, the Paus_kara, the Sv�ayambhuvas�utrasam
_graha, the
Rauravas�utrasam_graha, the K�alottaras, the Mata _nga, the Kiran
_a, the Mr
_gendra, the
Par�akhya, the Br_hatk�alottara, etc., (2) the V�amasaiva cult of Tumburu and his four
sisters taught in the V�ın_�asikha, (3) the Daks
_in_asaiva cult of Svacchandabhairava
taught in the Svacchanda, (4) the Y�amala cult of Kap�al�ısa and Can_d_�a K�ap�alin�ı
taught in the Picumata ð=Brahmay�amala), (5) the Trika cult of the goddesses Par�a,etc. taught in the Siddhayogesvar�ımata, the Tantrasadbh�ava, the M�alin�ıvijayottara,etc., (6) the K�al�ıkula cult of K�alasam
_kars
_an_�ı/K�al�ı taught in the Jayadrathay�amala
and the scriptures of the Krama (K�al�ıkulapa~ncasataka, K�al�ıkulakramasadbh�ava,etc.), (7) the cult of Kubjik�a taught in the Kubjik�amata, etc., (8) the cult of Tripura-
sundar�ı taught in the Nity�as_o_dasik�arn
_ava, etc., and (9) the cult of Amr
_tesvara and/
or Amr_talaks
_m�ı taught in the Netra.
Indo-Iranian Journal 47: 229–300, 2004.
DOI: 10.1007/s10783-005-2927-y
* Springer 2005
from scriptural texts claiming the authority of divine revelationthrough commentaries and treatises on these texts to manuals(paddhatih
_) of both transregional and local reach, we can make out
a detailed picture of the procedures they advocated and throughcomparative analysis arrive at some understanding of how thesemodel rituals changed over time, were adapted in different regions,and were related to those of the similar systems of ritual seen inthe literatures of the P�ancar�atrika Vais
_n_avas and the Mah�ay�ana-Bud-
dhist Way of Mantras (mantranayah_, mantray�anam).
But these sources are much less revealing about agency, socialmilieu, and historical context. They do provide us with some gen-eral rules of restriction and permission concerning which categoriesof person may or may not be initiated or officiate and concerningthe extent to which their mundane social status influences their sta-tus in the community of co-initiates, and these rules are different inthe different Saiva systems, which reveals something of their charac-ter and interrelation within the larger social world. But they provideno data, and we are not likely to discover any from other sources,that would enable us to judge, for example, what percentage of thepopulation in a given region and time was involved in the practiceor support of the religion, or how its followers and supporters weredistributed between castes, economic classes, age-groups, gendersand levels or type of involvement. In other words the texts tell uswhat was possible for various groups but not the extent to whichthese possibilities were put into practice. There is nothing here likethe evidence provided by the records of the government departmentsthat supervised the conduct of religion in China and Japan. Thekingdoms of South and Southeast Asia engaged in some such super-vision and must have maintained the sort of records that would haveenabled us to address these questions. But they have not been pre-served. All we have from that quarter are what happens to have sur-vived and come to light of inscriptions on stone or copper platesrecording major grants or pious works. This is crucial informationfor the historian of Saivism, as it is for students of all Indian reli-gious traditions, and in some areas, such as that of the Khmers, itand archaeology provide the only evidence that we have. But at bestit instantiates or challenges the model of possibilities conveyed bythe prescriptive literature. It does not enable us to go beyond itsrange into detailed social history. Nonetheless it is possible, I wouldsay necessary, to read the literature and inscriptions with the sort ofquestions in mind that a social historian would wish to ask.
230 ALEXIS SANDERSON
In this perspective it seems to me that active initiates are likely tohave been few in number and to have been concentrated among,though by no means confined to, brahmin men. Yet Saivism exertedan influence on the religious life of the Indian world that far exceedswhat might be expected of such a minority, especially from one out-side the mainstream of brahmanical observance. For there can beno doubt that for several centuries after the sixth it was the princi-pal faith of the elites in large parts of the Indian subcontinent andin both mainland and insular Southeast Asia. Only Mah�ay�ana Bud-dhism was able to rival it during this period; and when it achievedsuccess in this rivalry, either equalling or excelling Saivism as thebeneficiary of patronage, it was in a form led by the Way of Man-tras, a system of ritual, meditation and observance in which Bud-dhism had redesigned itself, if not in essence, then at least in styleand range of functions, on the model of its rival.
I attribute this success to three factors. The first is that thoughthe practice of the religion proper was restricted to the initiated,they cultivated the support of a wider community of uninitiated,lay devotees. An unpublished corpus of texts comprising principallythe Sivadharma and the Sivadharmottara, contains the observancesrecommended to this laity, revealing that following the example ofthe Buddhists the Saivas had propagated a lesser religion of merit-gathering that centred on the support and veneration of the personsand institutions of the religion proper, promising that those whofollowed it would be rewarded in death by a period in the paradiseof Siva (sivalokah
_, rudralokah
_) before returning to the world in the
most desirable of rebirths.2
The second is that the Saivism of the Mantram�arga developedin practice a thorough accommodation of the brahmanical religionthat it claimed to transcend, thus minimizing, even eliminating, theoffence it gave as a tradition whose scriptures, like those of the Bud-dhists, were seen to be, and claimed to be, outside the corpus of theVedas. These Saivas were to accept that the brahmanical traditionalone was valid in the domain it claimed for itself and that they
2 For further information on this corpus of texts and for evidence in it of inter-action with the king and his court see Sanderson, forthcoming.
231SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
were bound to follow its prescriptions and incorporate its ritualsbeside their own wherever practicable.3
Similarly, where they established themselves at the many Saivatemple sites that pre-existed them they did not attempt to reformworship by restricting it to the narrow pantheon that they propiti-ated as initiates. This they imposed on the worship of Siva in theLin_ ga at the heart of these foundations; but they also took over,preserved, and regulated in accordance with the expectations of theuninitiated laity a much wider range of ancillary deities, deities thathave no place in the scriptures and ritual manuals of the Man-tram�arga other than in treatments of their installation and iconog-raphy in this special context.4
The third and most vital factor is that the religion succeeded inforging close links with the institution of kingship and therebywith the principal source of patronage. I see four main elements
3 Transgressing the rules of the brahmanical socio-religious system, known tothe Saivas as the mundane religion (laukiko dharmah
_), is forbidden in a much-
cited passage from the lost Saiddh�antika Bh�argavottara: iti varn_�asram�ac�ar�an
manas�api na la _nghayet=yo yasminn �asrame tis_t_han d�ıks
_itah
_sivas�asane/sa tasminn
eva sam_tis_t_hec chivadharmam
_ca p�alayet ‘So he should not transgress the practices
of his caste-class and order of life even in thought. He should remain in the order
in which he was when he was initiated into the Saiva religion and [at the sametime] maintain the ordinances of Siva’. It is cited at, e.g., Naresvarapar�ıks
_�aprak�asa
ad 3.76. See also Sarvaj~n�anottara cited in Tantr�alokaviveka ad 4.251ab, and
Mata _ngap�aramesvara, Cary�ap�ada 2.2–7b. That the Saivas, at least those followingthe Saiddh�antika forms of observance, came to be widely accepted as co-religion-ists in traditional brahmanical circles is evident from the report of the Kashmirian
philosopher Jayantabhat_t_a, a contemporary of king Sa_nkaravarman (r. 883–902),
in Ny�ayama~njar�ı vol. 1, p. 636, l.15–p. 637, l.4; p. 637, ll. 16–19; p. 638, ll. 12–13.He defends the validity of the Saiva scriptures and claims that his position is that
of the society of respectable �Aryas (mah�ajanah_), which he defines as comprising all
who live within the system of the four caste-classes and orders of life in accor-dance with the ordinances of the Veda. See also Sanderson, 1995, pp. 27–38.
4 The principal of these ancillaries were Durg�a Mahis_�asuramardin�ı, Um�a,
Gan_esa, Skanda, Vis
_n_u, Brahm�a, S�urya, Laks
_m�ı, Sarasvat�ı, the Lokap�alas, the
Grahas, the Mother goddesses (M�at_rs), and a number of non-Mantram�argic Siva
forms: (1) a simple single-faced Siva with two or more arms, (2) Harihara or Sa_nka-ran�ar�ayan
_a, in which the left half of Siva’s body is Vis
_n_u, (3) Ardhan�ar�ısvara or
Gaur�ısvara, inwhichthis half is his consort Um�a, (4) the dancing Rudra, called vari-ously Nr
_tyarudra, Nr
_ttesvara, Nr
_tyesvara, Nat
_esvara, N�at
_akesvara and N�at
_yesvara,
and (5) Um�amahesvara, also called Umesa and Um�arudra, in which Um�a sits onSiva’s left thigh with his arm around her. Early sources that cover their iconographyare the Pratis
_t_h�atantras Devy�amata, Mayasam
_graha, Pi _ngal�amata, and Mohac�uro-
ttara, and the general scriptureKiran_a (Pat
_ala 52). See Sanderson, 2005, pp. 435–440.
232 ALEXIS SANDERSON
here: (1) the occupying by Saiva officiants of the office of RoyalPreceptor (r�ajaguruh
_) and in this position their giving Saiva initia-
tion (d�ıks_�a) to the monarch followed by a specially modified version
of the Saiva consecration ritual (abhis_ekah
_) as an empowerment to
rule beyond that conferred by the conventional brahmanical royalconsecration (r�ajy�abhis
_ekah
_); (2) the promoting by Saiva officiants
of the practice of displaying and legitimating a dynasty’s power bytheir officiating in the founding of Saiva temples in which the newSivas that they enshrined bore as the individuating first half of theirnames that of the royal founder or, where complexes of royal Sivatemples were established, those of the founder and any kin that hemight designate for this purpose; (3) the provision of a repertoire ofprotective, therapeutic and aggressive rites for the benefit of themonarch and his kingdom; and (4) the development of Saiva ritualsand their applications to enable a specialized class of Saiva officiantsto encroach on the territory of the R�ajapurohita, the brahmanicalexpert in the rites of the Atharvaveda who served as the personalpriest of the king, warding off all manner of ills from him throughapotropaic rites, using sorcery to attack his enemies, fulfilling themanifold duties of regular and occasional worship on his behalf,and performing the funerary and other postmortuary rites when heor other members of the royal family died.5
In a forthcoming monograph I have provided detailed evidenceof the first of these four factors.6 Here I consider the fourth, and,by way of introduction, the third. For the two overlap. As we shallsee, a Saiva Guru acting in the role of a king’s personal chaplainwas expected, like his brahmanical counterpart, to perform rites toprotect the king and kingdom. But the two factors must be distin-guished, since the performance of such rites was also commissionedfrom independent Saiva Gurus acting outside this role.
2. STATE PROTECTION BY INDEPENDENTSAIVA OFFICIANTS
An inscription of the fifth year of the reign of the Cola emperorR�aj�adhir�aja II (r. 1163–79 or 1166–82) tells us that when an armyfrom Sri Lanka had invaded the mainland, removed the door of theR�amesvaram temple, obstructed the worship, and carried away all
5 See Sanderson, forthcoming, for textual sources requiring the royal chaplain to be anAtharvavedin or expert in the apotropaic and other rites of the Atharvavedic tradition.
6 Sanderson, forthcoming.
233SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
the temple’s treasures, a certain Jn�anasiva, whose name shows him tohave been a Saiddh�antika Saiva officiant, was engaged by the em-peror to perform a ritual that would bring destruction on thoseresponsible for this desecration. According to the inscription the cere-mony was continued for twenty-eight days and at its end the invad-ing army was indeed defeated.7
Another example of a ritual performed by a Saiva officiant forthe good of the state is seen in a Sanskrit and Old Khmer inscrip-tion of AD 1052 from Sdok Kak Thom. This reports that a certainHiran
_yad�ama was commissioned by Jayavarman II (r. 802–c. 835)
at the time of his founding of the unified kingdom of the Khmersto perform a ritual of the V�amasaiva system that would guaranteethat unity and the kingdom’s independence from Jav�a:8
man vr�ahman_a jmah
_hiran
_yad�ama pr�aj~na siddhi vidy�a mok amvi jana-
pada pi vrah_p�ada paramesvara a~njen thve vidhi leha le _n kam pi kamvu-
jadesa neh_�ayatta ta jav�a ley le _n �ac ti kamrate _n phdai karom
_mv�ay guh
_ta j�a cakravartti vr�ahman_a noh
_thve vidhi toy vrah
_vin�asikha pratis
_t_ha
kamrate _n jagat ta r�aja vr�ahman_a noh
_paryyan vrah
_vin�asikha nayottara
sam_moha sirascheda sva _n man svat ta mukha cu _n pi sarsir pi paryann
ste _n a~n sivakaivalya nu gi
K. 235, Old Khmer text, ll. 71–74
Then a brahmin called Hiran_yad�ama, who was learned in the Mantras
that bestow Siddhi, came from Janapada. The Venerable Paramesvara[Jayavarman II] requested him to perform a ritual so that this land of
7 ARE 20 of 1899 at �Arpp�akkam, a village eight miles SSE of K�anc�ıpuram;
SII 4, no. 456; summary in Sastri (1984, p. 368). For a translation of the relevantpart see ARE 1899, paragraph 34.
8 Cœdes thought (1968, p. 100) that this Jav�a was the island of Java. But the iden-
tification is uncertain. Vickery proposes that it was Champa (Skt. camp�a), the land of
the Chams to the east in Southern Vietnam, claiming that ‘‘jav�a/chvea/ has been used
in Khmer until modern times to designate the Cham’’ (1998, p. 29; see also pp. 387
and 405). Champa and the Chams are frequently mentioned in the inscriptions but
never under this name, and in modern Khmer, according to the Cham scholar Phoen
(1987, p. 78), the term cited refers not to the Chams but to Malays descended from
Muslim immigrants from the Indonesian archipelago and the Malay peninsula in the
fourteenth and fifteenth centuries. The Khmer expression ‘‘cam jva’’ [i.e. cham chvea]refers, he reports, to the Muslim community in Cambodia in general as compris-ing both Chams and these Malays. Guesdon (1930, s.v.) gives chvea in the mean-
ing Java and by extension the Malay peninsula (‘‘malaisie’’). It is therefore moreprobable that the independence to which the inscription refers was from a king-dom in maritime Southeast Asia, probably that of the Sailendras of Sr�ıvijaya cen-tred in southeastern Sumatra.
234 ALEXIS SANDERSON
the Kambujas would no longer be subject to Jav�a and only one kingwould rule over it with sovereign power. That brahmin performed theritual [for those ends] following the venerable Vin�asikha9 and es-tablished the [image of the] Kamrate_n Jagat ta R�aja. The brahmin
[then] taught the Vin�asikha, the Nayottara, the Sam_moha and the
Sirascheda. He recited them from beginning to end so that they couldbe written down, and taught them to Ste _n an Sivakaivalya.
In none of these cases is it clear that the officiants engaged wereplacing themselves beyond the domain of occasional rites for thebenefit of others by becoming priests tied to the service of the kingor the state.
Much material in the Saivas’ prescriptive literature is similarlyambiguous. For example, the Uttarabh�aga of the Li _ngapur�an
_a, which
in spite of its claim to be a Pur�an_a, is in large part a thinly-disguised
Paddhati text of the Saiva Mantram�arga, teaches in addition to therituals of Saiddh�antika Li_nga worship and installation (1) fire-sacri-fices in which offerings are made to the Aghora aspect of Siva(in fact to the principal Daks
_in_asaiva deity Svacchandabhairava,
though this is not made explicit in the text) for the benefit of theking, to ward off danger from him and to restore his health,10 and
9 This is evidently the published V�ın_�asikha, the only complete V�amasaiva
scripture to have reached us. The form Vin�asikha in the Old Khmer text is con-
firmed by the Sanskrit (v. 28, cited below), where the metre requires the first syl-lable to be short. The error may be attributed to the passage of two and a halfcenturies between the introduction of these texts and the inscription. There is no
reason to assume that the Sanskrit original continued to be studied alongside thePaddhati based on it throughout this period.
10 See especially Li _ngapur�an_a, Uttarabh�aga, chapters 19–27 and 46–54. As a text
seeking acceptance as a Pur�an_a and thereby as a work within the corpus of Veda-
derived revelation, it disguises its properly Saiva character by jettisoning suchdistinctively Saiva doctrines as that of the thirty-six reality-levels (tattv�ani). It is, Isurmise, in the same spirit that it has avoided identifying its Aghora as Svaccha-
ndabhairava, the deity of the Svacchandatantra. That this is the true identity of thedeity is apparent in chapter 26, which is devoted to the worship of Aghora in the Li_ngaor, less desirably, on a Sthan
_d_ila, as an alternative to the regular Saiddh�antika Siva
worship in the Li_nga taught, with the necessary d�ıks_�a and ritual of Li_nga installa-
tion, in chapters 19–25 and 46–47. A royal fire-sacrifice to this deity is taught in 49and a ritual for destroying the king’s enemies in which the worshipper visualizeshimself as the same is taught in 50. The true name of the deity is not used, no doubt
because of its strongly non-Vedic associations, but his visualization (dhy�anam) in26.15–21b reveals his identity since it is of the five-faced and eighteen-armed Sva-cchandabhairava taught in Svacchanda 2.81c–97. All the hand-attributes are identi-
cal if, as I propose, we emend to mun_d_am_the inappropriate dan
_d_am_seen in 26.19c of
the published text.
235SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
(2) an elaborate S�akta Saiva procedure to guarantee that the kingwill be victorious when he goes into battle.11 All this is very muchwithin the purview of the purposes of the rituals of the brahmanicalroyal chaplain, but nothing in the text tells us whether the officiantenvisaged is a person acting in that role or an independent Gurucondescending to act for the benefit of the king in special circum-stances, like Jn�anasiva in the reign of R�aj�adhir�aja II or Hiran
_yad�ama
in that of Jayavarman II.
3. THE INSTITUTIONALIZATION OF STATE PROTECTION
In the last of these cases we have a record of a single ritual per-formed for state-protection by a Guru who was not in the king’sservice. But this instance also shows how such rituals could becomeregularized by transference to priests who were in such service. Forthe Sivakaivalya to whom Hiran
_yadama taught these four texts, the
principal scriptures of the V�ama division of the Saiva canon, wasJayavarman II’s R�ajapurohita.12 The king had him appointed toperform the regular worship of the image, the Kamrate_n Jagat taR�aja (Skt. Devar�aja), that Hiran
_yad�ama had established after this
ritual as the focus of a cult to protect the state; and it was agreedthat the duty and right to worship before this image should bepassed down through ascetics in Sivakaivalya’s matriline:
25 jayavarmmamah�ıbhr_to mahendr�a-
vanibhr_nm�urdhakr
_t�aspadasya s�ast�a
kavir �aryavar�a _ngavandit�a _nghrih_�sivakaivalya iti prat�ıtir �as�ıt
26 hiran_yad�amadvijapu _ngavo ’gryadh�ır
iv�abjayonih_karun
_�ardra �agatah
_ananyalabdh�am_
khalu siddhim �adar�atprak�a�say�am �asa mah�ıbhr
_tam_
prati
11 The large number of S�akta goddesses worshipped in the one thousand vasesfor the ‘consecration for victory’ ( jay�abhis
_ekah
_) taught in chapter 27 are drawn
from various parts of the Kaula Kubjik�amata (14.77–79, 81–85, 87, 91; 15.6–7, 20,22, 27, 30, 48; 9.3c–4, 5, 6; 10.120c–123, 124c–127; 21.16–20b; 22.16A, 1–9, 10–25–end; 2.59; 14.75, 77–79, 81, 84–85, 87, 91, etc.). But the procedure also includesSaiddh�antika, Daks
_in_a and Vaidika elements.
12 The Sanskrit part of the inscription refers to Sivakaivalya as the Guru andHotar of Jayavarman II (v. 25: jayavarmmamah�ıbhr
_to . . . s�ast�a=kavir . . .
sivakaivalya iti; v. 27bc: asmai/hotre). The Khmer refers to him as the Guru and
Purohita of the king (C ll. 61–62: ste _n a~n sivakaivalya ta aji pr�aj~na j�a guru j�apurohita ta vrah
_p�ada paramesvara).
236 ALEXIS SANDERSON
27 sa bh�udharen_�anumato ’grajanm�a
sas�adhan�am_siddhim adiks
_ad asmai
hotre hitaik�antamanah_prasattim
_sam_bibhrate dh�amavibr
_m_han�aya
28 s�astram_siraschedavin�asikh�akhyam
_sam_mohan�am�api nayottar�akhyam
tat tumvuror vaktracatus_kam asya
siddhyeva vipras samadarsayat sah_
29 dvijas samuddhr_tya sa s�astras�aram
_rahasyakausalyadhiy�a sayatnah_siddh�ır vvahant�ıh
_kila devar�aj�a-
bhikhy�am_
vidadhre bhuvanarddhivr_ddhyai
30 sa bh�udharendras sahavipravaryyastasmin vidhau dh�amanidh�anahetauv�ıt�antar�ayam
_bhuvanoday�aya
niyojay�am �asa mun�ısvaran tam
31 tanm�atr_vam
_se yatayas striyo v�a
j�at�a + + + tra niyuktabh�av�ah_tady�ajak�as syur na katha~ncid anya
iti ks_it�ındradvijakalpan�as�ıt
. . .
61 samadhikadhis_an_�as te s�urivaryy�as tad�a tair
dharan_ipatibhir abhyarn
_n_�arhan
_�abhyarhan
_�ıy�ah
_nagaranihitasam_sth�a devar�ajasya n�anye
sayamaniyamayatn�ah_pratyaha~n cakrur arcc�am
K. 235, 25–31, 61
King Jayavarman, who had made his residence on the summit of MountMahendra,13 had as his Guru a poet called Sivakaivalya, whose feet hadbeen honoured by [contact with] the heads of [prostrating] �Aryas. Hir-
an_yad�ama, an excellent brahmin, like Brahm�a himself in his great wis-
dom, being moved by compassion came and with due respect revealedto the king a Siddhi which no other had attained. To increase [the
king’s] splendour this brahmin, with the king’s permission, taught theSiddhi and the means of achieving it to that offerer of the [king’s] sacri-fices, [knowing that he was one] whose tranquil mind was devotedentirely to [his monarch’s] welfare. The Brahman revealed to him as
though by means of [this] Siddhi the four faces of Tumburu that are the[V�ama scriptures] Sirascheda, Vin�asikha, Sam
_moha and Nayottara, and
in order to increase the prosperity of the realm he carefully extracted the
essence of [those] texts through his mastery of the esoteric [teachings]and [with it] established the Siddhis that bear the name Devar�aja. Thenthe king with [the support of ] this excellent Brahman appointed [Siva-
kaivalya,] this lord among sages, to officiate in this ritual that is thecause of the treasure of power, in order that the realm should prosperwithout impediments. The king and the foremost of brahmins provided
13 Phnom Kulen.
237SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
that ascetics or women born in the latter’s maternal lineage, and no oth-
ers under any circumstances, should be appointed to this . . . and per-form its worship.
. . . . . . . . . . . . . . . . . .
Excellent scholars of the highest intelligence, settled by these kings inthe capital because they wished to have them nearby to venerate themas they deserved, they and they alone, performed the daily service of
the Devar�aja, zealously maintaining the major and minor restraints[of the ascetic’s discipline].
We see similar cases of regularization of rites of royal protection inour evidence for the Buddhist Way of Mantras. The Rgya gar chos’byu _n, the Tibetan history of Indian Buddhism completed by T�aran�athain AD 1608, reports that in order to protect his dynasty, expand its rule,
and spread the Buddhist religion the P�ala king Dharmap�ala (r. c. 775–812) had a fire-sacrifice performed regularly for many years by Tan-tric officiants under the direction of his Guru Buddhaj~n�anap�ada atan overall cost of 902,000 tolas of silver.14
An inscription of the reign of Jayavarman V (r. c. 968–c. 1000/1)reveals a similar arrangement in the Khmer court of Angkor. It tellsus that one K�ırtipan
_d_ita, a Mah�ay�anist scholar and adept of the
Buddhist Yogatantras, who had been adopted by the royal family astheir Guru, was frequently engaged by the king to perform apotro-paic, restorative and aggressive Mantra rituals within the royal pal-ace for the protection of his kingdom.15
4. SAIVA OFFICIANTS IN THE ROLE OF THER�AJAPUROHITA
Even in the case of the hereditary Khmer priests of the Kamrate_nJagat ta R�aja there is no reason to think that they were R�ajapuro-hitas in the narrow sense of the term, that is to say, personal chap-lains performing the whole repertoire of ritual duties, namely
14 For these reports see Chimpa and Chattopadhyaya (1970, pp. 274, 278–279).At present one tola (Skt. tul�a ) is approx. 11.7 grammes. By that standard theexpense said to have been incurred would have been that of 10,553.4 kg. of silver.
15 K. 111, 36: r�as_t_raman
_d_alaraks
_�artham
_satkr
_ty�ayu _nkta yan nr
_pah_/*mandir�abhyan-
tare (corr.: man_d_ir�abhyantare Ep.) bh�ıks
_n_am_�s�antipus
_t_y�adikarmmasu ‘In order to pro-
tect his realm the king bestowed honours on him and frequently engaged him withinthe palace to perform rituals for the quelling of dangers, the restoration of health
and the rest’. The ‘rest’ I presume to be abhic�arah_, that is to say rituals for the harming
of enemies. For evidence that K�ırtipan_d_ita was an adept of the Yogatantras see
Sanderson, 2005, p. 427, n. 284.
238 ALEXIS SANDERSON
(1) rituals to ward off dangers and ills of every kind from the kingand his kingdom (s�antikam
_karma), some of them simple rites to
protect the king’s person to be performed at various times everyday, others much more elaborate ceremonies to be performed peri-odically, (2) rituals to restore his health and vigour (paus
_t_ikam
_karma), (3) rituals to harm his enemies (�abhic�arikam_
karma), (4) theregular and occasional rituals (nityam
_karma and naimittikam
_karma)
required of the king,16 (5) reparatory rites (pr�aya�scitt�ıyam_karma),
and (6) postmortuary rites (aurdhvadehikam_karma) when the king or
any other member of the royal family dies.17
4.1 The Netratantra
For unambiguous evidence of Saiva Gurus tied to the service ofkings in that sense we are forced to turn from inscriptions recordingevents to a scriptural text regulating practice. This is the SaivaNetratantra, a work of approximately 1,300 stanzas that sets out rit-ual observances based on the propitiation of the deity Amr
_tesvara
(/Amr_tesa), also known as Amr
_tesabhairava, and/or his consort
Amr_talaks
_m�ı.18
It has come down to us in Kashmirian manuscripts with a learnedcommentary written from the non-dualistic Saiva point of view bythe Kashmirian Ks
_emar�aja in the early part of the eleventh century
and it was published therewith on the basis of two of these manu-scripts in the Kashmir Series of Texts and Studies in 1926 and 1939.19
16 The daily, periodic and occasional religious duties of the king and his per-sonal chaplain are set out in the Atharvavedaparis
_t_a. See also Vis
_n_udharmottara 2,
Adhy�ayas 5, 18–23, 132–144, 151–162, 176–177 (see n. 68 below for a listing ofthe rituals covered); N�ılamata 810–848; �Adipur�an
_a-Tithikr
_tya ll. 2618–3010;
Br_hatsam
_hit�a, Adhy�ayas 42 (Indra festival) and 43 (N�ır�ajanas�anti).
17 For this classification see Atharvavedaparisis_t_a 3.1.10: yasy�anyakulopayuktah
_purodh�ah_
s�antikapaus_t_ikapr�ayascitt�ıy�abhic�arikanaimittikaurdhvadehik�any atharvavi-
hit�ani karm�an_i kury�at.
18 Amr_tesvara is also known as Mr
_tyunjaya/Mr
_tyujit and as Netran�atha. The Aisa
form Amr_t�ısa (Amr
_t’�ısa) is also attested. The name Amr
_talaks
_m�ı is found not in the
Netra itself, where she is simply Sr�ı/Laks_m�ı, but in the ritual manuals based on that
text. Amr_tesvara and Amr
_talaks
_m�ı may be worshipped independently or as a pair.
19 Hitherto the only substantial scholarly attention paid to this work as a wholehas been an account of its contents published by Brunner (1974). While generallyaccurate that is a summary rather than an analysis and it is one that does not rec-
ognize what I identify as the distinctive and pervasive character of this text,namely that it envisages atypical Saiva officiants operating outside their traditionalterritory in that of the king’s chaplain.
239SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Its high standing in Kashmir is indicated by the composition andpreservation of this commentary, by the fact that the cult of its dei-ties, taught only in the Netratantra, is one of the two principal basesof the Saivism of the Kashmirian ritual manuals in use until recenttimes,20 by the survival of three previously unidentified images ofAmr
_tesvara and his consort Amr
_talaks
_m�ı in the small corpus of
known non-Buddhist Kashmirian bronzes,21 and by the fact that a
20 The other is that of Svacchandabhairava taught in the Svacchandatantra. In theKashmirian Saiva manuals of ritual the principal deities (mantracakram) are gener-
ally Svacchandabhairava (Sakalabhat_t_�araka [=Aghora] and Nis
_kalabhat
_t_�araka)
with Aghoresvar�ı and Amr_tesvarabhairava with Amr
_talaks
_m�ı at the centre of
the Y�aga surrounded by the Bhairavas of the Svacchandatantra. See, e.g.,Kal�ad�ıks
_�avidhi ff. 25v8–30v8; Agnik�aryapaddhati A f. 16r; and Sivanirv�an
_avidhi pp.
257, l.12–263, l. 11. The Kashmirian digest Nity�adisam_grahapaddhati, compiled by
R�aj�anaka Taks_akavarta at an unknown time after the completion of the Somasam-
bhupaddhati in AD 1095/6, that being the latest of the datable works cited by him,
distinguishes among the Saiva initiates he is addressing between followers of theSiddh�anta, followers of the Netratantra, and followers of the Svacchandatantra.
21 Two have been reproduced in Pal, 1975, plates 6 and 7. He assigns the first
to‘‘Kashmir or Afghanistan’’ and the tenth to eleventh century and the second to‘‘Kashmir or Himachal Pradesh (?)’’ and the tenth century. The first has also beenpublished by Reedy, who assigns it to Kashmir and the same period (1997, K85).
A third found its way to a Buddhist temple in Ladakh and has been reproducedin Snellgrove and Skorupski, 1977, vol. 2, p. 77, Fig. 66. These scholars were notaware of the identity of the image. Nonetheless, Pal, followed by Reedy, asserts
that it is ‘‘Um�a-Mahesvara’’, and Snellgrove and Skorupski that it is Vis_n_u with
Laks_m�ı.
According to the visualization-texts (dhy�anam) of these deities (Netra 3.17-23b;18.63–69), to which these three bronzes conform precisely, Amr
_tesvara is crowned,
white, one-faced, three-eyed, and four-armed, sitting on a white lotus at the centreof a lunar disc. In the proper right of his two inner hands he holds a vase of nec-tar at his heart and a full moon held at head height in the left, the upper arm hor-
izontal and the forearm vertical. The outer right and left hands show the gesturesof generosity and protection. The latter is invisible behind Amr
_talaks
_m�ı’s back in
the bronzes. Amr_talaks
_m�ı has the same appearance except that she carries the dis-
cus and the conch rather than the vase and moon in her inner right and left hands.Her gesture of protection is visible in the bronzes. She sits in Amr
_tesvara’s lap, on
his left thigh, and, in the bronzes is considerably smaller than her consort.
Among the other known non-Buddhist bronzes from Kashmir I am aware ofonly two others that belong to the domain of the Saiva Mantram�arga. Both areimages of the K�al�ıkula’s goddess Siddhalaks
_m�ı, identified in Sanderson (1990). No
bronze of the Siddh�anta’s Sad�asiva or the Daks_in_a’s Svacchanda has come to
light, though there are some modern paintings of the latter. The remaining non-Buddhist bronzes are P�ancar�atrika images of Vis
_n_u and images of Siva and other
gods proper to the domain of regular brahmanical observance.
240 ALEXIS SANDERSON
visualization verse for these deities recited in the Saiva rituals22 wasgiven pride of place in the non-Saiva fire-sacrifice of the Kashmirianbrahmins, being recited before pouring the oblations that accompanythe recitation of the Satarudriya of the K�at
_haka Yajurveda, the first in
a series of five Vedic hymns to Rudra (the rudrapa~ncakam).23
The Netratantra was not limited in its distribution and influenceto Kashmir. We have a Nepalese manuscript of the text from thebeginning of the thirteenth century;24 we have manuscripts in thesame region of two texts, one of the early thirteenth century andthe other probably so, that set out the procedures of its initiation cer-emony and of the regular postinitiatory worship of its deities; andthe inclusion of their worship in larger ritual contexts in the anony-mous manuals in Newari and Sanskrit used by the Tantric officiantsof the Kathmandu valley shows that the cult became and remainedan integral part of Newar Saivism. Thus in the autumnal Navar�atraceremonies Amr
_tesabhairava and his retinue are the deities of the
vase-worship (kalasap�uj�a) at the beginning of the installation of theroyal sword (khad
_gasth�apanavidhih
_) on the eighth day (mah�as
_t_am�ı ).25
The Nepalese texts that set out the procedures for initiation andsubsequent regular worship reveal that the cult was practised in theroyal family. The text on initiation, the Amr
_te�svarad�ıks
_�avidhi, envis-
ages no initiand but the king, since when it turns to the duties ofthe initiand on the day after the ceremony it requires him to returnto the Guru in a full military parade accompanied by his minis-
22 See, e.g., Sivanirvan_�avidhi p. 261: devam
_sudh�akalasasomakaram
_trinetram
_pa-
dm�asanam_ca varad�abhayadam
_susubhram/sa _nkh�abhay�abjavarabh�us
_itay�a ca devy�a v�a-
me ’ _nkitam_samanabha _ngaharam
_nam�ami ‘I prostrate myself to the god who frees
us from the torture of death, three-eyed, perfectly white, holding a vase of nectarand the moon, [showing the gestures of ] bestowing boons and protection, seatedon a lotus, marked on his left by the goddess adorned with a conch, [the gesture
of ] protection, a lotus and [the gesture of ] bestowing boons’.23 See Vedakalpadruma pp. 15–16.24 NAK MS 1-285: ‘Amr
_tesatantram’. In the colophons of this manuscript the
work is referred to as the Mr_tyujidamr
_t�ısavidh�ana. In Kashmirian sources it is also
known as the Mr_tyu~njaya or Mr
_tyujit, often with the honorific -bhat
_t_�araka; see,
e.g., Ks_emar�aja, Sivas�utravimarsin�ı ad 1.1, 1.13, 1.19, 3.16, etc. In citations of the
Netra below this manuscript will be referred to as N and the Kashmirian edition
as Ed. For the date of the manuscript see n. 28 below.25 Navar�atrap�uj�a f. 2r4–v8.
241SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
ters.26 Its author, Visvesvara,27 may well be the person of that namereported by the scribe of our Nepalese manuscript of the Netratantraas having commissioned the copying, which was completed in Febru-ary/March of AD 1220.28 If so, then Visvesvara may have produced histreatise in the context of initiation given to the Nepalese kings Arimalla(b. 1153, r. 1200/01–1216) and/or his son Abhayamalla (b. 1183, r.1216–1255).29 Be that as it may, it is certain that the latter received thisinitiation, since it was he who, at some time before his accession, com-posed our Nepalese ritual manual on the regular postiniatory worship.30
The fact that evidence of the text and its traditions is known fromKashmir and Nepal does not, of course, reveal its provenance; andin general one would not expect to be able to determine that fromthe text itself, since such works tend to lack the references to locat-able realities that allow us to draw conclusions of this kind. TheNetratantra, however, is exceptional in this respect. I propose that itcontains evidence sufficient to justify the conclusion that it was
26 For this reference to the newly initiated king’s procession in full militaryparade see Sanderson, forthcoming.
27 Amr_tesvarad�ıks
_�avidhi f. 19r1: bhairavasy�amr
_t�ısasya d�ıks
_�at_ippan
_akam
_sphut
_am/
visvesvaren_a racitam
_sajjan�as carcayantv idam.
28 Amr_tesatantra f. 89v4–5: sam�aptam idam
_mr_tyujidamr
_t�ısavidh�anam
_sam_p�urn
_am
iti subham: sam_vat 320 caitra sudi 9 sanidine ++ visvesvaralikh�apitam idam
_pusta-
kam_: pam
_d_itak�ırttidharalikhitam
_may�a. That this text is Nepalese is not certain, but
it is probable. The account of initiation is supplemented by a passage on the S�aktavedhad�ıks
_�a incorporated from the Kubjik�amata, a text of central importance in the
S�akta-Saiva tradition of the Newars (f. 16r11–v11 10.83–92b, 94, 96–98 and 100–107).29 For these dates see Petech (1958, pp. 76 and 82).30 See Amr
_tesvarap�uj�a f. 7v2–3: sr�ıdev�abhayamallena sad�ac�aryopadesin�a=sr�ımr
_tyu-
~njayadevasya nityap�uj�avidhih_kr_tah.
_ity amr
_tesvarap�ujanam
_sam�aptam (see also Petech,
1958, p. 89, citing its palm-leaf exemplar NAKMS 1–1365.5); Amr_tesvarap�uj�a f. 1v5–6
(v. 2): p�ıy�us_asindhulahar�ısatasiktapadmamadhye sphurattuhinarasmimar�ıcisubhram=
natv�a mahesam amalam_
kamal�asah�ayam abhyarcanam_
vitanute ’bhayamalladevah_.
NAK MS 1–1365.5 was copied on June 8, 1216 just before the end of the reign of Abha-yamalla’s father Arimalla (Petech, 1958, p. 84). Other relevant Nepalese manu-scripts are Amr
_tesvarap�uj�agnik�aryavidh�ana, Amr
_tas�uryap�uj�avidhi with drawings of the
deities, ‘P�uj�ak�an_d_a’, which contains inter alia an Amr
_tabhairav�arcanavidhi penned in
AD 1277/8, an Amr_t�ısabhairavabhat
_t_�arak�ahnikavidhi, and an Amr
_tas�ury�arcanavidhi.
Amr_tas�urya is Amr
_tesvara in the form of the Sun God, the worship of this ectype
being prescribed before that of the deity proper, as was standard procedure in the
Siddh�anta, whose Paddhatis prescribe a cult of Sivas�urya before that of Siva. Thereis no such preliminary in the Netra itself. The icon of Amr
_tas�urya created for this pur-
pose was the three-faced, eight-armed variant of the Sun God holding the weapons ofthe eight Lokap�alas taught (13.21c–25b) in the section of that text devoted to the ico-
nography of various non-Saiva deities, one of three forms of that God taught there.
242 ALEXIS SANDERSON
indeed produced in one of the two places in which we see evidenceof its presence, namely Kashmir, and that it was composed there be-tween c. 700 and c. 850, probably towards the end of that period.31
With the exception of an insertion of 95 verses seen in the Nepalesemanuscript32 the only major difference between the transmissions of theNetratantra in that and the Kashmirian manuscripts is that the deviationsfrom strictly grammatical Sanskrit that abound in the early Saiva scrip-tures are much fewer in the latter, most of whose divergent readings arebest explained as the result of rephrasing to remove such anomalies. In thecitations that follow I have therefore privileged the readings of the Nepa-lesemanuscript (N) as evidence of an earlier state of theKashmirian text.33
31 My reasons for proposing this provenance and date are set out in the Appendix.32 This insertion (ff. 47r1–53r2) is placed after 18.3 of the published edition. The
Nepalese manuscript treats this as the remainder of the 18th chapter (18.4–99). It
then gives the whole of the edition’s chapter 18, so repeating 18.1–3, as its chapter19, and so on to the end, so that it has 23 chapters rather than the edition’s 22. Thesubject-matter of the additional verses is hostile visualization rituals and fire-sacrificesin which the deity takes the form of Mah�abhairava. It has drawn on the Svacchanda:
18.62c–68b 9.62–67; 18.69–71a 9.71–73a; 18.72ab 9.76ab; 18.73–78 6.72c–78b;18.79–85a 6.85c–91c; 18.85b–87 6.92–94; and 19.92–95a 6.68c–71c.
33 Particularly notable among the Aisa usages accepted in my edition of the pas-
sages cited below is the use of genitives, instrumentals, locatives and ablatives/dativesplural side by side in a single construction without difference of meaning: e.g. pis�acaisc�apy anekasah
_=brahmaraks
_agrah�adibhyah
_kot_iso yadi mudrit�ah
_2.14bcd; nr
_p�an_�am_nr
_papatn�ın�am
_tatsut�an�am
_dvij�adis
_u 15.20cd; nr
_patau tatsut�an�am
_18.112ab; duh
_sva-
pnair m�atar�ıs_u ca 19.98d; gos
_u br�ahman
_araks
_�artham �atmanah
_svajanes
_u ca 19.104ab;
s�aly�adis_u ca sasyes
_u phalam�ulodakena ca=durbhiks
_avy�adhik�aryes
_u utp�atais c�apy
anuttamaih_19.108. This licence surely reflects the development of the case endings in
Middle Indo-Aryan as witnessed in the Apabhram_sa stage, in which the locative and
instrumental plural have merged, as have the dative, genitive and ablative plural (Ta-gare 1987, pp. 141–150). Other typical features of the MIA-influenced register of San-
skrit seen here are the extended stems ofm�atr�ıbhir 2.13c,m�atar�ıs_u 19.98d, d
_�avy�a 2.13d,
bh�ubhr_t�an�am
_12.7d, aris
_t_acihnit�atm�ano (nom. sg.; conj.; cf. Picumata f. 238v [52.15a]:
s�adhakas tu mah�atm�ano) 19.107a, and pasaves_u (conj.; cf. Picumata f. 224r [46.34b]:
pasav�an�am_) 19.120c; the contracted stems of sreyam in 19.105a and digv�asam in
13.10c; -eta for -ayeta in the optative in abhimantreta 19.90c, 19.117d, and 19.119a;the dispensability of final t=d revealed in ks
_ipe ’nale (18.118b) for ks
_iped anale; the loss
of declension in the numerals (catuh_for catv�arah
_in 18.121a); the neuter plural siddh�ıni
in 18.79c and 19.115c; so for sa in 16.113c; the non-causative base in the pseudo-causa-tive abhis
_i~ncayet in 19.109d; and ’yeta for -ayeta in p�ujyeta in 19.104c. Untypical is the
use of pratis_t_h�apyah
_in an active sense governing an object in the accusative in the
phrase pratis_t_h�apyah
_. . . guruh
_. . . bh�ıs
_an_am_r�upam ‘the Guru should install the fright-
ening [Bhairava] form’ in 18.119–120. For the termAisa (‘uttered byGod’) in this con-text see the references in Sanderson, 2002, n. 27. For the reader’s convenience I have
retained the chapter and verse numbers of the published edition (Ed.).
243SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
4.2 The Netratantra’s Saiva Officiant
Now the Saiva officiant in our Netratantra has very much the charac-ter of a personal chaplain to the king. He is presented as the per-former of rites for the protection and prosperity of all members ofsociety, but this wider constituency is generally mentioned only afterthe text has specified the king, his wives and their children, who arethe principal intended beneficiaries and in many cases the only ones.After various preliminaries the text introduces its subject as follows:
2.13 bh�utayaks_agrahonm�adas�akin�ıyogin�ıgan
_aih_bhagin�ırudram�atr�ıbhir d
_�avy�ad
_�amarik�adibhih
_14 r�upik�abhir apasm�araih_pis�acais c�apy aneka�sah
_brahmaraks_agrah�adibhyah
_kot_iso yadi mudrit�ah
_15 apamr_tyubhir �akr�ant�ah
_k�alap�asair jigh�am
_sit�ah
_r�aj�ano r�ajatanay�a r�ajapatnyo hy anekasah_16 vipr�adipr�an
_in_ah_sarve sarvados
_abhay�ardit�ah
_yena vai smr_tam�atren
_a mucyante tad brav�ımi te
13cm�atr�ıbhir corr. :m�atr�ıbhiN :m�atr�adi Ed. : 13d d_�avy�ad
_�amarik�adibhih
_N : d_�av�ıd
_�amarik�adibhih
_Ed. 14b anekasah
_Ed. : anekasaih
_N 14c
raks_agrah�adibhyah
_N : raks
_ograh�adyais ca Ed. 15b jigh�am
_sit�ah
_Ed. :
jigh�am_sat�aN
I shall tell you that [Mantra] by whose mere remembrance [the Guru]can free kings, their wives and children, and [indeed] all creatures begin-ning with learned brahmins, if they have been dominated by any of the
countless hordes of [possessing spirits:] Bh�utas, Yaks_as, Unm�adas, S�a-
kin�ıs, Yogin�ıs, Bhagin�ıs, Rudram�atars, D_�av�ıs, D
_�amar�ıs, R�upik�as, Ap-
asm�aras, Pis�acas, Brahmaraks_as, Grahas and the like, if they have been
attacked by Apamr_tyus, if they are in danger of falling victim to the
snares of Death, or are suffering from the danger of any ill.
From this point until the end of the sixth of the work’s twenty-twochapters we are told the Mantra, its worship, the ceremony of initia-tion to that worship and then a number of procedures for its applica-tion. The benefits specified are the restoring of physical vitality(pus
_t_ih_) (3.77), longevity (p�urn
_am �ayuh
_) (3.78), wealth (sr�ıh
_) (3.78),
rule (r�ajyam) (3.79), good health (3.79), rescuing the dying fromdeath (mr
_ty�utt�aran
_am) (6.9–11b), the warding off of all ills
(mah�as�antih_) (6.11c–13b) and the banishing of fevers ( jvaran�asah
_)
(6.15cd). The last half of the sixth chapter sets out a procedure forthe protection of the king (r�ajaraks
_�avidh�anam). A Yantra, that is to
say a diagram on birch bark inscribed with the Mantra and the nameof the beneficiary, is worshipped; a fire-sacrifice is performed; and
244 ALEXIS SANDERSON
the king is consecrated from a vase into which the Mantra has beeninfused. The ceremony is to destroy the pride of his enemies when hegoes into battle, to free him from all illnesses, and to bestow on himthe highest sovereignty.34
Chapters 7 and 8 outline subtler, meditational practices, but thecontext is unchanged, as it is in chapters 9 to 14, which teach thatthe Mantra is absolutely universal in its range and can therefore beused in conjunction with any deity and its retinue, not merely withAmr
_tesvara as visualized in the cult proper. So we are given the
appropriate substitute deity-visualizations for the four main divi-sions of the Saiva Mantram�arga, namely the Siddh�anta, the V�ama,the Daks
_in_a and the Kaula, and then, beyond Saivism, for the cults
of Vis_n_u, the Sun, Siva in the lay context, Brahm�a, the Buddha,
Skanda, K�ama, the Moon, Gan_esa, the Lokap�alas, ‘‘and all other
deities’’. The context, though mostly only implicit in this long sec-tion, surfaces in the chapter on the Kaula modification. There weare told that the eight Mothers (who form the circuit of Amr
_tesvara
in this case) must be worshipped with abundant offerings to bringabout the warding off of ills (s�antih
_), but with exceptional lavish-
ness when the beneficiary is the king, because it is by their favourthat he will enjoy untroubled rule:
12.6 pa _nktisth�a v�a yajed dev�ıh_sarv�abh�ıs
_t_aphalaprad�ah
_sarves_�am_caiva s�antyartham
_pr�an
_in�am
_bh�utim icchat�a
7 bh�uriy�agena yas_t_avy�a yath�ak�am�anur�upatah
_vi�ses_en_a tu yas
_t_avy�a bh�ubhr
_t�an�am
_tu daisikaih
_8 �as�am eva pras�adena r�ajyam_nihatakan
_t_akam
bhu~njate sarvar�aj�anah_subhag�a hy avan�ıtale
6c sarves_�am_caiva N : sarves
_�am eva Ed. 7b yath�ak�am�anur�upatah
_Ed. :
yath�akarm�anur�upatah_
N 7c vises_en_a tu N : vises
_�ad devi Ed. 7d bh�u-
bhr_t�an�am
_tu N : bh�ubhr
_t�am api Ed.
Alternatively [the officiant] should worship these goddesses thatbestow all the benefits one desires in a row [rather than a circle], inorder to bring about the warding off of ills from all, desiring the pros-
perity of [all] creatures. O Goddess, officiants should make lavish
34 Netra 6.35c–50, beginning r�ajaraks_�avidh�anam
_tu *bh�ubhr
_t�an�am
_(conj. [cf. 12.7d
below]: bh�ubhr_t�ama N: bh�ubhr
_t�am_
tu Ed.) prak�a�sayet=sam_gr�amak�ale varadam
_ri-
pudarp�apaham_
bhavet; 6.40: sarvama _ngalaghos_e_na sirasi hy abhis
_ecayet=sa mucyeta
na *sam_deho (N [Aisa sandhi] : sam
_dehah
_Ed.) sarvavy�adhiprap�ıd
_itah
_; 6.46d:
*bh�ubhr_to (N : bh�ubhr
_t�am_Ed.) r�ajyam uttamam.
245SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
offerings to them in accordance with the benefit desired, but especiallyon behalf of kings. [For] if a kingdom is free of enemies and fortunatekings enjoy [their sovereignty] on earth, it is by their favour.35
In the 15th chapter the text returns to the detail of protecting the royalhousehold (r�ajaraks
_�a), and this theme continues in the chapters that
follow. The Yantra taught in the 17th will bestow victory on kingswho are under attack from beyond their borders and should be used atall times to protect the king’s wives, his sons, brahmins and others.36
Chapter 18 teaches the procedures for the worship of Sr�ı(Amr
_talaks
_m�ı) without her consort. We are told:
sam_gr�amak�ale dhy�atavy�a khad
_gapatralat�asthit�a
18.86 jayam_
prayacchate ’va�syam_
ripudarp�apah�a bhavet
sam_gr�am�agre sad�a y�ajy�a parar�as
_t_rajig�ıs
_un_�a
87 ava�syam_jayam �apnoti devadevy�ah
_pras�adatah
_86a prayacchate ’va�syam
_N : prayacchate tasya Ed. 86d parar�as
_t_raji-
g�ıs_un_�a Ed. : ripudarp�apah�a bhavet N (dittography of 86b)
The Supreme Goddess should be visualized on the blade of the [king’s]sword at the time of battle. She bestows certain victory, crushing thepride of his enemies. A king who desires to conquer another kingdom
should always sponsor her worship before the battle. By her favour heis bound to win.37
The long 19th chapter details procedures for countering possessionby various classes of being. Here the Guru’s role is portrayed almostexclusively as that of priest to the royal family. He is told that hemay use his knowledge to help his own family and pupils but other-wise he is to do these rites only for the king, his queens and theirchildren. And there is a further restriction that emphasizes his tie to
35 According to Kalhan_a Circles of the Mothers (m�atr
_cakr�an
_i) were set-up at the
passes leading into the kingdom of Kashmir, no doubt as the guardians of the realm.R�ajataran_ gin
_�ı 1.122: dv�ar�adis
_u prade�ses
_u prabh�avogr�an
_y udagray�a=�ıs�anadevy�a
tatpatny�a m�atr_cakr�a _ni cakrire ‘�Is�anadev�ı, the wife of this [Jalauka, son of Asoka],
set-up Circles of the Mothers, terrible in their power, at the passes and other places’.36 17.5c–7: sarvas�antipradam
_cakram
_pus_t_isaubh�agyad�ayakam=�ayurv�ıryapradam
_*pun_yam
_(N : caiva Ed.) jvararogavin�asanam=parar�as
_t_ravibh�ıt�an�am
_nr_p�an_�am_
vija-
y�avaham=r�ajastr�ın_�am_*sut�an�am
_ca (N : tatsut�an�am
_Ed.) vipr�ad�ın�am
_ca sarva�sah
_=
raks_�a hy es
_�a prakartavy�a sarvopadrava*n�asan�ı (N : n�asin�ı Ed.)
37 A ritual for the empowering of the sword with the Mantra of K�al�ı before the
king goes into battle is part of the repertoire of the Paippal�ada AtharvavedinR�ajapurohitas of Orissa seen in the Paippal�adavas�adis
_at_karmapaddhati, pp. 71–72
( �A _ngirase K�alik�amantravidh�anam).
246 ALEXIS SANDERSON
them: he may exorcise his own family and pupils only if they aredevout, whereas he is obliged to exorcise the royal family regardlessof their personal qualities, for the king, we are told, is the head ofall the religious orders of life (sarv�a�sramagurutv�at).38
The chapter continues with services of protection that the offici-ant is to render to the king during the course of each day:
mukhe praks_�alite nityam
_tilakam
_svetacandanam
19.89 sapt�abhimantritam_
k�aryam_
dos_anivr
_ttaye tad�a
sn�anak�ale tath�a k�aryam_tilakam
_�svetabhasman�a
sam�alabhanapus_pam
_v�a t�amb�ulam
_v�abhimantritam
90 d�ıyate yasya tasyaiva na him_sant�ıha him
_sak�ah
_88d tilakam
_svetacandanam
_N : tilakah
_svetabhasman�a Ed. 89a sapt�a-
bhimantritam_
k�aryam_
N : sapt�abhimantritah_
k�aryo Ed. 89b dos_anivr
_-
ttaye tad�a N : m�atr_dos_anivr
_ttaye Ed. 89cd This line is in N only.39
89e sam�alabhana N : sam�alambhana Ed. 89f t�amb�ulam_v�abhimantritam
N : t�amb�ulen�abhimantritam Ed. 90a tasyaiva Ed. : tasyeha N
Whenever the [king’s] face has been washed [the officiant] should give[him] a forehead mark of white sandal paste that has been empowered
by reciting [the Mantra of Amr_tesvara] over it seven times, in order to
prevent assaults [by the Mothers40] at that time. He should also give[him] a forehead mark of white ash at the time of bathing. Harmful[spirits] will not attack [that king] to whom he gives betel or a flower
with fragrant powder.
38 19.86 bhakt�an�am_
svasut�an�am_
ca svad�ar�an_�am_
*tu (N : ca Ed.) k�arayet=svasi-s_y�a _n�am
_*tu (N : ca Ed.) bhakt�an�am
_n�anyath�a tu prayojayet/87 sarv�a sramagurutv�ac
ca bhupat�ın�am_ca sarvad�a=tatsut�an�am
_ca patn�ın�am
_*kartavyam
_tu hit�arthin�a (N : ka-
rtavyo hitam icchat�a Ed.) ‘He may do it for his sons, wife and his pupils only if theyare devoted [to Siva]. Otherwise he must not employ [this procedure]. [But] if hedesires the welfare [of all] he must always do it for kings, for the king is the headof all the orders, and likewise for the king’s sons and wives’. For this reference to
the king as the patron (-guruh_) of the orders (�asrama-), of the caste-classes (varn
_a-)
and of both (varn_�asrama-) cf. Kath�asarits�agara 12.6.85 (varn
_�asramaguruh
_) and the
sources cited in Sanderson, forthcoming, especially Brahmasambhu’s Paddhati
Naimittikakarm�anusam_dh�ana, in which it is said that the purpose of the Saiva modi-
fication of the royal consecration following the Saiva initiation of the king is to quali-fy him to hold office as the patron of the caste-classes and orders (f. 74v1 [4.118]:varn
_n_�an�am �a�sram�an
_�a~n ca gurubh�av�aya bh�upateh
_=yo bhis
_ekavidhih
_sopi procyate
d�ıks_it�atmanah
_).
39 N shows that the text seen in the edition has been corrupted by an eyeskipfrom the sveta of svetacandanam in 88b to the �sveta of svetabhasman�a in 89d. That
19.89 has three lines here is simply because I have kept the numeration of the edi-tion for the reader’s convenience.
40 That the assaults prevented are those of the Mothers is conveyed by the read-
ing m�atr_dos_anivr
_ttaye of the Kashmirian edition.
247SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
He must infuse the Mantra of Amr_tesvara into the king’s food be-
fore he eats and protect him with a visualization:
bhojanam_c�abhimantreta mantren
_�anena mantravit
19.91 ubhayoh_candrayor madhye bhu~nj�ano ’mr
_tam asnute
sarvavy�adhivinirmuktas tis_t_hate nr
_patih
_ks_itau
90c c�abhimantreta Ed. : c�abhimantryaiva N 91a candrayoh_
N : p�arsva-yoh_Ed.41
The Master of Mantras should empower [the king’s] food by recitingthis Mantra upon it. If the king eats between two [visualized] moondiscs he consumes the nectar of immortality42 and lives [long] on
earth, free of all disease.
He is also to use his art to protect the king’s person before hebegins his daily training in the arts of war:43
19.92 atha kr�ıd_anak�ales
_u gaj�asvasahites
_u ca
astrakr�ıd_�asu sarv�asu raks
_�artham
_kalasam
_yajet
93 kr�ıd_�artham
_vijay�artham
_ca raks
_�artham
_him_sak�adis
_u
yasm�ad dus_t_�as ca bahavo jigh�am
_santi nr
_pes_u ca
92b gaj�asvasahites_u em. : gaj�asvamahis
_es_u N : gaj�asvasahitasya Ed.
93a ca Ed. : tu N 93d nr_pes_u ca N : nr
_p�adikam Ed.
41 N’s reading candrayoh_has more attack. It is also supported by Ks
_emar�aja ad
loc.: anena mantren_a proktadr
_s�a sam
_put_�ık�arayukty�a dhy�ato ’bhimantritas candrad-
vayamadhyasthitam_
bhojanam_bhu~nj�ano ’mr
_tam asnute ’mr
_tatvam eti nr
_patih
_‘If the
king is empowered with this Mantra while visualized enclosed by it on either side in
the manner already taught and eats [his] food between two moons [likewise visual-ized on either side of it] he obtains ambrosia, i.e. becomes immortal’.
42 Literally ‘‘he obtains ambrosia’’. The point of the visualization is that the
moon is the embodiment of nectar (amr_tam), as is Amr
_tes�vara, who is visualized
at the centre of a lunar disc, holding in two of his hands a vase of nectar and alunar disc. See n. 21 above.
43 Such activities are prescribed as part of a king’s daily routine, after he has at-tended to affairs of state, heard petitions and the like, in Vis
_n_udharmottara 2.151.31–
32b: mantram_kr_tv�a tatah
_kury�ad vy�ay�amam
_pr_thiv�ıpatih
_=rathe n�age tathaiv�asve
khad_ge dhanus
_i c�apy atha=anyes
_u caiva sastres
_u niyuddhes
_u tatah
_param ‘When the
king has dealt with affairs of state he should exercise, by riding a chariot, an elephantand a horse, then in training bouts with sword, bow and other weapons’. Cf. 2.65.3d–4b concerning the training of the crown prince: dhanurvedam
_ca siks
_ayet=rathe ’sve;
ku~njare caiva vy�ay�amam_
k�arayetsad�a ‘[The king] should teach him the art of archeryand make him regularly exert himself in riding his chariot, his horse and his elephant’.
248 ALEXIS SANDERSON
Moreover, whenever [the king] engages in sport with elephants andhorses44 or takes part in contests with weapons [his officiant] shouldperform the vase-worship [of Amr
_tesvara] in order to protect [him.
Indeed he should do so whether the king is riding and fighting] for
recreation or for victory [in battle], in order to guard him against theharmful ones. For many are the evil [spirits] that seek to harm kings.
And he is to perform a ritual for protection in the king’s sleepingquarters when he retires for the night:
tatah_suptasya nr
_pater nidr�akalasam arcayet
19.95 raupyam aus_adhisam
_yuktam
_candan�agurucarcitam
ks_�ıren
_odakap�urn
_am_v�a yajen mr
_tyujitam
_param
96 sarvasvetopac�aren_a pus
_padh�up�arghap�ayasaih
_agre sthit�a mah�anidr�a jagatsam_mohak�arin
_�ı
97 sukh�artham_nr_pate r�atrau j�ırn
_�artham
_bhojan�adike
�arabdh�a devadevena �aj~n�a datteti bh�avayet98 tato r�atrim
_samagr�am
_tu tis
_t_hate nidray�a saha
yaks_araks
_ah_pis�ac�adyair duh
_svapnair m�atar�ıs
_u ca
99 bhayaih_sam_tr�asaduh
_khais tu muktas tis
_t_hed yath�asukham
lokap�ales_u s�astres
_u raks
_�artham
_nr_pasam
_nidhau
100 p�ujanam_c�arghapus
_p�adyaih
_kalase p�ujite sati
yasyaivam_satatam
_kury�aj j~n�anav�an daisikottamah
_101 p�urvoktam_sarvam �apnoti pr�aheti bhagav�a~n sivah
_94cd nr
_pater nidr�akalasam arcayet N : nr
_pate raks
_�artham
_kalasam
_yajet
Ed. 95a raupyam aus_adhi N : raupyam
_caus
_adhi Ed. 95b carcitam N :
lepitam Ed. 95c ks_�ıren
_odakap�urn
_am_v�a conj. : ks
_�ıras codakap�urn
_n_am_v�a
N : ks_�ıren
_a c�ambhas�a p�urn
_am_
Ed. 96c agre Ed. : gram_the N 97d
�aj~n�a datteti N : �aj~n�am_dattveti Ed. 98b tis
_t_hate N : tis
_t_hed vai Ed. 98d
duh_svapnair Ed. : dusvapne N � m�atar�ıs
_u ca N : m�atr
_sambhavaih
_Ed.
99a bhayaih_sam_tr�asa N : bhayais tattr�asa Ed.
Then he should worship a sleep-vase [to be set-up] for the king when
he sleeps. It should be silver, contain the [various protective] herbs, besmeared with sandal-paste and aloe-powder, and be filled with milk orwater. [In it] he should worship the supreme Mr
_tyujit [Amr
_tesvara]
with all his offerings white, with flowers, incense, guest-water and riceboiled with milk and sugar. [If the Lord has been worshipped in thisway45] [the Goddess] Mah�anidr�a (‘Great Sleep’) who deludes all the
world will be present [before the king]. [The officiant] should imagine
44 Literally ‘‘on occasions of sport that are accompanied by elephants andhorses’’. The reading of N adds buffaloes (mahis
_es_u) after the horses where I con-
jecture sahitesu ‘‘accompanied’’. But this is surely a corruption since buffaloes areinappropriate to the context. Ed.’s reading sahitasya supports the emendation. It
was no doubt substituted for sahites_u to improve the sense, since it is more natural
to describe the king than the occasions as accompanied by these animals.45 Ks
_emar�aja introducing 96c–97: ittham
_bhagavaty arcite.
249SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
that the God of Gods has commanded her to undertake [this task] sothat the king may sleep contented through the night and be able prop-erly to digest his food and [drink]. Then he will sleep the whole nightthrough. He will remain at ease, free of [the assaults of ] Yaks
_as,
Raks_ases, Pis�acas and the like, bad dreams, the Mothers, dangers,
and the sufferings caused by terror. When the vase has been wor-shipped [the officiant] should worship the Lokap�alas and their Weap-
ons near the king with guest-water, flowers and the rest. Lord Sivahas taught that a [king] for whom [this] learned and most excellentGuru performs these [services] attains all the [benefits] that have beenstated [in the course of this work].
Cognate rites of protection to be performed around the king’s bedare prescribed among the duties of the brahmanical royal chaplain.He is to install an image of the goddess Night in front of the king’sbed, worship it, scatter mustard seeds and sugar around the bed, givethe king a protective wrist-thread (pratisarah
_), and [a forehead-mark
of ] ash (bh�utih_), and then conduct him into the bed-chamber
(v�asagr_ham);46 and the placing of a silver sleep-vase (nidr�akalasah
_) at
the head of the royal bed is mentioned by the seventh-century poetB�an
_a in his description of the bed-chamber to which a prince and his
bride retire on the night of their wedding.
. . . sayanasirobh�agasthitena ca kr_takumudasobhena kusum�ayudhas�ah�aya-
k�ay�agatena sasineva nidr�akalasena r�ajatena vir�ajam�anam_v�asagr
_ham . . .
Hars_acarita, pp. 208–9
. . . the private apartment illuminated by a silver sleep-vase adorned
with lotuses placed at the head of the bed, as though it were themoon come to aid the flower-arrowed [God of Love] . . . .47
After speaking of these daily rites for the king’s personal protectionthe Netratantra goes on to the ceremonies that the Saiva officiantmust perform on special occasions for the more general benefit of the
46 Atharvavedapari�sis_t_a 4.3.1–4.5.16, 6.1.1–6.2.8. The same source teaches a night-
ceremony in which a lamp is to be carried three times round the king (7.1.1–11).47 According to the Kashmirian Kal�ad�ıks
_�avidhi the Guru is also to set up a vase
of this kind by the head of the initiand’s bed when he sleeps in the presence of the
deities in the hall of worship between the days of his initiation; f. 86r: nis�akalasam_c�asijaptam
_�sayy�a�sirah
_pradese pran
_avap�ujite sam
_sth�apya . . . ‘Having installed a
night-vase, empowered by reciting the Weapon [Mantra of Svacchanda] over it, on
a spot at the head of the bed, after worshipping that [spot] with OM_[thus prepar-
ing it as a throne] . . .’. This detail is not found in the Svacchanda, the text onwhich the Kal�ad�ıks
_�avidhi’s procedure for initiation is based.
250 ALEXIS SANDERSON
king and his kingdom. This part of the text begins with the rule thathe must undertake the worship of Amr
_tesvara on all such occasions:
nimittes_u ca sarves
_u amr
_tesam
_yajet sad�a
19.102 k�amar�up�ı bhaved yasm�at sarvak�am�an av�apnuy�at
101d yajet sad�a N : yajeta ca Ed. 102a k�amar�up�ı N : k�amar�upam_Ed.
� bhaved conj. : yajed B : sad�a Ed.
Since [Amr_tesvara] can take on any form at will [the officiant] should
always worship him on any festal day.48 [In this way] he will secure allthat he desires.
After this general rule the text sets out how he is to proceed in aparticular case. This is the royal festival of Indra’s pole (indro-tsavah
_, indradhvajotsavah
_) to be celebrated on the twelfth day of
the bright fortnight of the month Bh�adrapada (July/August).The procedures of the brahmanical prototype are described in
the Kashmirian Vis_n_udharmottara, Khan
_d_a 2, chapter 155.49 Ac-
cording to that account the rites start on the first day of the lightfortnight of Bh�adrapada. First the king worships Indra and hisconsort Sac�ı on Pat
_as.50 Then the pole is prepared by felling an
appropriate tree and fetched from the forest on a cart drawn by
48 I understand nimittam, literally ‘an occasion requiring’ [special worship], torefer here to all days that occasion a naimittikam
_karma in the sense of a calendrical-
ly fixed recurrent non-daily act of special worship (vi�ses_ap�uj�a). Cf. Y�aj~navalkyasmr
_ti
1.203ab: d�atavyam pratyaham p�atre nimittes_u vi�ses
_atah
_. That this is the sense is
apparent from the specific occasions that exemplify this rule in the verses 19.102c ff.49 For evidence that the Vis
_n_udharmottara originated in Kashmir or its neigh-
bourhood see the Appendix.50 A Pat
_a (Skt. pat
_ah_) is a tanka (Tibetan tha _n ka), a painting of a deity or deities
on burnished cotton cloth to which several layers of a gesso have been applied;Pi _ngal�amata (f. 27v4–6 [5.2–5]): bhogamoks
_aprasiddhyartham
_pat_am_
k�arp�asikam_varam=*ke�saj�ady anyath�a (conj. : ko�saj�ady�anyath�a Cod.) devi vipar�ıt�adis�adhane=pre-
tavastr�adikam_
�slaks_n_am_*sada�sam
_(corr.: sadasam
_Cod.) dvistriks
_�alitam=khalitam
_pin_d_itam
_mr_dyam
_�sa _nkh�adyena su�sobhane=tintad
_�ıb�ıja sam
_gr_hya susvinnam
_p�ıs_ayed
budhah_=tasyordhvam
_kharparam
_pis_t_v�a *caik�ıkr
_tv�a (corr. : cek�ıkr
_tv�a Cod.) tu marda-
yet=svacchodakena c�alod_ya tena vastram
_pralepayet=vajralepah
_smr_to hy es
_a punah
_punah_sam�acaret. When a Pat
_a is not in worship its painted surface should be con-
cealed by a layer of cloth; see op. cit. f. 29r3 (5.45c–46b): vastrair �acch�adayen nityam_sarvacitres
_u yatnatah
_=p�uj�adhy�anajapak�ale udgh�at
_ya vidhim �acaret. Tibetan practice
indicates that it was stitched on to the upper edge of the Pat_a’s cloth border and
rolled up and secured with ties at the time of worship.
251SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
cows or by men. On the eighth the king enters the city followed bythe citizens carrying fruits and wearing their best clothes. The capi-tal must be decorated with banners and flags, the royal highwaysprinkled, and the children adorned. It must be full of actors anddancers, and its deities, both public and domestic, must be wor-shipped to the accompaniment of loud music. The pole is placed onthe ground prepared for it facing east, covered with fine cloths, andworshipped until the twelfth. On the eleventh the king has fastedand held a vigil with his chaplain, his astrologer, and all the citi-zens. Dramatic spectacles must be staged all over the city duringthe night and the king must worship Indra with dance and song.On the twelfth he bathes his head and has the pole raised. He wor-ships the pole and the Pat
_as of Indra and Sac�ı with various Balis
and by means of the honouring of brahmins. The chaplain per-forms a fire-sacrifice with the Mantras of Indra and Vis
_n_u and wor-
ships Indra with dance and song. The king honours brahmins withgifts of money, particularly his chaplain and astrologer. On the fifthday of the festival the pole is dismissed. After offering reverence toit in the presence of his army he has it born away by elephants anddisposed of with the two Pat
_as into a river. The citizens celebrate
by playing in the water.51
Now the Netratantra tells us that on this occasion the Saiva offi-ciant is to worship not Indra but Amr
_tesvara as Indra:
praj�an�am_raks
_an_�arth�aya �s�al�ın�am
_sasyasampade
19.103 sutapatn�ıs_u raks
_�artham �atmano r�as
_t_ravr
_ddhaye
indrar�upam_yajet tatra vijay�artham
_nr_pasya ca
102d sasyasampade N : c�api sam_pade Ed. 103c indrar�upam
_Ed. :
indrar�up�ı N
For the protection of the [king’s] subjects, for abundant crops of rice[and other] grains, for the protection of his [king’s] sons and wives,
for the prosperity of the kingdom and the king’s victory [in war] heshould worship [Amr
_tesvara] on that [day]52 in the likeness of Indra
(indrar�upam).
51 For detailed accounts of this festival see also (1) Br_hatsam
_hit�a, Adhy�aya 42,
following Garga, and (2) Atharvavedapari�sis_t_a 19a (indramahotsavah
_). In the sec-
ond Khan_d_a of the Vis
_n_udharmottara Adhy�ayas 154–157 are devoted to it.
Adhy�aya 154 is introductory. 155 covers the procedures. 156 deals with overcom-
ing the dire consequences for the king and citizens if the pole falls or is damagedin some way. Adhy�aya 157 gives the text of the Mantra of Praise (stavamantrah
_)
that the king must recite when the Indra pole is being raised.52 Ks
_emar�aja ad 103c: tatreti naimittike indradine.
252 ALEXIS SANDERSON
Since the text has just stated that Amr_tesvara can take on any form I
infer a rule that on all calendrical occasions on which the worship of acertain deity was required, the Saiva officiant was to worship Amr
_tes-
vara (and/or Amr_talaks
_m�ı ) as that deity. This is certainly how Ks
_e-
mar�aja understands the matter. He explains that when the text says in19.102a that Amr
_tesvara is able to take on any form at will (k�ama-
r�upam in his version) it means that he ‘‘assumes the form of whicheveris the deity of the special occasion in question (tattannaimitti-kadevat�ak�aram)’’. In other words the officiant’s cult is always that ofAmr
_tesvara (and/or Amr
_talaks
_m�ı), unchanging in its essence, since
that resides in the Mantras, but it can be inflected to take on the formof any other cult as required, by substituting the form and other exter-nals of the appropriate deity with or without his or her consort.53
I propose that this inference provides the key to understanding whythe text did not restrict itself to the icons of Amr
_tesvara and Amr
_ta-
laks_m�ı but after setting out the cult of Amr
_tesvara in chapters 2 to 8
devoted chapters 9 to 13 to his visualization first as the deities of thefour specific Saiva divisions (Siddh�anta, V�ama, Daks
_in_a and Kaula)
and then, in chapter 13, as the principal deities beyond the boundariesof the Mantram�arga, including the non-�Agamic, lay forms of Siva him-self.54 For these deities outside the five Saiva systems of the text (thefour and the uninflected cult of Amr
_tesvara) are evidently those of
brahmanical calendrical worship, among whom Siva himself is num-bered. I therefore interpret the absolute universality of Amr
_tesvara,
53 I propose that if the deity were male then Amr_tesvara alone would be in-
voked; if female, then Amr_talaks
_m�ı; and if accompanied by a female consort, as
in the case of Indra and Sac�ı in the Indra festival of the Vis_n_udharmottara, then
Amr_tesvara and Amr
_talaks
_m�ı. Evidence that Amr
_talaks
_m�ı was invoked when the
worship of a goddess was required will be presented in due course in the case ofthe worship of the royal sword.
54 These are the four-armed form of Siva (13.29–30) cited in the Appendix as
relevant to the date of the text, the multi-armed dancing form (=Nr_tyarudra,
Nr_tyesvara, etc.) (n�at
_yastham
_19.31ab), Ardhan�ar�ısvara/Gaur�ısvara (um�ardha-
dh�arin_am_
19.31c), Harihara (vis_n_u-r-ev�ardhadh�arin
_am_
19.31d), Siva and P�arvat�ı at
their wedding (viv�ahastham_) 19.32a) (19.32a), and Siva and P�arvat�ı side by side (?)
(sam�ıpastham_
18.32b, = Um�amahesvara?). Cf. n. 4 above.
253SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
much vaunted in the text,55 not as an expression of ontological tran-scendence, though the liberationist Saiva learning of the non-dualistscould overcode it in that sense,56 but as a device to enable the officiantto penetrate the territory of brahmanical observance, shadowing therites of the brahmanical royal chaplain at every step or subsuming themwithin his office. For the Kashmirian �Adipur�an
_a-Tithikr
_tya requires the
king to offer worship, that is to say, to have worship performed by hischaplain, for the whole range of brahmanical deities on the days in thelunar month that are sacred to them.57
It might be urged against this interpretation that the Netratantraincludes the Buddha among the forms that may be assumed byAmr
_tesvara. For the Buddha is evidently not a brahmanical deity.
That objection might hold for other areas of the Indian world butnot for Kashmir. For in its account of the local religious calendarthe Kashmirian N�ılamatapur�an
_a requires the worship of the Bud-
dha in celebration of the days of his birth and Nirv�an_a during the
3 days of the moon’s passing from Pus_ya to Magh�a in the bright
half of Vais�akha.58 Moreover, the Netratantra refers to the Buddhaat the end of its description of his iconic form as ‘‘bestowing thereward of liberation upon women’’.59 This suggests that the wor-ship of [Amr
_tesvara as] the Buddha was a duty that the Saiva offi-
ciant was required to perform for the special benefit of the womenof the palace. Patronage of Buddhism in Kashmir was not providedby royal women alone, but in the political history of the kingdomcompleted by the poet-historian Kalhan
_a in AD 1148/9 they do
figure conspicuously in this role in his account of events immedi-ately before and during the K�arkot
_a dynasty (c. 626–855/6), the
55 Netra 9.17 b: *sarvas�adh�aran_o hy es
_a (N : sarv�as t�a + + + + hy es
_a Ed.);
13.44: sarvas�adh�aran_o devah
_sarvasiddhiphalapradah
_/sarve�s�am eva *var�n�an�am_ (N :
mantr�a�n�am_ Ed.) j�ıvabh�uto yata�h sm�rta�h; 13.46ab: vikalpo naiva kartavya�h sarva-s�adh�ara�no yata�h; 14.8ab: s�adh�ara�no mantran�atha�h sarve�s�am eva v�acaka�h; 16.23c–24:dvait�advaitavimisre v�ap�ı�s�to vai siddhido bhavet=yasm�at sarvagato deva�h visvar�upoma�nir yath�a=s�adhakasyecchay�a ce�s�ta�h siddhido bhavati dhruvam; and 19.82cd: sarva-tantre�su s�am�anyo m�rtyujit praka�t�ık�rta�h.
56 See, e.g., K�semar�aja ad 6.8cd (param_
sarv�atmakam_caiva mok�sadam_ m�rtyu-
jid bhavet): mah�as�am�anyamantrav�ıryar�upatv�an m�rtyujinn�athasyettham_ nirdesa�h. sa-rv�atmakam
_param�advayam:
57 �Adipur�an_a-Tithikr
_tya ll. 2828–2843.
58 N�ılamata 689–695.59 Netra 13.36cd: *dhy�atv�a hy evam
_prap�ujyeta (N : evam
_dhy�atah
_p�ujitas ca Ed.)
str�ın_�am_moks
_aphalapradah
_‘He who bestows the reward of liberation on women
should be visualized in this way and then worshipped’.
254 ALEXIS SANDERSON
period towards whose end I hold the Netratantra to have beencomposed.60
Immediately after instructing the officiant to worship Amr_tesvara as
Indra on the occasion of the pole festival theNetratantra goes on to rulethat on the Great Ninth (Mah�anavam�ı), the ninth day of the brighthalf of the next month, �Asvayuja (August/September), he shouldmake lavish offerings to the deity and worship the king’s weapons:
19:104 gos_u br�ahman
_araks
_�artham �atmanah
_svajanes
_u ca
mah�anavamy�am_p�ujyeta bh�uriy�agena vesmani
105 p�urvoktasreyam �apnoti �ayur�arogyasam_padah
_astray�agam_prayatnena kartavyam
_siddhihetutah
_106 astrasiddhim av�apnoti prayokt�a phalam asnute
104a gos_u br�ahman
_a conj. : gobh�ubr�ahman
_a N : gobr�ahman
_es_u Ed.61
105a p�urvoktasreyam �apnoti N : p�urvoktam_
samav�apnoti Ed. 105b
sam_padah
_N : sam
_padam Ed. 105cd astray�agam
_prayatnena kartavyam
_siddhihetutah_
N : astray�agah_
prakartavyah_
prayatn�at siddhihetave Ed.
106b prayokt�a Ed. : prayukt�a N
On the Great Ninth he should worship [Amr_tesvara] with lavish offer-
ings in his home for the protection of cows, the land, brahmins, himselfand his household. He will attain the above-mentioned benefits, long
60 Amr_taprabh�a, queen of Meghav�ahana, probably early in the sixth century,
constructed the monastery Amr_tabhavana for foreign Buddhist monks (R�ajatara _ngin
_�ı
3.9); his wife Y�uk�adev�ı competed with her fellow-wives by founding a splendid Bud-dhist monastery at Nad
_avana (3.11); Indradev�ı, another wife of this king, founded
the monastery Indradev�ıbhavana and a St�upa (3.13); many other monasteries were
built in their names by Kh�adan�a, Samm�a and other wives of his (3.14). Amr_ta-
prabh�a, wife of Ran_�aditya-Tunj�ına III (probably in the late sixth century), installed a
Buddha statue in a monastery built by Bhinn�a, another of Meghav�ahana’s wives
(3.464). Ana_ngalekh�a, wife of Durlabhavardhana (r. c. 626–662), founded the monas-
tery Ana_ngabhavana (4.3) and Prak�asadev�ı, wife of Candr�ap�ıd_a (r. c. 712–720/1),
the monastery Prak�asik�avih�ara (4.79). Support for Buddhism within Kashmirianroyalty appears from the R�ajatara _ngin
_�ı to have reached its highest point during
the reign of Lalit�aditya (c. 725–761/2). The king himself, though personally aBh�agavata, founded several Buddhist monasteries and St�upas and installed Bud-dha images (4.188, 4.200, 4.203–04, 4.210), as did his Central-Asian chief minister
Ca_nkun_a (Chin. jiangjun ‘General’) (4.211, 4.215, 4.262). There is no evidence of
royal support for Buddhism after this reign in the R�ajatara _ngin_�ı. It records no
Buddhist foundations or installations for the period of the Utpala dynasty (855/
6–980/1) and thereafter only one, the construction of a monastery by Bhadre-svara, the chief minister of Sam
_gr�amar�aja (r. 1003–1028) (7.121).
61 The conjecture gives an irregular syntax; but it is one seen repeatedly in theNepalese manuscript of this text, and it agrees with the sense of Ed.’s reading.
That is readily explained as an attempt to remove this anomaly and N’s gobh�u asa scribal error prompted by common usage in contexts of donation, as here ingobh�uhiran
_yavastr�adyaih
_(16.112c).
255SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
life, good health and wealth. He should perform the ceremony of wor-shipping the weapons [on that day62] with special care in order to bringabout Siddhi. He will indeed accomplish their Siddhi, and he who com-missions [the ceremony] will achieve [victory in battle as his] reward.
Ks_emar�aja interprets the attaining of the Siddhi of the weapons to
be their transformation into weapons with celestial power (divy�ani)and the person who commissions the ceremony to be ‘‘the king orthe like’’ (r�aj�adih
_). ‘‘The like’’, I propose, are others with troops
under their command, such as provincial governors (man_d_ale�sah
_).63
Since the purpose of the ritual is that they should be victorious inbattle, the weapons can only be theirs.
The deity of this autumnal festival, which marked the beginning ofthe season of military campaigns and did indeed include a ceremony inwhich the royal weapons and insignia were worshipped, is the martialgoddess Bhadrak�al�ı. According to the Vis
_n_udharmottara’s account of
this festival64 the king should have a pavilion for the worship of Bha-drak�al�ı (bhadrak�al�ıgr
_ham) constructed on the northeast side of his cap-
ital. He should worship her there on a painted Pat_a on the ninth day
of the bright half of the month after worshipping the weapons, armour,parasol, banner and all the other royal insignia (r�ajali _ng�ani) on theprevious day.65 The N�ılamatapur�an
_a tells us that the weapons are to
be worshipped in a shrine of Durg�a during the preceding night.66
No doubt when Amr_tesvara was made to take on this form, or
indeed that of any other goddess in the calendar, he did so in his
62 Ks_emar�aja ad 105cd: mah�anavamy�am eva.
63 Ks_emar�aja ad 106ab: divy�any astr�an
_i mantraprabh�av�at sam
_p�adayati: r�aj�adis ca
vijayam �apnot�ıty �aha: prayokt�a p�urvoktay�ajayit�a ‘Through the power of the Mantra [of
Amr_tesvara] he makes the weapons celestial. He now states that the king or other [com-
mander] achieves victory in the words ‘‘He who commissions will achieve [his] reward’’.‘‘He who commissions’’ is the person who has the aforesaid sacrifice performed’.
64 Vis_n_udharmottara 2.158.1–8.
65 Vis_n_udharmottara 2.158.4: tatraiv�ayudhavarm�adyam
_chattram
_ketum
_ca p�ujayet=
r�ajali _ng�ani sarv�an_i tath�astr�an
_i ca p�ujayet. The same is seen in Agnipur�an
_a 268.13–14:
bhadrak�al�ım_
pat_e likhya p�ujayed �asvine jaye=suklapaks
_e tath�as
_t_amy�am �ayudham
_k�ar-
mukam_dhvajam=chatram
_ca r�ajali _ng�ani sastr�adyam
_kusum�adibhih
_.
66 N�ılamata 780–782. This practice of worshipping the royal weapons and otherinsignia during the Navar�atra festival was not restricted to Kashmir. See, e.g.,
Sivapriyananda (1995), plates 55–58, 91–92, and 96 for photographs of the royalswords, the royal crown and fly-whisk installed for worship beside the image ofC�amun
_d_esvar�ı, the lineage goddess of the Mah�ar�ajas of Mysore, in their royal pal-
ace during the Navar�atra festival that culminates on this ninth; and Tod (1920,p. 683) for the worship of the royal sword, shield and spear on Mah�anavam�ı inthe royal palace in Udaipur.
256 ALEXIS SANDERSON
female aspect, through his consort Amr_talaks
_m�ı. This would have
prevented an awkward clash of genders. But it is suggested indepen-dently by the ruling seen above that it is Amr
_talaks
_m�ı that is to be
worshipped in the king’s sword. For the underlying identity of thedeity of that weapon is indeed Bhadrak�al�ı. The worship of Bhad-rak�al�ı on the king’s sword before he goes to war is treated at somelength in a text attributed to the �A _ngirasakalpa of the Atharvavedins’ancillary literature and included in the Orissan Paippal�adavas�adi-s_at_karmapaddhati, a work that sets out a large number of rites that
should or may be performed by chaplains for their royal patrons.67
That the Netratantra should mention only these two calendricalceremonies, the Indra festival and the worship of the royal weapons[and Bhadrak�al�ı] on the Great Ninth (Mah�anavam�ı), is in keepingwith the proposition that the officiant in this text is one who isworking in the territory of the king’s personal chaplain, since thesetwo are the principal festivals that engage the king. That can be seenfrom the fact that in the detailed account of the king’s ritual obliga-tions in the Vis
_n_udharmottara they are the only calendrically fixed
annual ceremonies with a marked civic dimension apart from theVais
_n_ava festivals that mark the four months of Vis
_n_u’s sleep.68
After prescribing the worship of Bhadrak�al�ı and the royal weaponsand insignia on the Great Ninth the Vis
_n_udharmottara goes on to
67 See Paippal�adavas�adis_at_karmapaddhati pp. 105–113. I am very grateful to Dr.
Arlo Griffiths of the University of Groningen for sending me first a copy of thispublication, of which, according to its Sanskrit title page, he was the promoter( prots�ahakah
_), and then an electronic text of the same.
68 The Vis_n_udharmottara briefly lists the king’s periodic ritual duties (nityakarma)
in 2.152.1–7. They are (1) a monthly ritual bath when the moon is in the asterismunder which he was born ( janmanaks
_atrasn�anam) and (2) another when it is in the
asterism Pus_ya (pus
_yasn�anam), (3) worship of S�urya (the Sun) and Candra (theMoon)
on the days on which the sun moves from one zodiacal sign into the next, (4) the wor-ship of a planet (Graha) when it has been eclipsed by the Sun, (5) worship to be offered
on the day of the heliacal rising of the star Agastya (Canopus) (agastyap�uj�a), (6) theworship of Vis
_n_u during the 4 months mentioned, (7) an annual Ghr
_takambala-
Kot_ihoma, a fire-sacrifice requiring a number of priests working simultaneously over
many days to make 10 million oblations timed to end at the end of the 4 months ofVis_n_u’s sleep followed by a ritual in which the king is covered with a blanket (kamba-
lam) and then first has melted butter (ghr_tam) poured over him from eight, twenty-
eight or one hundred and eight vases, and then, after the blanket has been removed, is
bathed with consecrated water, (8) a ritual for Rudra (rudrap�uj�a) at the end of eachregnal year, and (9–10) the celebration of these two public festivals. Chapters 153–158then cover the major topics in detail. Chapter 153 deals with the worship of Vis
_n_u
during the 4 months, chapters 154–157 with Indra’s pole festival and chapter 158 withthe worship of Bhadrak�al�ı and the royal weapons onMah�anavam�ı.
257SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
another ceremony to be performed at this time. This is the lustration(n�ır�ajanas�antih
_) of the king’s soldiers, horses and elephants, which, we
are told, should be done for their welfare, but also for the greater pros-perity of the kingdom and the destruction of its enemies.69 A section col-ophon after N�ılamata 780–82 confirms the association of this ceremonywith the Great Ninth by identifying that day as n�ır�ajananavam�ı ‘the lus-tration ninth’. The section requires that the weapons be worshipped inthe temple of Durg�a during the night of the eighth and that the lustra-tion take place on the next day according to the procedure of the Atha-rvaveda, that is to say, the Veda of the king’s personal chaplain. Thelustration of the king’s horses and elephants is indeed scheduled for thisday in the section on the annual royal ceremonies to be performed bythe king’s domestic chaplain in the Atharvavedaparisis
_t_a.70
In the Netratantra too a lustration ceremony is taken up immedi-ately after its treatment of the rites of the Great Ninth, though hereit is a lustration of the king himself. Moreover there is no statementthat it is to be performed during the autumnal festival. The only ex-plicit instruction is that it is to be adopted when there is some ill toward off such as a life-threatening illness of the king or other mem-ber of the royal family:71
yad�a mr_tyuvas�aghr�atah
_k�alena kalito nr
_pah_19:107 aris
_t_acihnit�atm�ano desam
_v�a tatsut�adayah
_br�ahman_�adis
_u sarves
_u paurajanapades
_u ca
69 This is taught in Vis_n_udharmottara 2.159.1–47 and the ascribed benefits are
declared there in 46–47: s�antir n�ır�ajan�akhyeyam_kartavy�a vasudh�adhipaih
_=ks_emy�a
vr_ddhikar�ı r�ama naraku~njarav�ajin�am = 47 dhany�a yasasy�a ripun�a�san�ı ca sukh�avah�a
�s�antir anuttam�a ca=k�ary�a nr_pai r�as
_t_ravivr
_ddhihetoh
_sarvaprayatnena bhr
_guprav�ıra.
70 Atharvavedaparisis_t_a 17.1.1–8, 18.1.1–18.3.12, and 18b.2.1–9 (18b.2.1: mah�ana-
vamy�am_hastyasvad�ıks
_�a ‘the lustration of the [king’s] elephants and horses is on
the Great Ninth’). The 6th-century Br_hatsam
_hit�a of Var�ahamihira says in the
chapter devoted to this ceremony (n�ır�ajanas�antih_) that it should take place on
the twelth, eighth or fifteenth day of the bright half of K�artika or, as here, during[the bright half of] �Asvayuja (43.2). K�at
_hakagr
_hyas�utra 57.1 rules that one should
honour horses and all transports on the full moon day of �Asvayuja: �asvayujy�amasv�an mahayanti sarv�an
_i ca v�ahan�ani; and �Adityadarsana ad loc. explains ‘all trans-
ports’ as as elephants, mules, buffaloes, camels and the like’: sarv�an_i ca v�ahan�ani:
hastyasvataramahis_akharos
_t_r�ad�ıni ca.
71 We see a lustration prescribed both on the ninth of �Asvayuja and as a specialrite to be performed when the need arises in Artha�s�astra 2.30.51: n�ır�ajan�am �asvayujek�arayen navame ’hani=y�atr�ad�av avas�ane v�a vy�adhau v�a s�antike ratah
_‘Devoted to
rites for the warding off of ills [the superintendent of the king’s horses] should have alustration ceremony performed [by the Purohita not only] on the 9th day of �Asva-yuja, [but also] at the beginning or end of a military expedition or in time of sickness’.
258 ALEXIS SANDERSON
108 �s�aly�adis_u ca sasyes
_u phalam�ulodakena ca
durbhiks_avy�adhik�aryes
_u utp�atais c�apy anuttamaih
_109 tad�a nair�ajanam_k�aryam
_r�aj~n�am
_r�as_t_rasya vr
_ddhaye
p�urvavad yajanam_
kr_tv�a kalasen�abhis
_i~ncayet
110 nih_sa _nkam
_nirjane r�atrau subharks
_e ca tath�am
_sake
jayapun_y�ahasabdena vedama _ngalasvastikaih
_111 abhis_i~nceta r�aj�anam
_siddh�arth�a~n juhuy�ad bah�un
nair�ajanavidh�anena n�am�a _nkam_juhuy�at priye
112 vahnau sam_kruddhamanas�a aj�am
_s ca proks
_ayed bah�un
tr_ptyartham
_bh�utasa _nghasya mantr�ı raks
_�artham udyatah
_113 s�akunokty�am_sagaty�a v�a vij~n�aya sakunam
_hitam
yaks_endrasivav�arun
_y�am
_niry�atah
_sarvasiddhidah
_114 atha p�urvoktavidhin�a gr_he y�agam
_tu k�arayet
y�avat sapt�ahnikam_devi bh�urihomena siddhidam
115 tasy�acal�a mah�alaks_m�ı r�ajyam
_v�a yad abh�ıpsitam
bhaum�antariks_asiddh�ıni pr�apnuy�an nr
_patih
_sukh�ı
116 tad�a nair�ajanam_khy�atam
_sarvasreyaskaram
_param
p�urvoktam_nasyate dos
_am_devi n�asty atra sam
_sayah
_106d kalito nr
_pah_
Ed. : kalitam_nr_pam
_N 107a aris
_t_acihnit�atm�ano conj. :
aris_t_acihnit�ad�ana N : aris
_t_acihnit�atm�a vai Ed. 107b de�sam
_N : deso Ed.
107d paurajanapades_u ca N : n�ase janapadasya ca Ed. 108a s�aly�adis
_u ca
sasyes_u Ed. : s�alic�urn
_�adisasyes
_u N 108b phalam�ulodakena N : phalam�u-
lodakes_u Ed. 108d utp�atais c�apy anuttamaih
_N : utp�ates
_u mahatsu ca Ed.
109a nair�ajanam_N : n�ır�ajanam
_Ed. 109b r�aj~n�am
_N (cf. 19.129b): r�aj~no Ed.
� r�as_t_rasya vr
_ddhaye N : r�as
_t_ravivr
_ddhaye Ed. 109d kalasen�abhis
_im_cayet
N : kalasen�abhis_ecayet Ed. 110a nih
_sam_kam
_N : nih
_sa _nko Ed. 110b su-
bharks_e ca tath�am
_sake Ed. : subham etat tu daisikah
_N 110c sabdena N :
sabdais ca Ed. 110d svastikaih_N : nih
_svanaih
_Ed. 111a abhis
_i~nceta conj. :
abhis_im_cata N : abhis
_i~ncet tu Ed. 111c nair�ajana N : n�ır�ajana Ed. 111d
n�am�am_kam
_juhuy�at pr
_yeN (=15.8ab) : n�am�a _nke sam
_skr_te priye Ed. 112a
sam_kruddhamanas�a corr. (=15.8c) : sam
_kruddhamanaso N : sam
_ruddha-
manas�a Ed. 113a s�akunokty�am_sagaty�a Ed. : s�akunom
_ty�asagaty�a N 113c
yaks_endra Ed. : yajem
_dra N � v�arun
_y�am
_N : v�arun
_y�a Ed. 114a p�urvokta
Ed. : p�urvoktena N 114b gr_he Ed. : gr
_ha N 115a tasy�acal�a N (sic also 18.79
in N and Ed.) : asy�acal�a Ed. 115c siddh�ıni N (sic also 18.79c in N) : siddh�ısca Ed. 116a nair�ajanam
_N : n�ır�ajanam
_Ed. 116c p�urvoktam
_nasyate dos
_am_conj. : p�urvoktam
_na pasyate dos
_am_N : p�urvokt�an n�asayed dos
_�an Ed.
If the king is touched by the power of death, if time has him in his sway, ifhe, the country or his sons or [wives] are marked by signs of impending
death, if all the inhabitants, both of the capital and elsewhere, both brah-mins and others, the rice and other crops of grain together with fruit, rootsand water [are in danger], or if there arises famine, an epidemic, or any
other ominous abnormality of nature (utp�atah_),72 then [the officiant] should
72 For utp�atah_in this sense see Br
_hatsam
_hit�a 45.1–2.
259SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
undertake the ceremonies of lustration (n�ır�ajanam) so that kings and theirkingdoms may prosper. After worshipping as before he should do anAbhis
_eka [by sprinkling with water] from the [consecrated] vase. When the
asterism and degree are auspicious he should give this Abhis_eka to the king
in an isolated place to the accompaniment of cries of ‘‘Victory!’’, ‘‘Meritori-ous Day!’’,73 auspicious [chanting of ] the Vedas, and [the eulogies of ]bards.74 He should [then] offer into the fire a large quantity of mustard
seeds.75 Beloved, he should offer them into the sacrificial fire with the pro-cedure of lustration, indicating the beneficiary’s name,76 with angry mind.The officiant, determined to accomplish this protection, should [then] asper-
ge [and sacrifice] numerous goats to gratify the hostile spirits and the horde[of Mothers, Yogin�ıs and the rest77]. If when he has determined the auspi-cious moment following the instruction of the [royal] astrologer or by cal-culating the degree [of the zodiac] the officiant goes out [of the capital
accompanied by the king78] to the north, the northeast, or the west, hewill bestow all supernatural benefits [on him]. Then, O Goddess, fol-lowing the procedure taught above he should perform the Siddhi-
73 The text refers here to the brahmanical practice of pun_y�ahav�acanam, the
uttering of the word pun_y�aham ‘meritorious day’ thrice at the beginning of any rite
to promote its success (ma _ngalam).74 For these auspicious, apotropaic accompaniments at the time of Abhis
_eka cf.,
e.g., Br_hatsam
_hit�a 47.49 concerning the king’s Pus
_yasn�ana: vandijanapauravipra-
praghus_t_apun
_y�ahavedanirghos
_aih_=samr
_da _ngasankhat�uryair ma _ngalasabdair hat�anis
_-
t_ah_; and N�ılamata 824 concerning the brahmanical consecration of the king: sn�anak�ale
ca kartavyam_mahat kalakalam
_tath�a=v�aditrasa _nkhapu
_ny�ahas�utavandijanaih
_saha. The
same applies in the Abhis_eka of an initiate; see, e.g., Bhojadeva, Siddh�antas�arapaddhati
f. 34r4–v2: bhadr�asanam_
vinyasya tasmin sis_yam
_vinyasya sa _nkhat�uryav�ın
_�aven
_usvasti-
pun_y�ahavedadhvanibhih
_kr_tama _ngalam
_. . . abhis
_ecayet.
75 Mustard seeds, also called sars_apah
_and raks
_oghnah
_, are believed to have the
power to fend off evil. See Netra 15.7–11.76 Ks
_emaraja ad loc.: amukasya n�ır�ajanam astu sv�ah�a ity atra prayogah
_‘The for-
mulation here is ‘‘May there be lustration of ‘[name]’, SV�AH�A’’ ’. For ‘[name]’ (amu-ka-) the officiant is to substitute the name of the king and, of course, to precedethis formulation with the Base-Mantra (m�ulamantrah
_) of Amr
_tesvara. The expletive
SV�AH�A is the closure ( j�atih_) required when making a regular oblation into the fire,
taking the place of the NAMAH_
at the end of other offerings. Thus for King Sa_nka-ravarman, for example, the Mantra to be uttered with each oblation would be:
OM_
JUM_
SAH_
SR�ISAN_ KARAVARMAN_O N�IR�AJANAM ASTU SV�AH�A.
77 The most natural understanding of the expression bh�utasa _nghah_is as a Tatpu-
rus_a meaning ‘the horde of hostile spirits’. Ks
_emar�aja, however, no doubt with rit-
ual procedure in mind, takes it as a singular Dvanda meaning ‘hostile spirits andthe horde’ specifying the latter as that of the Mothers, Yogin�ıs and others(bh�ut�ani ca sam
_ghas ceti sam�asah
_: sam
_gho m�atr
_yoginy�adigan
_ah_).
78 That the officiant goes out with the king is a detail added by Ks_emar�aja ad
loc.: vijay�abhimukhena r�aj~n�a saha niry�atah_‘gone forth with the king intent on vic-
tory’. The sense is that the king and his chaplain enact the king’s matching forthto war after lustration.
260 ALEXIS SANDERSON
bestowing ritual of [Amr_tesvara’s] worship in the [S�anti] temple79
for seven days, together with lavish offerings into the sacrificial fire.He [for whom this sacrifice is performed80] will achieve permanentgreat wealth, sovereignty or whatever else he may desire. The king
will be contented and attain the Siddhis of both earth and sky.81
Then, O Goddess, it is said that he has received the highest lustra-tion, that which bestows all benefits. The aforesaid evils cease to exist.
Of this, O Goddess, there is no doubt.
There is no reference here to the lustration of the king’s soldiers,horses and elephants, as there is in the Vis
_n_udharmottara, but the
Netratantra follows its lustration of the king with instruction in themeans by which the Saiva officiant should protect the king’s cattle,horses, elephants, goats and other livestock:
19:117 gos_u madhye yajed yasm�at sad�a vardhati gokulam
sind�uram_gairikam
_v�api abhimantreta mantravit
118 yoktavyam_gos_u raks
_�artham
_�sr__ngordhve sarvados
_ajit
asv�an�am_raks
_an_�arth�aya p�urvoktavidhin�a yajet
79 The Netra says only that the ritual should take place gr_he ‘in the house’ or
‘in the temple’. Ks_emar�aja understands this to mean r�aj~no gr
_he, i.e., ‘in the royal
palace’. I conjecture that the unspecified gr_ham is the temple known as the
�s�antigr_ham, the temple for the performance of S�anti rituals to protect the king
and the kingdom. This does not necessarily contradict Ks_emar�aja’s opinion,
which might be expected to be well-informed on such a point. For according tothe Saiva Pratis
_t_h�atantras this temple could be built in the northeast quarter of
the royal palace or of the residence of [his] Saiva Guru. See Mayasam_graha f.
33v (5.188abc): atha bh�ubhr_nniv�aso ’tra kury�ad vedhasi tad gr
_ham ais�anye
s�antigr_ham
_; Pi _ngal�amata f. 74v4 (10.151ab) (concerning the cumbakagr
_ham ‘the
residence of the Guru’): b�ahye ny�asam_
punar devi tatrese s�antikam_
gr_ham.
80 This is Ks_emar�aja’s interpretation ad loc.: yadartham
_caivam ijyate ‘‘asy�acal�a
mah�alaks_m�ı . . .’’.
81 The same language is used in 18.79 to describe the benefits that accrue tosomeone who has received Abhis
_eka from a vase of water in which [Am
_r_ta]
laks_m�ı has been installed and worshipped, except that to Siddhis of earth and
sky that passage adds those of heaven: tasy�acal�a mah�alaks_m�ı r�ajyam
_v�a yad
abh�ıpsitam=bhaum�antariks_a *siddh�ıni (N : siddhim
_ca Ed.) divy�am
_*caivai�svar�ım
_(Ed. : caivesvar�ı N) �subh�am_. I have not encounted the notion of the Siddhis of
these realms elsewhere in the literature and Ks_emar�aja offers no explanation on
either of its two occurrences in this text. Elsewhere the adjectives bhaumah_,
�antariks_ah_=antariks
_agah
_and divyah
_occur together with reference to phenomena
that portend calamities (utp�at�ah_) (e.g. Atharvavedapari�sis
_t_a 2.2.3: divy�antariks
_a-
bhaum�an�am utp�at�an�am; Br_hatsam
_hit�a 47.53ab) or to hostile spirits (e.g. Svacchanda
3.20). The sense is probably that the king gains power over these phenomena in the
sense that he is immunized against their influence.
261SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
119 abhimantreta kalasam_siras tes
_�am_prad�apayet
siddh�arthakam_japitv�a tu kan
_t_he k�aryam
_tu m�urdhani
120 sarvados_avinirmukt�an gaj�am
_s caiva tu raks
_ati
ajes_u pasaves
_v evam
_raks
_�am_sarvatra k�arayet
121 sarvapr�an_is_u raks
_�artham
_yoktavyam
_nr_pateh
_sad�a
mah�as�antir bhavet tes_�am_durbhiks
_am_nasyate sad�a
117b vardhati N : vardheta Ed. 117c gairikam_Ed. : gaurikam
_N 117d
abhimantreta N : abhimantryaiva Ed. 118a yoktavyam_N : yojayed Ed.
118c asv�an�am_raks
_an_�arth�aya N : asv�an�am api raks
_�artham
_Ed. 119a abhi-
mantreta N : abhimantryaiva Ed. 119b siras tes_�am_prad�apayet N :
m�urdhni tes_�am_prap�ataye Ed. 119c siddh�arthakam
_japitv�a tu N : si-
ddh�artho mantrajaptas tu Ed. 119d k�aryam_
tu N : k�aryo ’tha Ed. 120bcaiva tu N : caiva ca Ed. 120c ajes
_u pasaves
_v evam
_conj. (Aisa): ajes
_u
pa�savo hy evam_
N : aj�adis_u pasus
_v evam
_Ed. 120d sarvatra N : sarves
_u
Ed. 121b yoktavyam_
N : yoktavyo Ed. 121d nasyate sad�a N : na-syati ks
_an_�at Ed.
The officiant should worship [Amr_tesvara] in the midst of the [king’s]
cows, since [by this means] his herd will constantly increase. He
should empower vermilion powder or red chalk with the Mantra andapply it to the tips of their horns to protect them, for it will overcomeall evils. To protect the [king’s] horses he should offer the cult in the
manner stated above, empower with the Mantra a vase [filled withwater] and pour it[s contents] on their heads. He should empowermustard seeds by repeating the Mantra over them and then place
them on their necks and heads. [Mantra-empowered mustard seeds82]also protect the [king’s] elephants, [so that they become] free of allevils. He should do the same rite of protection for the [king’s] goatsand [all his other] domestic animals.83 He should employ his Mantra
at all times for the protection of all the king’s living creatures. Theywill benefit from a general warding off of ills (mah�a�s�antih
_). Famine
will cease forever.
The Netratantra also requires its Saiva officiant to perform the wor-ship of Amr
_tesvara as a S�anti ritual whenever the realm
(man_d_alam) is affected by an earthquake, the falling of a meteor
(ulk�ap�atah_), a drought, excessive rains, a swarm of mice or other
pests, phenomena such as the untimely appearance of flowers, thedestruction or splitting of an image of a god, fevers, [illnesses
82 I follow Ks_emar�aja in taking these mustard seeds (siddh�artho mantritah
_) to
be the subject here.83 I take the causative k�arayet here in the non-causative sense, a licence com-
monly seen in such scriptural texts. See, e.g., Svacchanda 423c–4: tato ghr_tena
sam_pl�avya abhim�anam
_tu k�arayet=aham eva param
_tattvam
_par�aparavibh�agatah
_=
tattvam ekam_
hi sarvatra n�anyam_
bh�avam_
tu k�arayet.
262 ALEXIS SANDERSON
caused by] spiders and other poisonous insects (l�ut�adi), and un-timely deaths (apamr
_tyuh
_) (19.122–124b); or if there is suffering
caused by some evil action in the past (karmados_ah_), by a seizing
spirit (grahados_ah_), by some offence against a god or Guru
(tirobh�avah_), by an error in the propitiation of a Mantra (mantra-
cchidram), by the poison of snakes and the like, by such ills as skineruptions caused by insect bites (k�ıt
_avisphot
_ak�adayah
_), by imbal-
ances of the humours, piles, eye-diseases, contagious skin diseasesand the like (visarpak�adayah
_), by every kind of illness, by grief and
other states causing insanity, and if the brahmins or others [in therealm] have been cursed by a god or [a sage] (19.122–128).
Moreover, even when no emergency has arisen:
19:129 pratyaham_havanam
_k�aryam
_r�aj~n�am
_r�as_t_ravivr
_ddhaye
sukhena bhujyate r�ajyam_n�atra k�ary�a vic�aran
_�a 84
He should offer a fire-sacrifice for the prosperity of the king and thekingdom every day. [If he does so the king] will enjoy a happy reign.
There can be no doubt of this.
For:
19:130 sakr_tp�ujanam�atre
_na nasyante him
_sak�adayah
_nas_t_�a dasa diso y�anti sim
_hasyeva mr
_g�adaya
_h
131 satat�abhy�asayogena d�aridryam_nasyati kul�at
yasmin dese ca k�ale ca nivasen mantravit sad�a132 �ıtayo vy�adhayas caiva kh�arkhod�as tasya v�a grah�ah
_s�akinyo vividh�a yaks_�ah_pis�ac�a r�aks
_as�as tath�a
133 b�alagrah�as ca visphot_�a vyantar�as c�apar�as ca ye
sarv�an_i vis
_aj�at�ani durbhiks
_am_grahap�ıd
_anam
134 sarvam_
na prabhavet tatra mantravitsam_nidh�anatah
_Even if he worships [Amr
_tesvara in this way] only once, the harming
spirits and other [afflictions] will be crushed and will flee in all tendirections85 like deer and other [prey] from a lion. If he constantly
repeats [the cult], poverty will be removed from the lineage. When [aGuru who is] a master of this Mantra is in permanent residence in aland, his mere presence will ensure that no calamities, diseases,Kh�arkhodas, siezers (Grahas), none of the various kinds of S�akin�ı,Yaks
_as, Pis�acas, R�aks
_asas, seizers of infants (B�alagrahas), boil[-causer]s
and other Vyantara beings, no poison, famine, or oppression by planets,will have any power over the [king] there.
84 I have not been able to collate 19.125–139c with N, since they are the content
of N’s folios 68v and 69r, which are lacking in my copy of the NGMPP microfilm.85 In the cardinal directions, the intermediate directions, upwards, and down-
wards.
263SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Finally, as is the case with the brahmanical royal chaplain, thefunctions of the Saiva officiant prescribed by the Netratantra donot end with the king’s life. If he or any of the princes dies the offi-ciant should perform a special form of postmortuary initiationknown as the Rescue of the Dead (mr
_toddh�arad�ıks
_�a). He may offer
worship to Amr_tesvara at the time of the cremation itself, installing
an image of [Amr_tesvara as] Bhairava in the cremation ground
where the body has been burned; and he should perform the subse-quent Sr�addha rites:
18:112 . . . nr_patau tatsut�an�am
_. . .
. . .mr_tasyoddharan
_�arth�aya d�ıks
_�artham
_paramesvarah
_116 yas_t_avyah
_p�urvavad devo vises
_�at tatra c�akr
_tih_kartavy�a r�ajat�avasyam
_sadr
_s�ı dv�adas�a _ngul�a
117 k�ary�a v�a gomayair devi kusair v�a sn�anasodhit�ad�ıks
_aiva tatra sam
_sk�arah
_vy�apty�a yatrastham �anayet
118 an_um_ca yojayet tasmin p�urn
_�ahuty�a ks
_ipe ’nale
yojany�a sivatattve tu tatah_s�ayojyat�am
_labhet
119 sr�addhe sam_p�ujayed devam antyes
_t_�av athav�a yajet
pratis_t_h�apyas tath�a devi dagdhapin
_d_e smas�anake
120 p�urvoktadravyasambh�arair p�urvoktavidhin�a guruh_p�urvoktam
_bh�ıs
_an_am_r�upam
_saktidvayasamanvitam
121 catus c�as_t_au thav�a devi p�urvadhy�an�avalokit�ah
_p�urvoktam_phalam �apnoti ity �aj~n�a p�aramesvar�ı
112d tatsut�an�am_
N : tatsutes_u Ed. 115c mr
_tasyoddhara
_n�arth�aya N :
mr_tes_�uddhara
_n�arth�aya Ed. 115d paramesvarah
_Ed. : paramesvaram N
116a yas_t_avyah
_p�urvavad devo Ed. : yas
_t_avyam
_p�urva devesam
_N 116b
vi�ses_�at Ed. : vises
_as N 116c r�ajat�avasyam
_em. (= reading rejected by
Ks_emar�aja ad loc.: r�ajatetety apap�at
_hah_) : rajat�avasyam
_N : rajas�ava-
syam_
Ed. 117a gomayair (conj.) : gopaye N : gomay�ad Ed. 117d
yatrastham �anayet conj. : yavastham �anayet Ed. : yatra samam_nayet N
118a a_num
_ca yojayet tasmim
_N : a _n�um
_s ca yojayet tasy�am
_Ed. 118b
ks_ipe’nale N (Aisa for ks
_iped anale) : saha ks
_ipet Ed. 118d s�ayojyat�am
_labhet N : s�ayujyabh�ag bhavet Ed. 119ab devam antyes_t_�av athav�a Ed. :
devam_mam
_tes_t_itveti v�aN 119c pratis
_t_h�apyasN : pratis
_t_h�apyam
_Ed. 120a
p�urvokta N : p�urvoktair Ed. 120b p�urvoktavidhin�a guruh_
N : gurun_�a
pr�agvidh�anatah_
Ed. 121a catus c�as_t_au thav�a N : catasro ’s
_t_�av atho Ed.
121b dhy�an�avalokit�ah_
Ed. : dhy�an�avalokitam_
N 121c p�urvoktam_
N :p�urvokta Ed.
To accomplish the initiation to rescue the dead for . . . the king or [anyof ] the princes . . .86 he should worship the Supreme Lord as above but
86 The passages omitted list other classes of dead who should receive this form
of initiation.
264 ALEXIS SANDERSON
with the difference that he must fashion a silver simulacrum [of thedeceased] twelve A _ngulas [approx. 21 cm] in length. Alternatively itmay be made with cowdung or blades of Kusa grass.87 He should purifyit with a bath. He should then perform the ceremony of initiation upon
it.88 By [meditating on himself as Siva] pervading [the universe](vy�apty�a) he should bring back the soul [of the deceased from] wherever
87 The option that the simulacrum should be made of silver is that of a readingr�ajat�a rejected by Ks
_emar�aja. I have retained it because it is supported by N
(rajat�a). Ks_emar�aja explains his preferred reading rajas�a ‘with powder’ as meaning
‘with rice-flour’ (s�alic�urn_ena), but that is not supported by other accounts of this
ritual. In his treatment Abhinavagupta gives an open list of materials that may be
used to make these simulacra, mentioning cow-dung, blades of Kusa/Darbha grass(Poa cynosuroides), fruit (Tantr�aloka 21.22d–23a, 33, 35, 40, 43) and ‘such thingsas the nutmeg (Myristica fragrans)’ (21.36ab: j�at�ıphal�adi yad kim
_cit tena v�a
dehakalpan�a). The Kashmirian Br_hatk�alottara, borrowing from the P�a~ncar�atrika
Jay�akhyasam_hit�a, matches the Netra in saying that the simulacrum should be
twelve A _ngulas in length, but differs in saying that it should be made with all itslimbs by shaping it out of white earth mixed with the five products of the cow and
water, or out of the wood of the trees Pal�asa (Butea frondosa) or Asvattha (Ficusreligiosa), or with a spray of flowers (pallavah
_) (B f. 195v4–5): tatah
_�svetamr
_d�alod
_ya
pa~ncagavyena c�ambhas�a=dv�adas�a _ngulam�atram_tu m�urttim
_kury�at *tad�akr
_tim (em. :
tad�akr_ttim
_Cod.)=�ap�ada*c�ulik�antam
_(corr. : c�ulik�atam
_Cod.) ca *sarv�a _ng�avaya-
v�anvit�am (em. : sarv�avayav�anvitam_
Cod.)/pal�as�asvatthajenaiva d�aru_n�a *pallavena
(em. : canavena Cod.) v�a. The Nepalese manuscript of the Jay�akhyasam_hit�a adds a
third wood as an option, but the reading is evidently corrupt: *tatah_(em. : tatada-
ta Cod.) svetamr_d�alod
_ya pa~ncagavyena c�ambhas�a=dv�ada�s�a _ngulam�atr�ım
_tu m�urttim
_*kury�at (corr. : kury�a Cod.) tad�akr_tim=�ap�ad�ac c�ulik�ant�a~n ca sarvv�a _ng�avayav�a-
nvit�am=*pal�as�asvatthayd�arbhyotthad�arun_�a (conj. : pal�asosvatthad�arbhotth�ad�arun
_�a
Cod.) *pallavena (em. : pavaluvena) v�a (f. 81r3–4). The Jay�akhyasam_hit�a published
on the basis of south-Indian manuscripts makes this third wood that of the birch(Baetula bhojapatra) (24.86cd): pal�as�asvathavalkotthad�arun
_�a). In the same tradition
is the simulacrum made of Ficus leaves and flowers known as a pus_pali _nga that is
animated with the soul of the deceased by the Saiva officiant in the Balinese postcre-mation ritual of the purification of the soul (mukur, nyekah, neles, etc.); see Hobartet al. (1996, pp. 125–6); and Stuart-Fox (2002, pp. 92–3); also the yogic ensouling of
the ‘flower-body’ (pus_pasar�ıra) in the Saiva-Bauddha postcremation rites of the
Javanese queen of Majapahit described in the Old Javanese Desawarn_ana (64.5, 67.2
[Robson, 1995, pp. 71 and 74]).88 Literally ‘initiation alone’ (d�ıks
_aiva). Ks
_emar�aja takes the point of the restric-
tive particle eva to be that in this case there is no need for the preliminary ritesknown as adhiv�asah
_that normally take place during the day before the initiation
proper.
265SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
it may be and place it in that [simulacrum].89 As he makes the FullOblation [after completing the oblations that eliminate the possibility ofreincarnation on any level of the universe] he should cast [the simula-crum] into the fire while [raising the soul through his own central
energy-channel and then] uniting it with the level of Siva. By this means
89 Neither the reading of N (yatra samam_
nayet) nor that of Ed. (yavastham�anayet) makes good sense. N’s reading might be a corruption of yatra samam
_na-
yet, confusion of the graphs sa and sa occurring so pervasively in manuscriptscopied by Newar scribes that it is arduous to record it. We could then take this
and the next quarter verse to mean ‘he should place (yojayet) the soul [of thedeceased] (an
_um) into that (tatra) into which (yatra) he should bring it to rest
(samam_
nayet)’, understanding the site of the placing to be the reality-level to
which the soul is to be raised through initiation and bringing it to rest to meancausing it to be one with that level. But at least two problems obtrude. The first isthat the phrase samam
_nayet, though common, always denotes elimination (as of
diseases or poison in the medical literature) or dissolution, as when a Yogin medi-tates on a lower reality-level being withdrawn into the one above it; see, e.g.,Jay�akhyasam
_hit�a 16.263d (ks
_m�akhyam
_tattvam
_samam
_nayet) and Laks
_m�ıtantra
35.8d and 53.7b (aham_k�are samam
_nayet), and 53.8b (m�ul�avyakte �samam
_nayet).
This is not the natural idiom for the uniting of the soul with a reality-level. Thesecond problem is that the interpretation leads to pleonasm: the action of fusingthe soul will be mentioned again in 118c and as fusion with the reality-level of Siva
(yojany�a sivatattve tu). I turn therefore to the reading yavastham �anayet in Ed. This Ipropose is an error, probably of the compositor of the Devan�agar�ı edition rather thanthe Kashmirian scribes, for yatrastham �anayet, tra and va being graphs that are
more readily confused in Devan�agar�ı than in S�arad�a. That Ks_emar�aja had this in
his text of the Netra is not certain, since he does not gloss it directly, confirmingonly �anayet with �an�ıya in his introduction to the next quarter verse. But I proposethat he does so indirectly in the eleven-line citation from the Ham
_sap�aramesvara
that he gives in his comment on vy�apty�a. For that describes the Great Net pro-cedure (mah�aj�alaprayogah
_) by means of which the officiant is to catch the soul of
the deceased in whatever other state of incarnation it resides and place it in the
heart [of the simulacrum] uttering the seed-syllable [of M�ay�a (HR�IM_)] and the soul’s
name: yatra srot�antare sthitam/gr_h�ıtv�a tat prayoge
_na mah�aj�alena yuktitah
_=gr_h�ıtam
_*hr_daye (em. : hr
_dayam
_Ed.) sth�apyam
_b�ıj�abhikhy�asamanvitam. I propose that for
Ks_emar�aja the point of this part of citation was that it clarified the meaning of
yatrastham in the Netra.The yogic procedure for catching the soul through visualization and the recita-
tion of the seed-syllable HR�IM_
(m�ay�ab�ıjam) is described by Abhinavagupta inTantr�aloka 21.25–26 and by the passage cited from the Ham
_sap�aramesvara. The
officiant is to meditate on himself as Siva pervading the universe, exhale, inhale,hold his breath, raise the vital power through the central channel to the point
twelve finger-breadths above his head, and then visualize this power moving outthrough all the worlds to find the soul. He should utter the syllable HR�IM
_and take
hold of that soul, visualizing it as resembling a drop of water.
266 ALEXIS SANDERSON
it will attain union [with Siva].90 He should [also] worship Siva [for the
deceased] in the Sr�addha ceremonies. Optionally he may do this wor-ship during the cremation ceremony itself. [In the latter case], O god-dess, the Guru should install an image of the frightening [Bhairava]
form taught above91 attended by two Saktis92 in the cremationground where the body was burned, employing the various offeringsalready mentioned and the aforesaid rites. Alternatively, O goddess,
there may be four or eight [attendant Saktis] contemplated with thevisualizations already taught.93
90 This kind of Saiva initiation was a conspicuous feature of Kashmirian life if
we may judge from satirical references to its practice in the eleventh-century worksof Ks
_emendra (Desopadesa 8.50: mr
_takoddh�are; Narmam�al�a 3.43: mr
_toddh�ara-
d�ıks_�a). Moreover, it may have been distinctive of that region. For the only prescrip-
tions of the practice known to me from Saiva sources other than the Netra andTantr�aloka are also Kashmirian: (1) S
_at_ka 3 of the Jayadrathay�amala (f. 156r1–v6
in the Catu�scatv�arim_satid�ıks
_�apat
_ala in the Ghoraghoratar�acakra)—its second and
third lines have been cited in this context without attribution by Jayaratha adTantr�aloka 21.6–9b — and (2) the Antes
_t_imr_toddh�arapat
_ala of the Br
_hatk�alottara
(B ff. 195r3–197r1, a section of that eclectic text borrowed with superficial adjust-ments from the P�ancar�atrika Jay�akhyasam
_hit�a24.76–105b.
Commenting on this passage of the Netra Ks_emar�aja refers to a non-Yogic alter-
native method for catching the soul of the deceased. A circular diagram is drawnwith OM
_(n�adah
_) at the centre and the syllables of the syllabary drawn in six cir-
cuits around it. The Sivanirv�an_avidhi, which gives the text of the Saiva cremation
ritual followed in Kashmir, illustrates this diagram and gives the full ritual proce-dure, Mantras, and deities. The last are M�ay�adev�ı, who is to be worshipped in a
dish full of offerings placed on a lamp that rests at the centre of the diagram ontop of OM
_HAM
_SAH
_followed by the name of the soul to be drawn in, and the
eight Ks_etrap�alas, who are to be worshipped around the periphery (pp. 242–246).
91 The reference is to the five-faced, ten-armed black Bhairava taught in 10.1–6bas the form assumed by Amr
_tesvara in the Daks
_in_a division. See also Tantr�aloka
26.7–8, commenting on such public installations.92 According to Ks
_emar�aja ad loc. the two Saktis of the Bhairava to be installed at
the place of cremation are ‘the full-bodied and the emaciated’ (saktidvayam_kr
_sasth�ulam). These, I propose, are the last two of the eight Mothers, C�amun
_d_�a
and Yogesvar�ı, since they are so described in the Kashmirian Br_hatk�alottara A f.
251r2: atip�urn_n_�a tu c�amun
_d_�a khad
_gak�adyasamudyat�a=sav�ar�ud
_h�a nr
_tyam�an�a savasrag-
d�amaman_d_it�a=sus
_k�a yogesvar�ı k�ary�a; and f. 251r3–4: evam
_vidh�a tu yoges�ı sir�al�a
vikr_t�anan�a.
93 The four are Siddh�a, Rakt�a, Sus_k�a and Utpalahast�a, and the eight are these
together with their four companions (D�ut�ıs): K�al�ı, Kar�al�ı, Mah�ak�al�ı and Bha-drak�al�ı. Their visualizations are given in Netra 10.17–37a.
267SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
The Netratantra, then, has shown us Saiva officiants active in almostall the areas of observance assigned by the Atharvavedic traditionto the brahmanical royal chaplain: rituals for the protection of theking, from rites attending his bathing, eating, exercise and sleep todaily and periodic sacrifices, rituals for his invigoration and victory,rituals of regular worship on the king’s behalf, including the greatpublic ceremonies of the Indra and Bhadrak�al�ı festivals, and rituals
for the king’s benefit after his death. Only the incidental function of
performing reparatory rites (pr�ayascitt�ıyam_
karma) receives no atten-tion here.94
Nor is the text backward in urging that the officiant’s servicesshould be lavishly rewarded. In the fifteenth chapter we read:
nr_p�an_�am_
nr_papatn�ın�am
_tatsut�an�am
_dvij�adis
_u
15:21 �ac�aryah_kurute yas tu sarv�anugrahak�arakah
_mantraj~nah_s�adhako v�atha sa p�ujyah
_sarvath�a prabhuh
_22 sam_m�anair asamair nityam
_d�anair vividhavistaraih
_22a asamair conj. : asanair N: vividhair Ed:95
Any �Ac�arya or S�adhaka compassionate to all and possessing masteryover this Mantra who does these rites of protection for kings, their
wives, their children, brahmins and the rest, should be constantly ven-erated with unequalled marks of distinction and with gifts both vari-ous and abundant.
and in the sixteenth:
gobh�uhiran_yavastr�adyaih
_key�urakat
_ak�adibhih
_16:113 p�ujyo ’sau paray�a bhakty�a s�antipus_t_yor vises
_atah
_yasm�an mantramayo so vai sivah_s�aks
_�at tu dai�sikah
_114 tena p�ujitam�atre_na sarve siddhiphalaprad�ah
_bhavanty avitatham_bhadre satyam etan na sam
_�sayah
_115 anyath�a siddhih�anih_sy�at kr
_tam_caiva nirarthakam
112c vastr�adyaih_Ed. : vastr�an
_i N 113b s�antipus
_t_yor conj: : s�antipus
_t_y�a
Ed. : s�antipus_t_ir N 113c so vai N : vai sa Ed: 114d satyam etan na sam
_-
sayah_N : satyam
_me n�anr
_tam_
vacah_Ed:
That [officiant] should be honoured with the greatest devotion with
gifts of cows, land, gold, cloth and the like, with armlets, braceletsand other [ornaments], particularly when he performs rites to ward offills or restore to health. For the Guru embodies the Mantra[-deities]. He is
94 For the six areas of the royal chaplain’s Atharvanic rituals see n. 17 above.95 This emendation is supported by a parallel in 19.135ab: d�anap�ujanasam
_-
m�anair asamaih_
p�ujyate yad�a.
268 ALEXIS SANDERSON
Siva made manifest. My beloved, if one honours him, then for that alone
all [the Mantra-deities] will certainly bestow the success of the Siddhi [onedesires]. This is the truth. There is no doubt. Otherwise the Siddhi will belost and one’s effort will be in vain.
and in the 19th chapter, after the passage cited above (19.139–134b) in which we are told that the permanent residence of such anofficiant in a kingdom will render it immune to all conceivablecalamities:
sa p�ujyah_sarvajant�un�am
_bh�upat�ın�am
_ca sarvad�a
19:135 d�anap�ujanasam_m�anair asamaih
_p�ujyate yadi
tena p�ujitam�atren_a sarve mantr�as ca p�ujit�ah
_136 bhavanti sukhad�as tatra
All men should honour that [officiant], and kings should do so con-stantly. When they have honoured him with unparallelled gifts, dem-
onstrations of respect, and marks of distinction, then by this alonethey will have honoured all the Mantra[-deities], who will reward themwith happiness in that [realm].
All this is very much in the style that had been adopted in the brah-manical context to promote the interests of the king’s personalchaplain, as can readily be seen from the following passage ofAtharvavedapari�sis
_t_a 4:
4:6:1 yasya r�aj~no janapade atharv�a s�antip�aragah_nivasasty api tadr�as
_t_ram_vardhate nirupadravam
2 yasya r�aj~no janapade sa n�asti vividhair bhayaih_p�ıd
_yate tasya tad r�as
_t_ram_pa _nke gaur iva majjati
3 tasm�ad r�aj�a vi�ses_en_a atharv�an
_am_
jitendriyamd�anasam
_m�anasatk�arair nityam
_samabhip�ujayet
1c tadr�as_t_ram_
corr: : tad r�as_t_ram_
Ed:
The kingdom of that king in whose realm dwells an Atharvavedic
master of the rites for warding off ills will prosper, free of all calami-ties. The kingdom of that king in whose realm he is not present isoppressed by diverse dangers. It sinks like a cow in the mud. There-
fore to that Atharvan [chaplain] whose senses are controlled the kingshould show exceptional honour at all times, by means of gifts, marksof distinction, and demonstrations of respect.
I take the ‘marks of distinction’ (sam_m�anam) to which this passage
and its Saiva parallels refer to be those insignia that served to dis-tinguish high dignitaries in the court culture of South and South-east Asia, attributes such as palanquins, white parasols andfly-whisks with golden handles, which would be displayed wheneversuch persons appeared in public. The term is used in this sense in the
269SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Khmer inscriptions, a corpus rich in records of such honours;96 andwe may compare Atharvavedaparisis
_t_a 3.1.17:
hastyasvam_
naray�anam_
divyam �abharan_am_
�atapatram_
hiran_yam
_ks_itigodhanadh�anyaratn�adikam
_ca gurave dady�at
[The king] should give his chaplain an elephant and a horse, a palan-
quin, the finest ornaments for his person, a gold[-handled] parasol,and [valuables] such as lands, cows, coin, grain and jewels.
Such insignia were calibrated as to status. The Vis_n_udharmottara
(2.13.7–9b) specifies that the pole of the king’s parasol should besix cubits (ca. 2.6 metres) in length, those of the royal chaplain,royal astrologer and head of the army (sen�apatih
_) five, and those of
the chief queen (mahis_�ı) and the crown prince (yuvar�ajah
_) four and
a half. It seems highly probable that the Saiva officiant would haveexpected no less than is promised to the royal chaplain in this pas-sage, thereby aspiring to recognition as a dignitary second in rankonly to the monarch himself.
5. CONCLUSION
In depicting Saiva officiants in the role of the traditional royal chap-lain, the Netratantra indicates the existence of a new class of Saivaspecialists envisaged nowhere else in the corpus of the surviving
96 See K. 762, 6: svasv�aminaf pras�ad�at sa ca r�ajasabh�adhipatyakr_tan�am�a sauvarn
_a-
kalasakara _nkasit�atapatr�adisanm�anah_[ f=Upadhm�an�ıya] ‘who had received the title
R�ajasabh�adhipati by the king’s favour and been honoured with the golden vase, the[golden] cup, the white parasol and other [insignia]’; K. 809, 43, concerning Indravar-
man’s Saiva officiant Sivasoma: þ iþþþþ yasya r�ajena sr�ındravarmman_�a*sita
(conj. : + + Ep.) chatraprad�an�adisanm�ananam ak�arayat ‘he caused him to be hon-oured by King Indravarman with such marks of distinction as the white parasol’; K.725, 20: *�atapatr�adisanm�anair (conj.: þþþ tra d�ıpanm�anair Ed.) asakr
_t tena
satkr_tah_‘Honoured by himmore than once with such marks of distinction as the par-
asol’. That the sanm�anam=sanm�ananam of the Khmer inscriptions is used in the samemeaning as sam
_m�anam=sam
_m�ananam here is evident from parallels in which it is
linked, as in the Netra and Atharvavedaparisis_t_a, with d�anam, p�ujanam (/satk�arah
_)
and synonyms; see K. 436, 17: p�uj�aprad�anasanm�ana; and K. 81 A 22: visrambha-d�anasanm�anaih
_yogyo yaf paryyatr
_pyata. For the golden-handled parasol see, e.g.,
Pi _ngal�amata f. 75r1 (10.159a): hemada_nd_am_
sitam_
chattram_. The granting of such a
parasol by the king is frequently mentioned in the Khmer inscriptions; see, e.g., K.273, 29 , K. 289 C, 54, and K. 323, 80, in the last of which those with this honour(hemadan
_d_�atapatrin
_ah_) are assessed as a distinct class for the purpose of fines.
270 ALEXIS SANDERSON
Saiva scriptures.97 But what was the nature of this encroach-ment? Several scenarios are conceivable.
We might imagine that the officiant of this text had taken theplace of the brahmanical chaplain altogether or that he coexistedwith him, providing the monarch with parallel Saiva observances todouble the chaplain’s. In the latter case the Saiva officiant mighthave matched all or only some of the chaplain’s activities. It is alsopossible that the encroachment of the Saiva officiant led to aretrenchment of the brahmanical chaplain’s activites, leaving somedomains in the hands of the Saiva officiant alone.
In the absence of independent historical evidence – as yet I knowof none – it is impossible to determine exactly the situation underly-ing the textual regulation. But I find the first scenario the least plausi-ble, since the dominant tendency in Indian religion has been one ofaccumulation rather than substitution. Furthermore, though theNetratantra shows us the officiant at work in nearly all the areasassigned to the chaplain it does not match every one of his activitiesin each. Thus in the area of daily activities we see a close match onlyin the rituals of protecting the king while he sleeps. There is no men-tion, for example, of the early-morning routine of giving the kinghis garments, ornaments, and perfume, annointing his eyes with col-lyrium, and then ritually bestowing on him his horse, his elephant,his palanquin, his sword and other royal insignia.98 Similarly, in thecase of the major periodic ceremonies, the Netratantra covers the In-dra festival and the autumnal festival of the Goddess, but does notmention the great biannual and annual fire-sacrifices of one hundred
97 It may be objected that the Netratantra is a prescriptive text and that it is there-fore illegitimate to infer practice from it, since a prescription may be an exhortationthat neither reflects nor brings that about. This is true in principle, but the probabil-
ity that the Netratantra was the blueprint for an institution that never existed isextremely remote. It is surely much more probable that its purpose, like that of theSaiva scriptures in general, was to authorize and regulate an already existent tradi-tion of practice that hitherto lacked adequate scriptural sanction. The principal
defect of such materials is not fantasy but schematization. The greater the range ofpractice that they seek to bring within their scope the greater their tendency to avoidthe level of detail that characterizes actual implementation, since in this way they
can avoid contradicting the specifics of current variants and instead provide a matrixof prescription within which all these variants can comfortably be accommodated.
98 These activities are set out in Atharvavedaparisis_t_a 4.1.1–24.
271SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
thousand and ten million oblations (laks_ahomah
_and kot
_ihomah
_) that
are among the brahmanical chaplain’s principal periodic duties.99
One might dismiss these discrepancies by saying that the Netratantragives only some examples of the officiant’s obligations rather than afull account. But that would be plausible only if we had some furtherreason to suppose that this was so. The alternative would be to sup-pose that the reason why rituals such as these fire-sacrifices were nottaken over by the Saiva officiant is that they were no part of thechaplain’s duties. But that is improbable, since they are included inthe accounts of the rituals to be performed for the king by his chap-lain in the N�ılamatapur�an
_a and �Adipur�an
_a-Tithikr
_tya, both texts con-
cerned with Kashmirian practice. The evidence tends therefore to theconclusion that the brahmanical chaplain retained areas in which healone operated.
At the same time it is possible that there were areas of retrench-ment. For while rituals such as those of the two great festivals ofIndra and Bhadrak�al�ı might have been carried out by both the brah-manical chaplain and the Saiva officiant working simultaneously, evenside by side, it is harder to imagine such co-ordination in the case ofthe intimate domestic rituals to prepare the king’s bed-chamber forhis sleep. Here perhaps the Saiva rite had ousted the brahmanical.
Whether this new institution was present beyond Kashmir I amunable at present to determine. The existence of an early Nepalesemanuscript of the Netratantra, of a manual based on this text for thedaily cult of Amr
_tesvara and Amr
_talaks
_m�ı attributed to the Malla
99 The procedure for these two sacrifices is taught in Atharvavedaparisis_t_a 30,
30b, and 3. Vis_n_udharmottara 2.152.6 requires the Kot
_ihoma annually: sam
_vatsar�at
kot_ihomam
_kury�ac ca ghr
_takambalam (2.161 is devoted to this procedure [ghr
_ta-
kambalakot_ihomah
_/ghr
_takambalas�antih
_’]). The �Adipur�an
_a-Tithikr
_tyav requires two
Laks_ahomas each year and one Kot
_ihoma (ll. 2801–2803): dvau laks
_ahomau kur-
v�ıta tath�a sam_
vatsaram_
prati=ekam_
tu [ko]t_ihomam
_tu yatn�at sarv�abhayapradam=
atharvavedavidhin�a *sammantrya (em : sammantryam_
Ed.) ca [pu]rohitaih_. The
N�ılamata probably required the same (813): sam_vatsarasy�atha *k�aryau laks
_ahomau
(conj.: k�aryo laks_ahomo Ed.) mah�ıks
_it�a=kot
_ihomas tath�a k�arya eka eva dvijottama=
tayor vidh�anam_
vij~neyam_
kalpes_v �atharvan
_es_u ca. Perhaps these references to the
Atharvanic procedures are to Atharvavedaparisis_t_a 30a, 30b, and 31. For references
to Laks_ahomas and Kot
_ihomas performed for the Khmer and Nepalese monarchs
see Sanderson, forthcoming.
272 ALEXIS SANDERSON
king Abhayamalla, of a manual for royal initiation into this cult,and of other textual evidence of the integration of the worship ofAmr
_tesvara and Amr
_talaks
_m�ı into the larger framework of Newar
S�akta Saivism only shows that this tradition took root there in themanner of any other Saiva system, that is to say, as a form of initia-tion and regular worship. It is of course possible that Saiva officiantsin the royal palaces of the Kathmandu valley were serving their kingsin the manner envisaged in the Netratantra, but the mere presence ofa manuscript of that text is not sufficient to prove this, since to befollowing a tradition of initiation and worship based on theNetratantra would be enough to motivate its copying. If evidencewere to come to light that the cult of Amr
_tesvara and Amr
_talaks
_m�ı
did extend in Nepal beyond the shared essentials of initiation andworship to include encroachment into the territory of the brahmani-cal royal chaplain—and this possibility cannot be excluded sincemany Nepalese liturgical texts in Newari and Sanskrit remain to bestudied—then it would be probable that it was established in yetother regions of the subcontinent, at least in the North and East.
APPENDIX
THE PROVENANCE AND DATE OF THE NETRATANTRA
I have asserted above that the Netratantra was composed in Kashmir and at some timebetween about AD 700 and 850, probably towards the end of that period. Here I setforth the considerations that have led me to these conclusions. In the course of doing
so I shall bring forward evidence of the provenance of certain other scriptural texts,notably the Jayadrathay�amala, the Br
_hatk�alottara, and the Vis
_n_udharmottara.
THE ATTRIBUTES IN SAD�ASIVA’S HANDS
Evidence of the Netratantra’s provenance is found in its information on the iconicforms under which Siva and other deities should be visualized. In 9.17–25 it pre-scribes the image of Sad�as iva, the five-faced and ten-armed form under whichSiva is worshipped in the Siddh�anta and under which Amr
_tesvara should be visu-
alized when it is necessary to worship him in that context. Sad�asiva is nearlyalways five-faced and ten-armed in our sources. But there is variety in the prescrip-tion of the objects and gestures to be exhibited by the ten hands. Now the Netra-
tantra shows a strongly distinct tradition in this regard:
tris�ulam utpalam_
b�an_am aks
_as�utram
_ca mudgaram
9.22 daks_in_es_u kares
_v evam
_v�ames
_v evam atah
_param
khet_ak�adarsac�apam
_ca m�atulu_ngam
_kaman
_d_alum
273SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
21d ca mudgaram_
N : samudgaram Ed. 22b v�ames_v evam atah
_param N :
v�ames_u sr_n_v atah
_param Ed. 22c khet
_ak�adarsac�apam
_ca conj. : khet
_ak�adars
_a-
c�apogram_
N : sphet_ak�adarsac�apam
_ca Ed.
In the right hands are a trident, a blue lotus, an arrow, a rosary and a cudgel.
Next [those] in the left hands, as follows: a shield, a mirror, a bow, a citron
and an ascetic’s water-vessel.
This tradition I have seen elsewhere only in the Vis_n_udharmottara:
3.44.18 dasab�ahus tath�a k�aryo devadevo mahesvarah_aks
_am�al�am
_tris�ulam
_ca saram
_dan_d_am athotpalam
19 tasya daks_in_ahastes
_u kartavy�an
_i mah�abhuja
v�ames_u m�atulu_ngam
_ca c�ap�adar�sau kama
_nd_alum
20 tath�a carma ca kartavyam_
devadevasya s�ulinah_
18d saram_
em. : sara Ed.
And Mahesvara, the God of the Gods, should be made ten-armed. O great-
armed [hero], one should place a rosary, a trident, an arrow, a cudgel and a
blue lotus in his right hands. In the left hands of the Trident-wielder, the God
of the Gods, one should place a citron, a bow, a mirror, an ascetic’s water-
vessel, and a shield.100
100 The same attributes are taught in 3.48.9–16 with the information that theten are five pairs, one held by each of the five deities said here to be fused in thefive-faced Sad�asiva image (3.48.3–8): Mah�adeva facing forward (the rosary and
ascetic’s water-vessel), Sad�asiva above (the bow and the arrow), Bhairava lookingto the right (the cudgel and citron), Nandin at the rear (the shield and trident),and the goddess Um�a looking left (the mirror and blue lotus). Addorsed images of
Siva with the lateral faces of Um�a, Bhairava/Mah�ak�ala and Nandin behind are afeature of local Kashmirian tradition as seen in material evidence of the sixth toseventh centuries. We have examples in stone from the Siva temple at Fattegarh(Siudmak 1994, pl. 39a,b) and the Sailaputr�ı temple in Wushkur village ( Hus
_ka-
pura) (Siudmak 1994, pl. 40a,b), and a related bronze (Pal 1975, pl. 4a,b). The tra-dition is also represented in Kashmirian praise of the holy site of Bh�utesvara, alsocalled Nandiks
_etra, located below Mount Harmokh. See Nandiks
_etram�ah�atmya
f. 14r1–4 (vv.165–168): sarvanandimah�ak�aladev�ıvadanaman_d_itam=bh�utesvaram
_bh�u-
tapatim_
dr_s_t_v�a martyo vimucyate=pascime vadane v�ıra mama vatsyasi yat sahe=
bh�utesvarah_sarvabh�utah
_sut�ırth�antargato vibhuh
_=sr�ıkan
_t_hah_p�urvavadane mah�ak�alo-
’tha daks_in_e=pa�scime nandirudras tu dev�ı saumye pratis
_t_hit�a=bh�utesvarasya devasya
nandiks_etramah�aphalam=dr
_syante vadanes
_v ete dev�ınandimah�asiv�ah
_‘Mortals are lib-
erated by seeing Bh�utesvara, the Lord of Creatures, adorned with the faces ofSarva [=Siva], Nandin, Mah�ak�ala and the Goddess. I allow, O hero, that you
should reside in my face at the rear. Bh�utesvara, [though he] is all things, theall-pervading Lord, resides within [this holy place] Sut�ırtha. Sr�ıkan
_t_ha [=Siva]
is established in his east-facing face, Mah�ak�ala in the south-facing, Nandirudra
in the west-facing [at the rear] and the Goddess in the north-facing. In the facesof the god Bh�utesvara one beholds as the great reward of the Nandiks
_etra these
[four]: the Goddess, Nandin, Mah�a[k�ala] and Siva’; cf. N�ılamata 1119c–1120.
274 ALEXIS SANDERSON
and the Sarv�avat�ara (f. 8v14–15):
s�ulotpales_udan
_d_�aks_as�utrakodan
_d_adh�arin
_e
kaman_d_alukarasph�araphaladarpan
_ap�a
_naye
[Obeisance] to [him] who carries a trident, a blue lotus, an arrow, a cudgel, a
rosary and a bow, whose hands hold an ascetic’s water-vessel, a shield, a fruit,
and a mirror.
The Sarv�avat�ara can only have been written in Kashmir, since its subject matter is
restricted to the glorification of Saiva sacred sites in that region, most of whichhave no place on the pan-Indian pilgrimage map.101
That the Vis_n_udharmottara was written in Kashmir or a neighbouring region
follows from a number of factors. (1) There is a strong correlation between the
T�ırthas invoked in the Vis_n_udharmottara’s Mantra for the Royal Consecration
(r�ajy�abhis_ekamantrah
_) taught in chapter 22 of its second Khan
_d_a and those sacred
sites, mostly Kashmirian, mentioned in the local N�ılamata. (2) There is close agree-ment between chapter 35 of the second Khan
_d_a (str�ıdevat�ap�ujananir�upan
_am) and
the religious calendar of Kashmir taught in the N�ılamata. (3) Where the Vis_n_u-
dharmottara prescribes domestic Vaidika rites it adheres to the distinctive proce-dures of the K�at
_hakagr
_hyas�utra, also called Laug�aks
_igr_hyas�utra, the authority
followed for these rituals by the brahmins of Kashmir. Thus its Vaisvadeva deities(2.92.3–15) are those of the Kashmirian brahmins as prescribed in K�at
_hakagr
_hya-
s�utra 4.14.1–20.102 The same applies to the Sr�addha rituals, as can be seen by com-paring Vis
_n_udharmottara 1.140.8–43 with Laug�aks
_igr_hyas�utramantrabh�as
_ya, vol. 2,
pp. 332–363. (4) It fuses the old Kashmirian iconography of the Siva image, with itssecondary faces of Um�a, Bhairava and Nandin,103 with the pan-Indic tradition ofthe Saiva Mantram�arga, which equates the five faces of Sad�asiva with the five Vedic
Brahmamantras (3.48.1–6). And (5) The principal Vis_n_u form in its prescription of
the images of deities is the four-faced Vaikun_t_ha, in which the forward-facing
101 Among the sacred places of Kashmir praised in this text are Mah�adevagiriand Sr�ıdv�aragiri, the mountain-ridge along the east side of the D
_al lake, with its
various T�ırthas, notably Jyes_t_hesvara and Tripuresvara at Tripar and Suresvar�ı
near Isha_ ba_r.
102 For the Kashmirian Vaisvadeva ritual see K�asm�ırikakarmak�an_d_apaddhati f.
192v and the Saivavaisvadevavidhi.103 See n. 100 above.
275SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
anthropomorphic face is flanked by the faces of Var�aha and Narasim_ha, with the
face of the sage Kapila at the rear (3.44.9c–13, 3.85.42c–45). This is the principalKashmirian Vis
_n_u image and it is seldom seen elsewhere.
The inference that the Netratantra’s prescription of the hand-attributes of Sad�a-siva is that of a Kashmirian tradition outside the Saiddh�antika mainstream isstrengthened by two further items of evidence. The first is that this type of Sad�asiva has left a trace in no Saiddh�antika scripture other than the Br
_hatk�alottara. That
teaches a close variant of the hand-attributes seen in the Netratantra—it differsonly in that a sword (khad
_gah_) takes the place of the cudgel (dan
_d_ah_) in one of the
hands104—and it contains other indications that it was redacted in Kashmir or
under Kashmirian influence, notably the imprint of the non-dualistic S�akta Saiva
doctrine and terminology seen in the Spandak�arik�a, a seminal work of that tradi-tion composed in Kashmir towards the end of the ninth century.105
The second is that we have another variant of the Netratantra’s Sad�asiva in
the Kashmirian liturgical tradition. This is the image of Bahur�upabhairava, who isworshipped with his consort M�ay�adev�ı in the Kashmirian Saiva cremation ritual(sivanirv�an
_avidhih
_). In the hands of this variant a sword takes the place of the blue
lotus (n�ılotpalam) and the gesture of bestowing boons (varadamudr�a) that of the
mace (gad�a).106
104 Br_hatk�alottara B f. 17r5–6: n�ılan�ırajan�ar�acakhad
_g�aks
_avalay�abhayam=satris�u-
lam_
harasyoktam_
*daks_in_e (em. : daks
_in_a Cod.) pa~nca b�ahavah
_=b�ıjap�uram
_dhanus
carma varada~n ca kaman_d_alum
_=v�ame tu devadevasya b�ahavah
_pa~nca k�ırtit�ah
_.
105 For this imprint see Sanderson 2001, pp. 17–18, n. 19. As further evidenceof the Br
_hatk�alottara’s Kashmirian origin one may cite its knowledge of the pair-
ing (to be discussed below) of the two sets of four goddesses associated in Kashmi-rian tradition with the V�ama and Daks
_in_a divisions of the Saiva scriptures. Also
consistent with this origin is its use of the term kh�arkhodah_, to be discussed below,
and its dependence on the P�ancar�atrika scripture Jay�akhyasam_hit�a demonstrated in
Sanderson, 2001, pp. 38–41. That that scripture was produced in Kashmir is highly
probable, though not certain.106 See Sivanirv�an
_avidhi p. 235, l.8–p. 246, l.8 (mun
_d_ay�agah
_and m�ay�aj�alap�uj�a). For
the visualization see p. 237, ll. 1–4: sitam_
tryaks_am_
pa~ncavaktram_
dasab�ahum_sasaktikam=s�ul�aks
_as�utres
_ukhad
_gavarair daks
_akarair vr
_tam=m�atulu _ngadhanuscarma-
kumbhadarpan_av�amakaih
_. The names of the two deities are revealed in their Mantras:
9-M_½¼H-S-KS
_-M-L-V-R-Y�UM
_� BAHUR�UPABHAIRAV�AYA SVADH�A NAMA
_H (p. 237,
l. 4) and HR�IM_
M�AY�ADEVYAI SVADH�A NAMAH_(p. 237, l. 9). They are worshipped
surrounded by the eight Mothers and the eight Bhairavas (p. 237, l. 9 – p. 238, l. 7).
276 ALEXIS SANDERSON
THE GODDESSES SIDDH�A, RAKT�A, SUS_K�A
AND UTPAL�A
In chapters 10 and 11 the Netratantra teaches substitutes for the standard icon ofAmr
_tesvara to be adopted when the officiant has reason to adapt this Mantra-
deity to the context of the V�ama and Daks_in_a Saiva systems. In these cases
Amr_tesvara should be visualized as Tumburu and Bhairava respectively, those
being the presiding deities of those systems, and he should be worshipped sur-
rounded by the retinues of Saktis proper to those two. These, to mention onlythe primary circuit of goddesses, are four in each case: Jay�a, Vijay�a, Ajit�a andApar�ajit�a in the V�ama (11.12–18) and Siddh�a, Rakt�a, Sus
_k�a and Utpalahast�a
(10.17c–34) in the Daks_in_a. The first set of goddesses is given as the inner reti-
nue of the V�ama’s Tumburu in all accounts of this cult, and these come fromwidely separated areas of the Indic world: Kashmir, Gilgit, Nepal, Kerala, Ta-milnadu, Cambodia and Bali. But the second set’s association with the Daks
_in_a’s
Bhairava, indeed the second set itself, is far less well attested, in spite of the fargreater abundance of the textual materials that have survived from this division.Now the only other texts known to me in which these four goddesses, or variants of
them, are mentioned are Kashmirian: Jayaratha’s commentary on Abhinavagupta’sTantr�aloka, the �Anandesvarap�uj�a, the Br
_hatk�alottara, the Moks
_op�aya, and the
Jayadrathay�amala.
Where the Kashmirian Abhinavagupta (fl. c. 975–1025) says that in the Kaulaworship of the Trika the deities that surround the central triad of the Goddesses
may be twelve, sixteen, four, or indeed whatever set one prefers,107 his compatriotJayaratha (fl. c. 1250) comments: ‘‘The four here are either those beginning withSiddh�a or those beginning with Jay�a’’;108 and where Abhinavagupta describes
wine as ‘‘Mah�abhairava fully radiant with the four Saktis’’,109 Jayaratha com-ments that the four to which he refers are ‘‘the set of four beginning with Siddh�a’’,‘‘for’’, he adds, ‘‘these are white, red, yellow and black in colour’’.110 The �Anandes-
107 Tantr�aloka 29:51 : antar dv�adasakam_
p�ujyam_
tato ’s_t_�as_t_akam eva ca=catus
_kam
_v�a yatheccham_
v�a k�a sam_khy�a kila rasmis
_u. The sequel reveals that the expression
as_t_�as_t_akam is to be taken as a Dvandva compound meaning ‘eight-and-eight’,
referring to the eight Kaula Mothers with their eight Bhairavas.108 Tantr�alokaviveka ad 29.51c: catus
_kam iti siddh�adi jay�adi v�a.
109 Tantr�aloka 37.42d: sakticatus_t_ayojjvalam alam
_madyam
_mah�abhairavam.
110 Tantr�alokaviveka ad 37.42d: sakt�ıti �siddh�adicatus_kam (em.: siddh�acatus
_kam
Ed.). tad dhi sitaraktap�ıtakr_s_n_avarn
_am. The colours were no doubt thought to be
significant as those of four varieties of wine.
277SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
varap�uj�a has this same set of four worshipped as the retinue of �Anandesvarabhair-ava and his consort Sur�adev�ı, the deities of wine.111
The Br_hatk�alottara covers the four goddesses beginning with Jay�a and the four
beginning with Siddh�a, without associating them with Tumburu and Bhairava or
the V�ama and Daks_in_a divisions, in a chapter devoted to the iconography of
images of the Mothers.112
The Moks_op�aya identifies the four beginning with Siddh�a as those that sur-
round Bhairava in the Daks_in_a division (daks
_in_asrotah
_), pairs them with the four
surrounding Tumburu in the V�ama division (v�amasrotah_), and asserts that these
eight are the foremost of all the Mother goddesses.113
In the Jayadrathay�amala the two sets of four form the sixty-first and sixty-second cycles of the eighty-one cycles of the goddess R�avin
_�ı that occupy ever
higher centres along the axis of the worshipper’s body in the fourth of the four
divisions of six thousand stanzas (S_at_kas) that comprise that text. The first set, in
the order Sus_k�a, Siddh�a, Utpal�a and Rakt�a, is said there to comprise the Saktis of
111 �Anandesvarap�uj�a f. 59[2]r4–6: bh�an_d_atale: ha-sa-ra-ks
_a-ma-la-va-ya-�um
_�ana-
ndesvarabhairav�aya vaus_at_: sa-ha-ra-ks
_a-ma-la-va-ya-�um
_sur�adevyai vaus
_at_: om
_* si-
ddh�ayai (conj.: siddh�arth�ayai Cod.) vaus_at_: evam
_sus_k�ayai vau. rakt�ayai: utpal�ayai.
Cf. in a prescription for the worship of �Anandesvara after the completion of thefire-sacrifice Agnik�aryapaddhati B f. 130r10: bh�an
_d_e devyas catasro.
112 The chapter, called m�atr_bhairavavartan�a in its colophon, unnumbered in the
manuscripts, is the seventy-seventh by my count. The section on the two sets offour is as follows (Br
_hatk�alottara A f. 252r2–4; B f. 219r4–219v2):27 caturbhuj�a
caturvaktr�a jay�a kundendusannibh�a=is_ukodan
_d_asam
_yukt�a pretasth�a nr
_tyatatpar�a=
28 evam_
tu vijay�a k�ary�a rakt�abh�a�svoparisthit�a/evam_
jayant�ı b�ıbhats�a s�urya*bh�ımo-paristhit�a (conj.: bh�ımaparisthit�a AB)/29 *megha (conj.: moha AB) sth�a *cotpal�ak�ar�a(A: cotpat�ak�ar�a B) nr
_tyant�ı kr
_s_n_avarcas�a (A: kr
_s_n_avarcas�ı B)=apar�ajit�a pra-
kartavy�a n�an�atody�anuvartin�ı=30 pretasth�a (A: pratasth�a B) caiva nirm�am_s�a siddh�a
kundendusannibh�a=khad_gacarma *dhar�ı (A: dhur�a B) dev�ı ks
_urik�amun
_d_abh�us
_it�a=31
evam_
rakt�a kim_
tu *rakt�a (A: bhukt�am_
B) sus_k�am
_*k�al�ım
_(conj.: k�antim
_) tu
k�arayet=*utpalaprabhavadan�a (conj: utpalaprabh�avadan�a A: utpalaprabhav�ad�at�a B)utpal�a _ngotpalasthit�a.
113 The relevant passage of this unpublished part of the Moks_op�aya has been
edited in Hanneder, 1998a, p. 69: jay�a ca vijay�a caiva jayant�ı c�apar�ajit�a=v�amasro-togat�a et�as tumburum
_rudram �asrit�ah
_=siddh�a sus
_k�a ca rakt�a ca utpal�a ceti devat�ah
_/
sroto daks_in_am �asritya bhairavam
_rudram �asrit�ah
_=sarv�as�am eva m�at�r
_n_�am as
_t_�av e-
t�as tu n�ayik�ah_.
278 ALEXIS SANDERSON
the Lord of the Daks_in_a (daks
_in_esvarah
_), and the second set, in the order Jay�a, Vi-
jay�a, Apar�ajit�a and Jayant�ı, that of the Lord of the V�ama (v�amav�ıresvares�anah_).114
The four beginning with Siddh�a appear again in that Sat_ka as the first of a series of
32 Saktis comprising the sequence called ‘With-Support’ (s�alambakramah_) in a vari-
ant of the Krama system of the K�al�ıkula.115
In the second Sat_ka they (but with Rakt�a under the name C�amun
_d�a) are the
first four of the twelve Mothers that form the retinue of K�al�ı Catuscakresvar�ı‘Ruler of the [Three] Cycles of Four’. The other eight are the four beginning withJay�a, followed by V�am�a, Jyes
_t_h�a, Raudr�ı and Bhadrak�al�ı.116
That Jayaratha’s commentary on the Tantr�aloka is Kashmirian requires no
demonstration. The �Anandesvarap�uj�a is part of the corpus of Kashmirian ritualtexts, and the form of worship it teaches is a regular ancillary element in the Kash-mirian Saiva ritual of initiation.117 The Kashmirian origin of the Br
_hatk�alottara has
114 Jayadrathay�amala, S_at_ka 4, f. 91v7–92r: v�amadaks
_in_ac�ar�abhy�am
_kalitau cakra-
*n�ayakau ðconj: : n�ayike Cod.)/yau pam_cadh�a sures�ani tatra s�a sphurit�a yad�a=tad�a
vyakt�avyaktatar�a sr_s_t_isam
_h�ara*k�arik�a (corr.: k�arak�a Cod.)/karam
_kin_�ı mah�araudr�a
daks_in_esvara *sam
_yut�a (conj.: s�am
_pratam Cod.)/sus
_k�asiddhotpal�arakt�arasmin�atham
_param_
mahat=tatrodgatam_
k�alan�aya jagaty asmin car�acare=v�amav�ıresvares�anam_
jay�aca vijay�a tath�a=apar�ajita jayam
_t�ı s�a cety evam
_pam
_cakam
_smr_tam_=atra *sr
_s_t_ivat�ı
(em.: siddhivat�ı Cod.) dev�ı prodit�a paramesvar�ı=sam_h�ara*dh�amni (em.: dh�astri Cod.)
y�a k�al�ı s�a vyakt�a p�urvvacakratah_=evam
_c�aradvaye k�al�ı sr
_s_t_isam
_h�arak�arin
_�ı.
115 Jayadrathay�amala, S_at_ka 4, f. 202r2–6: s�alambam evam �ay�ati *vyaktim
_(em.:
vyaktih_
Cod.) suravar�arcite=atra rasmisam�uhasya vibh�agam_
sr_n_u s�am
_pratam=si-
ddh�a rakt�a susus_k�any�a utpal�a parik�ırttit�a=k�al�ı ca k�alar�atr�ı ca k�ala*dh�ar�a(em.: dh�are
Cod.) kalesvar�ı=sim_havaktr�a ca m�arj�ar�ı us
_t_r�a k�ap�alin�ı tath�a=khara*r�up�avir�up�a ðem: :
rup�a ca virup�a Cod.) ca mes_ar�up�a mahorag�a=rakt�aks
_�ı raktav�as�a ca lam
_bakarn
_n_�ı tat
haiva ca=pr_thodar�ı tv ekanetr�a lokan�ath�a bhayam
_kar�ı=ul�ukavadan�a c�any�a *kolavaktr�a
(conj.: k�alavaktr�a Cod.) ca khim_khin�ı= karam
_k�a bhadrak�al�ı ca tathaiv�any�a mah�a-
bal�a=bharud_�a hy at
_t_ah�as�a ca r�aks
_as�ı hy �asur�ı tath�a=et�a eva smr
_t�a rasmyo *dvidh�a-
s_t_�as_t_akabhedatah
_(conj.: dr
_s_t_v�aj~n�as
_t_akabhedatah
_Cod.)=kulavidy�ap�urvvayukt�a svan�a-
makr_ta*madhyak�ah
_(conj.: madhyag�ah
_Cod.)/*p�ad�ant�ah
_(conj.: pad�arth�a Cod.)
p�ujan�ıy�as t�ah_
samyak*s�alambasiddhid�ah_
(em.: pr�alam_basiddhid�ah
_Cod.) sphuracca-
krakram�anta*sth�ah_(corr.: sth�a Cod.) svasth�ane *pravijr
_mbhit�ah
_(corr.: prajijr
_m_bhitah
_Cod.)=praks_�ın_abh�ava*vr
_ndaugh�ah
_(corr.: vr
_m_daugh�a Cod.) sarvv�ah�ar�ah
_sulampat
_�ah_/
iti jayadrathay�amale s�alam_bacakrakramavidhibhedah
_.
116 Jayadrathay�amala, S_at_ka 2, f. 12r8–9 (3.45–47b): tatrasth�am
_p�ujayen mantr�ı
antak�antakar�ım_
par�am=dv�adas�are tata�s cakre sam_p��ujy�a m�atarottam�ah
_=c�amun
_d_�a ca
tath�a sus_k�a siddh�a caivotpal�a tath�a=jay�a ca vijay�a caiva jayant�ı c�apar�ajit�a=v�am�a
jyes_t_h�a tath�a raudr�ı bhadrak�al�ı gan
_�ambik�a.
117 See Kal�ad�ıks_�avidhi f. 58r9–10, in the context of the concluding of the rites of
the first day (adhiv�asadinam): kr_tv�a ca vaisvadev�anandesvarabhairavap�uj�adi br�ahma-
n_ap�ujanam
_ca kr
_tv�a; f. 235r16–v3, in the context of the closing rites of the last day
of the initiation: tatah_
pr_thaksthale �anandesvarabhairavap�uj�am
_taduktavidhin�a kr
_tv�a
ks_etrap�al�am
_s c�agrelikhitaks
_etrap�alapaddhatikramen
_a sam
_p�ujya.
279SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
been argued above. That the Moks_op�aya was composed in Kashmir has been estab-
lished by Jurgen Hanneder, who has also tied the time of its composition to afew years immediately after the reign of the Kashmirian king Ya�saskara (r. AD 939–948).118
As for the provenance of the Jayadrathay�amala, we must distinguish betweenthe first S:at
_ka of six thousand verses and the eighteen thousand verses of the
remaining three. The first was originally an independent whole. It presents itself as
such, predicts no sequel, and is distinct from the other S: at_kas, which are closely
related to each other in style, terminology and concepts.119 In the first I see noth-ing that enables us to fix the region of its composition. But the rest of the later
text shows clear signs of Kashmirian origin. In the second S:at_ka the only sign I
see is the collocation of these two sets of four Saktis. But in both S:at_kas 3 and 4
there is further evidence.When the fourth S:at
_ka sets out the procedures and rituals that must accom-
pany the copying of a manuscript of the Jayadrathay�amala it assumes that thecopying will be done on sheets of birch bark (bh�urjapatr�an
_i).120 This was the stan-
dard writing material only in Kashmir and adjacent areas of the northwest.
The third S:at_ka contains a chapter devoted to the use of the Mantra of the
goddess Ghoraghoratar�a in order to gain access to the subterranean paradises ofP�at�ala (p�at�alasiddhih
_). It lists seventeen sites where there are Sr�ımukhas, special
apertures in the earth (bilam) through which this feat can be achieved. The firstseven are at sites of pan-Indian fame: Pray�aga, Gay�a, Sr�ısaila, Man
_d_alesvara,
Hariscandra, the Narmad�a river, and the K�alinjara mountain. The last ten are
118 Hanneder, 2003, pp. 40–52.119 Sanderson, 2002, p. 2 and n. 13.120 Jayadrathay�amala; S:at
_ka 4, f. 208v4: bh�urjapatr�an
_i c�amam
_trya kr
_takautu-
kamam_galah
_=likhed varn
_�ani ‘Having empowered the leaves of birch-bark with the
Mantra and tied a protective thread about his wrist he should trace the letters’.
280 ALEXIS SANDERSON
said to be at S�ulabheda, Vijaya, Var�aha, Jyes_t_ha, the Uttaram�anasa [lake], near
Tu _ngasv�amin, on Mah�adevagiri, at P�atram�ula, Padmasaras, and Mah�am�a-y�uraka.121 Two of these ten, Mah�am�ay�uraka and Tu_ngasv�amin, are unknown tome, but the remaining eight can be identified as sacred sites within Kashmir. S�u-labheda, also known as S�ulagh�ata, is the spring of the N�aga N�ıla (n�ılakun
_d_am),
the source of the river Vitast�a/Vyath/Jhelum, so named because Siva is believedto have split open the earth (-bheda) here by striking it (-gh�ata) with his trident
(S�ula-) so that the river could emerge from the underworld;122 Vijaya is Vija-yaks
_etra on the right bank of the Vitast�a, the site of the temple of Siva Vijayes-
121 Jayadrathay�amala, S:at_ka 3, f. 162r4–7 (in Ghoratar�as�adhana; P�at�alasi-
ddhipat_ala): evam
_bilavibh�agam
_sy�ad deses
_v adhun�a mucyate=pray�age ca gay�ay�am
_ca
sr�ısaile man_d_alesvare=*hariscandre (em. : hariscam
_dra Cod.) narmad�ay�am
_tath�a
k�ali~njare girau=kasm�ır�ay�am_s�ulabhedam
_*toyap�urn
_am_
(conj. : rotap�urn_n_a Cod.) bilo-
ttamam_=vijaye ca var�ahe ca jyes
_t_he cottaram�anase=tu _ngasv�amisam�ıpe tu mah�adevagi-
rau tath�a=p�atram�ule padmasare mah�am�ay�urake tath�a=evam�adis_u deses
_u sr�ımukh�as te
prak�ırtit�ah_. The term sr�ımukham, here masculine, denotes the superior among such
apertures. Ibid. f. 162r4–5: uttamam_sr�ımukham
_j~neyam
_bahugarbhapur�acitam=ma-
dhyamam_
bilasam_j~nam
_sy�ad antah
_purasatair yutam=s�am�anyam
_vivaram
_proktam
_siddhadravyasat�avr_tam.
122 N�ılamata 1302, 1389; Haracaritacint�aman_i 12.16c–17, referring to it as ‘‘the
supreme aperture’’ (bilam uttamam): *n�ılakun_d_am_
(em. : n�ılakan_t_ham
_Ed.) vitast�a-
khyam_s�ulagh�atam iti tribhih
_=*abhidh�anaih
_(em. : abhidh�anam
_Ed.) prasiddham
_tad
ady�api bilam uttamam; Vijayesvaram�ah�atmya f.11v4–6: s�ulena bhittv�a p�at�alam_tasm�at
sth�anavar�ac chubh�at=uddhr_t�a s�anad�ı pun
_y�a paramabrahmac�arin
_�ı=var�ahatanay�a devi
muktid�a sarvajantus_u=s�ulabheda iti khy�atam
_tat t�ırtham
_parvat�agrimam; Stein, 1961,
vol. 2, p. 411. It is located near V�ern�ag.
281SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
vara, recognized beyond the Kashmir valley as the principal Siva of the region;123
Var�aha (Var�ahaks_etra/Var�ahat�ırtha) is the site of the shrine of Vis
_n_u �Adivar�aha
just above the gorge through which the Vitast�a leaves the valley;124 Jyes_t_ha is the
site of Siva Jyes_t_hesvara adjoining that of �Siva Bh�utesvara (Butish�er), a major
Kashmirian pilgrimage site below the Harmo�kh glaciers;125 Uttaram�anasa is theGan_ gabal lake at the foot of those glaciers;126 Mah�adevagiri is the mountainpeak of that name located in the ridge that separates the valleys of the Sindhu
123 See R�ajatara _ngin_�ı 1.38; Skandapur�an
_a, N�agarakhan
_d_a (6), Adhy�aya 109 (list-
ing the names of the Sivas at each of 68 Sivat�ırthas throughout the subcontinent),13a: vijayam
_caiva k�asm�ıre; N�ılamata 1056, 1303; Haracaritacint�aman
_i, chapter 10;
Tantr�aloka 37.39cd; Kath�asarits�agara 39.36; 51.48; 66.5; Desopadesa 4.28; Stein,
1961, vol. 2, pp. 463–464. This is the eponymous Siva of the modern town ofV�ejabr�oru ( Vijayabhat
_t_�araka). The view that this is the pre-eminent Li _nga of Siva
in Kashmir is also expressed in the Vijayesvaram�ah�atmya (f. 2r3: kasm�ıraman_d_ale
pun_ye vijaye li _ngam uttamam), which also claims pan-Indian pre-eminence for the
site by saying that of all the Sivaks_etras of the subcontinent (Kum�ar�ıdv�ıpa), the five
sets of eight and the sixty-eight—for these see Sanderson, 2005, nn. 199–203 and207–209—four are supreme: Avimukta (in Benares), Mah�ak�ala (in Ujjain), Varis
_t_ha
(in Is_t_ak�apatha/Is
_t_ik�apatha?), and Vijayesvara (f. 2r15–v8): ye ca pa~nc�as
_t_ak�a guhy�a
ye v�as_t_�as_as_t_isa _nkhay�a=sth�an�as susobhan�a hr
_dy�ah
_sarvak�amaphalaprad�ah
_=jant�un�am
_bh�avayukt�an�am_bhogad�a muktid�as sad�a=tes
_�am_
madhy�an mah�adevi proktam_ks_etraca-
tus_t_ayam=avimuktam
_mah�ak�alam
_varis
_t_ham
_vijayesvaram=�agneyaman
_d_al�antasstham
_kum�ar�ıdv�ıpam �asritam=bharatasya tu madhye sya catv�aro vasthit�a iha=mah�asmas�a-nasam
_j~n�as ca sth�an�a ete prakalpit�ah
_=tes
_�am_
vibh�agam_vaks
_y�ami yath�akramam anu-
ttamam=avimuktah_
sthitah_
p�urve mah�ak�alo tha daks_in_e=*varis
_t_hah_(corr. : varis
_t_ah_Cod.) pascime bh�age tasy�ante vijayesvarah
_.
124 See R�ajatara _ngin_�ı 6.206; 7.1310; N�ılamata 1158–59; Haracaritacint�aman
_i
12.43; Kath�asarits�agara 39.37 (v�ar�aham_ks_etram
_); Stein, 1961, I, p. 251, n. on
6.186. This is the source of the name of the surrounding town of Warahmul/B�ar�am�ula ( Var�aham�ula).
125 R�ajatara _ngin_�ı 1.113; Nandiks
_etram�ah�atmya of the Sarv�avat�ara, ff. 12r7–15r1
(vv. 142–175); N�ılamata 1032, 1111–1136 (Bh�utesvaram�ah�atmya); Kath�asarits�agara39.36 (Nandiks
_etra); 51.48 (Nandiks
_etra); Stein, 1961, II, pp. 407–408.
126 See N�ılamata 899, 960, 1124–30; Haracaritacint�aman_i 4.87ff.; Kath�asarits�agara
39.38; Jayantabhat_t_a, Ny�ayama~njar�ı, vol. 2, p. 376, l.14; Moks
_op�aya, Vair�agya-
prakaran_a 1.2.36b; the Northern recension after Mah�abh�arata 13.26.56 (on K�alo-
daka, Nandikun_d_a, Uttaram�anasa and the image of Nand�ısvara [at Bh�utesvara/
Jyes_t_hesvara]); K�urmap�ur�an
_a 2.36.41c–42b; Stein, 1961, vol. 1, p. 111, n. on 3.448.
282 ALEXIS SANDERSON
and the Arrah;127 Padmasaras is the great lake at the northern end of the valleynow known as the W�olur;128 and P�atram�ula can only be the T�ırtha of the N�agaP�atra that the N�ılamata places on the course of the Vitast�a between its conflu-ence with the Sindhu and its entry into the W�olur Lake.129 The site appears
under the name P�atram�ula in the Kashmirian pilgrimage text Vitast�am�ah�atmya,which tells us that it is here that the Greater Gan_ g�a (Mah�agan_ g�a) emerged intothe world from the subterranean paradise and that it is here that the demon
Gay�asura disappeared into that underworld when pursued and struck by Vis_n_u
with his mace.130
Two of these sites, Vijaya and Uttaram�anasa, were famous outside Kashmir,131
but others, such as P�atram�ula, are registered only in local tradition. It is thereforevery unlikely that this is the work of any but a Kashmirian addressing a Kashmirianreadership.
Since, therefore, the only sources other than the Netratantra that know the setof the four goddesses Siddh�a, Rakt�a, Sus
_k�a and Utpal�a are Kashmirian, and since
there is an abundance of non-Kashmirian sources in which their absence is signifi-
cant, it is highly probable that the Netratantra too is a work of this region.
THE ICONOGRAPHY OF VIS_N_U
Further support for this conclusion is provided by the text’s iconography of Vis_n_u
and Brahm�a. Of its forms of the former the first is one-faced and four-armed,holding the conch, discus, mace and lotus (13.2–4). The second (13.5–9) is a three-
faced version of the four-faced Vaikun_t_ha, with a central anthropomorphic head
flanked by those of the Boar (Var�aha) and the Man-Lion (Narasim_ha), sur-
rounded by the goddesses Laks_m�ı, K�ırti, Jay�a, and M�ay�a, and accompanied by
his consort Laks_m�ı (13.5–9). The third (13.10–13b) is a naked, ithyphallic, eight-
127 See N�ılamata 1337; Haracaritacint�aman_i 10.258; Sarv�avat�ara ff. 3–5 (Adhy�aya
3); Kath�asarits�agara 51.48; Sivas�utravimarsin�ı, p. 1; Stein, 1961, vol. 2, p. 422.128 N�ılamata 985–997, 1351, 1353; R�ajatara_ngi _n�ı 4.592–617; Stein, 1961, vol. 2,
pp. 423–24. It is more usually called Mah�apadmasaras (‘the lake of the [N�aga]Mah�apadma’), but we see Padmasaras in R�ajatara _ngin
_�ı 8.2421.
129 N�ılamata 1349–50.130 Vitast�am�ah�atmya (assigned to the Bhr
__ng�ısasam
_hit�a, a traditional locus of attri-
bution for Kashmirian Puranic materials), A f. 25r9–11: anvadh�avac ca tad raks_o
devadevo jan�ardanah_=gaday�a c�api tam
_jaghne raks
_asam
_bhagav�an harih
_=anvadravat
punas tam_ca *y�avad (conj. : t�avad Cod.) vai p�atram�ulakam=tatraiva raks
_ah_p�at�alam
_pradadr�ava mahesvari=tatrodbh�ut�a mah�aga _ng�a p�at�al�aj jagad�ı�svari=tatra sn�atv�a *nare(conj. : naro Cod.) devi muktibhukt�ı na sam
_sayah
_.
131 See nn. 123 and 126 above.
283SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
armed child riding a ram, playing with women, and attended by four naked god-desses, Karp�ur�ı, Candan�ı, Kast�ur�ı, and Ku_nkum�ı.132
Judging from surviving stone and bronze sculptures we see that the first and thesecond were the standard forms of Vis
_n_u in Kashmir133 and that the second is
found almost only there, becoming four-faced from about the middle of the ninthcentury through the addition of an addorsed head of the sage Kapila.134 Ks
_emar�aja
identifies this form and its four attendant goddesses as following the prescription of
the P�a~ncar�atrika Jay�akhyasam_hit�a, and he is right to have done so, except that this
text teaches the four-faced form with those goddesses. The Netratantra’s image re-flects Kashmirian practice prior to the addition of the Kapila face.135
132 Netra 13.10–13b: athav�as_t_abhujam
_devam
_p�ıtavarn
_am_
susobhanam=mes_opa-
risthitam_
*devi (Ed. : devam_N) *digv�asam �urdhvali _nginam (corr. :digv�asam
_m �urdhva-
lim_ginam
_N: digvastram
_cordhvali _nginam Ed.) /11 sr
__ngam
_vas_t_abhya caikena *vy�a
khy�anodyatap�an_ikam (conj. : coy�anodyatap�an
_ikam N: cey�arodyatap�an
_ikam Ed.)=
b�alar�upam_
yajen nityam_
*kr�ıd_am�anam
_hi yos
_it�am
_(N: kr�ıd
_antam
_yos_it�am
_gan_aih_Ed.)/12 caturdiks
_u sthit�a devyo*digv�as�as tu ðN: digambara Ed.) manoram�ah
_=ka-
rp�ur�ı candan�ı caiva kast�ur�ı ku _nkum�ı tath�a= 13 *tadr�upadh�arin_�ır dev�ır (conj. : tadr�u-
padhar�ari _n�ım_
dev�ım_
N : tadr�upadh�arik�a devyo Ed.) icch�asiddhiphala *prad�ah_
(Ed. :prad�a N).
133 For instances of the first see Pal, 1975, pl. 10 (9th century) and Siudmak1994, pl. 31 (c. 500–550), pl. 34 (c. 550–600), pl. 38 (c. 525–550), pl. 50 (c. 600–625), pl. 52 (c. 575–600), pl. 55 (early 7th century), pl. 58 (c. 525–550), pl. 60 (c.600–625), pl. 72 (c. 675–700), pl. 123 (c. 825–850) For instances of the second see
Pal, 1975, pl. 9 (three-faced, c. 800), pl. 12a,b (four-faced, 11th century), pl. 84a,b,c(from neighbouring Chamba, four-faced, 9th century), and Siudmak, 1994, pl. 118(three-faced, c. 700–725), pl. 120 (three-faced, c. 775–800), pl. 121 (three-faced, c.
775–800), pl. 122 (three-faced, c. 825–850), pl. 124 (four-faced, c. 850), pls. 140–143(all four-faced, c. 850–55, Avantisv�amin temple), pls. 155–56 (both c. 875–900), pl.170 (c. 1000–1025), Huntington, 1985, fig. 17.19 (c. 12th century).
134 Siudmak, 1994.135 Netroddyota after 13.8c–9: evam
_sr�ıjay�asam
_hit�adr
_s_t_yoktv�a. Ks
_emar�aja’s Jay�a-
sam_hit�a is evidently the Jay�akhyasam
_hit�a. The Kashmirian Bh�agavatotpala refers
to the work as Jay�a in his Spandaprad�ıpik�a, p. 91 ( proktam_
hi sr�ıjay�ay�am. The
citation that this introduces is Jay�akhyasam_hit�a 10.69). The visualization of the
four-faced Vaikun_t_ha, the central deity of that scripture, is prescribed in 6.73–76.
It is highly improbable that Ks_emar�aja knew this text in an earlier redaction in
which the image had only three faces as in the Netra. For the text does not merelyteach a four-faced image. It teaches a system of rites in which the distinctionbetween the three subsidiary faces of Narasim
_ha, Var�aha and Kapila, each with
its own Mantra, is central.
284 ALEXIS SANDERSON
The third form, which is not named in the Netratantra, is said by Ks_emar�aja to
follow the prescription of the M�ay�av�amanasam_hit�a, another P�a~ncar�atrika scrip-
ture, but one not known to have survived.136 This too is very probably a Kashmiri-an tradition. For I see evidence of it elsewhere only in the Haracaritacint�aman
_i, a
collection of local Kashmirian variants of Saiva myths composed in the thirteenthcentury by R�aj�anaka Jayadratha. In that text’s account of the origin of the Kash-mirian variant of the Sivar�atri festival a two-armed, red-clad form of this mysteri-ous child Vis
_n_u riding a ram, subsequently identified as a manifestation of the deity
Sam_kars
_an_a, comes with Narasim
_ha to rescue the Goddess P�arvat�ı when the
Yogin�ıs had magically extracted her from Siva’s heart and sacrificed her to him inhis Bhairava form without his knowledge:
sivas tad v�ıks_ya puratah
_parij~n�aya vimr
_sya ca
hr_dayam
_p�arvat�ıs�unyam
_cintayan ks
_obham �ayayau
45 evam_vidh�am
_tato v�art�am adhigamya jan�ardanah
_�aruhya garud_am_mes
_ar�upam
_dvibhuja �ayayau
46 sa b�alar�upah_
sauvarn_os_n_�ıs_o rakt�ambaro ’pi ca
kr�ıd_an sam�ayayau vis
_n_us tath�a sim
_hatanur narah
_47 nr_sim_h_atanun�a s�akam
_ks_obhayan yogin�ıgan
_am
jagarja ghoragambh�ıram_
n�ar�ayan_a itas tatah
_Haracaritacint�aman
_i 31.44–47
When Siva saw that before him, understood it, and reflected upon it, he
became greatly disturbed, contemplating his heart that was now empty of
P�arvat�ı. Then when Vis_n_u had learned that this had come to pass he arrived in
the form of a child, playing, two-armed, wearing a golden turban and a red
robe, riding Garu_da in the form of a ram. Narasi
_mha [came] too. Accompanied
by his Narasim_ha form N�ar�aya
_na emitted a deep and terrifying roar [rushing]
to and fro amid the band of Yogin�ıs, causing them to quake [with fear].
The Yogin�ıs, eight in number, meditate on the eight Mothers that are their sources.
These come forth and join the Yogin�ıs in placating Siva with a hymn. The terrible
Siv�ad�ut�ı then arises to devour the Yogin�ıs and a celestial voice calls on Siva toremember his own true nature. This he does and immediately the Supreme Sakti
136 Netroddyota after 13.8c–9, introducing 10: evam_
sr�ıjay�asam_hit�adr
_s_t_yoktv�a
m�ay�av�amanik�asthity�apy �aha: . . . ‘Having taught [Vis_n_u] according to the doctrine
of the Jay�asam_hit�a he says, following the rule of the M�ay�av�amanik�a: . . .’. The
form M�ay�av�amanik�a (/*M�ay�av�aman�ı ) is an abbreviation for M�ay�av�amanasam_hit�a
of the type Bh�ıma for Bh�ımasena (‘bh�ımavat’). The full form M�ay�av�amanasam_hit�a
is seen in Spandaprad�ıpik�a, p. 92. Dropping -sam_hit�a and transferring the feminine
ending to the preceding word is a common practice when citing the names of bothP�a~ncar�atrika and Saiva scriptures. Cf. Paus
_kar�a for Paus
_karasam
_hit�a and S�atvat�a
for S�atvatasam_hit�a in Spandaprad�ıpik�a, pp. 85 and 98, and K�amik�a for K�amika-
sam_hit�a in Tantr�aloka 4.25c.
285SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
emerges in her terrible, universal aspect [as K�alasam_kars
_an_�ı137]. Evidently this is
the true nature that Siva had forgotten. The Yogin�ıs restore P�arvat�ı; a secondSakti of Vis
_n_u (in addition to Siv�ad�ut�ı) comes forth to serve her; her two defenders,
the first, Vis_n_u, now identified more specifically as Sam
_kars
_an_a, sing a hymn of
praise to the terrible Goddess; and she rewards them by granting them the honourof residing on her person as ear-pendants.138
Two features beyond the mere fact of the inclusion of this myth in the Haraca-ritacint�aman
_i point to its Kashmirian character. The first is the reference to the
Goddess’s having rewarded Sam_kars
_an_a and Narasim
_ha by adopting them as her
ear-pendants, evidently an aetiological explanation of a detail of her iconography.This detail I have encountered only in manifestations of K�alasam
_kars
_an_�ı taught
in the Kashmirian part of the Jayadrathay�amala.139
137 She is not so named directly in this text. But in v. 59 we learn that at theclose of these events Siva worshipped her in the midst of the Mothers, offering upTime (k�alah
_Þ as the sacrificial victim
_: p�ujit�a m�atr
_madhye s�a kruddh�a dev�ı kap�alin�a=
upah�ar�ıkr_
tastatra pasuh_
k�alas ca duh_sahah
_. This element of the myth is surely
intended as a semantic analysis of the name K�alasam_kars
_an_�ı ‘Withdrawer of Time’.
138 Haracaritacint�aman_i 31.48–58: tatas t�ah
_ks_obhit�as t�abhy�am
_yoginyo bhayak�a-
tar�ah_=asmaran sv�am
_sa ev�antah
_saran
_am_
bh�avan�abal�at=49 tadbh�avan�abal�at *svasva(conj : svam
_svam
_Ed.) prakat
_�ıkr_ta*vigrah�ah
_(em. : vigrahamÞ=ath�avir�asan yuga-
pad brahm�an_y�ady�as ca devat�ah
_=50 par�as t�ah
_s_od_aso devyah
_pran
_amya paramesva-
ram=astuvann a~njal�ır baddhv�a vicitraih_
p�avanaih_
stavaih_=51 stut�ır vidh�ayavidhin�a
bh�uyo ’py et�a ath�avadan=*stutav�ıryo (conj.: stutiv�ıryam_
Ed.) nijam_
v�ıryam_
smaradeva nir�akulah
_=52 iti stute yogin�ıbhir mah�adeve samudyayau=d�arit�asy�a siv�ad�ut�ı yogi-
n�ıbhaks_an_odyat�a=53 athodabh�ut par�a v�an
_�ı smara rudra nij�am
_tanum=katham
_sivo-
citam_
r�upam_
vismr_tam_
te vimr_syat�am=54 tay�a gir�a mah�adevo nijam
_sasm�ara
vigraham=udyayau ca par�a saktir adbhut�ak�arar�upin_�ı=55 ghor�a sahasracaran
_�a bhak-
s_ayant�ı car�acaram=brahm�an
_d_akot
_�ır nirmathya pibant�ı bh�uri son
_itam=56 tat ks
_an_e
yogin�ıvargo n�ıtap�urv�am_
him�adrij�am=punar utp�aday�am �asa *svayogena bhay�anvit�ah_(em.: svayogen�abhay�anvit�ah
_Ed.)/57 udyayau vais
_n_av�ı saktir apar�a sevitum
_ca
t�am=sim_hasam
_kars
_an_�abhy�am
_ca par�a saktis tad�a stut�a=58 bhakty�a viracitastotr�a
dev�ı varayati sma tau=svadh�ama dehe karn_�abhy�am
_bh�us
_an_�artham adhatta ca.
139 See Jayadrathay�amala, S_at_ka 2, f. 82r7 (visualization of J�ıvak�al�ı): nr
_sim_hasam-
_kars_an_akarn
_alambin�ı; S
_at_ka 2, f. 85v8 (visualization of Ardhamun
_d_�a/Mahes�anak�a-
l�ı): v�amakarn_e pralam
_bantam
_sam
_kars
_an_am avasthitam=daks
_in_e narasim
_ham
_sy�ad;
S_at_ka 2, f. 99v5 (visualization of V�ıryak�al�ı): sphurannr
_sim_ha*sam
_kars
_’n_apr�a-
lambasrutisobhit�am (conj.: sam_karn
_apr�alam
_bhasobhin�am Cod.); S
_at_ka 3, f. 92r4
(visualization of Matacakresvar�ı): sa _nkars_an_amah�asim
_ha*sava(em. : sarva Cod.)
karn_�avalam
_bin�ım.
286 ALEXIS SANDERSON
The second is that I see no trace outside Kashmir of this novel myth
of extraction, sacrifice and restoration, while in Kashmir itself it appears ina number sources concerned with local traditions. The �Adipur�an
_a-Tithikr
_tya
gives it in the context of Um�acaturth�ı, the festival of the goddess Um�aon the fourth day of the light half of M�agha (December/January).140 Thescripture D�utid
_�amara,141 the Suresvar�ım�ah�atmya of the Sarv�avat�ara,142 the
140 �Adipur�an_a-Tithikr
_tya, ll. 2145, 2147: um�acaturthy�am
_m�aghe tu sukl�ay�am
_yogi-
n�ıgan_aih_=pr�ag bhaks
_ayitv�a *sr
_s_t�a (Cod.: sr
_s_t_v�a conj. Ed.) ca bh�uyah
_sv�a _ng�am
_-
sajair gan_aih_(I have dropped l. 2146 as a misplaced double of l. 2152) ‘On the bright
fourth sacred to Um�a in the month of M�agha the hordes of Yogin�ıs first devouredher and then re-created her with Gan
_as that were partial incarnations of their own
bodies’.141 D�utid
_�amara f. 71v11–12 (vv. 15–17 of this section): mayi nr
_tyati dev�ıti
tatra cchidram_
prakalpitam=m�atr_bhih
_tv apahr
_tya tv�am
_dev�ıcakre nivedit�a/ 16 bha-
ks_it�a yogin�ıbhis ca tato hr
_s_t_�as tu devat�ah
_=nr_tyanti ca may�a s�ardham
_y�avad eva
dinadvayam=tatas samast�a visr�ant�a hr_di tvam
_cintit�a may�a=na pasy�ami ca devi
tv�am_
vismayam_
paramam_
gatah_
‘O Goddess, by dancing there [in the cremationground] I made myself vulnerable to [their] entry. [So] the Mothers extracted you
[from your hiding place within me] and offered you up to the Cakra of the Goddess.The Yogin�ıs devoured you. Then the deities were delighted and danced with me forthe next two days. Then they all ceased and I thought of you [, believing that you
were still hidden] in my heart, and when I could not see you there I was greatlyastonished’. Bhairava in his rage smashes the Cakra of the sacrifice. The terrifiedYogin�ıs propitiate him with offerings and finally restore the dismembered Goddessto him whole (tatas samagradev�ıbhis sam
_dhit�a paramesvar�ı 30ab). He is delighted
and founds the Sivar�atri festival to commemorate these events.142 Sarv�avat�ara f. 12: *dadur (em.: dadhur Cod.) dh�up�aya preyastv�at karn
_�a-
mburuhakot_ar�at=t�am �akr
_s_ya svasakty�a vai prahars
_otphullalocan�ah
_‘With eyes wide
with joy [the Yogin�ıs ] extracted her [P�arvat�ı ] from [her hiding-place in] the inte-rior of his lotus-like ear and offered her up [to Bhairava] out of their love for him,as the sacramental fumigant [prepared from her flesh]’.
287SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Haran�agavarn_ana143 of the Vitast�am�ah�atmya, and Jayaratha’s commentary on
the Tantr�aloka144 give versions of it in the context of the Sivar�atri festival ascelebrated in Kashmir during the last five days of the dark half of Ph�algu
_na
(January/February).
143 Vitast�am�ah�atmya A f. 4r12 ff.: gav�am_
kot_isahasrasya d�anapun
_yam
_labhen
narah_=yah
_sn�ati ca vitast�ay�am
_harat�ırthasya sa _ngame=yogin�ın�am
_gan_air yatra y�age
vai *saivar�atrike (em. : sivar�atrike A : s�avar�atrike B)=v�arun_y�a saha devesi dev�ım
*�alabhya (A : �ar�adhya B) vai pur�a=p�ujayitv�a bhairav�aya *balir datto (em. : balim_datto
B : balim_dattv�a A) mah�atmane=*tad (B : tam
_A) dr
_s_t_v�a devadeve�sah
_param
_ks_obham
*av�apa sah_
(B: ag�at punah_
A) =dr_s_t_v�a ks
_obham
_*mahe�s�ani (B: param
_devi A)
bhairavasya mah�atmanah_=yogin�ın�am
_gan_ah_
s�ıghram_
pal�ayanaparo ’bhavat=s�ulamutth�apya bhagav�an yogin�ın�am
_gan_am_
tad�a=pal�ayanaparam_
dr_s_t_v�a ’nvadh�avat
svagan_air vr
_tah_=dr_s_t_v�a tath�anudh�avantam
_bhairavam
_yogin�ıgan
_ah_=m�ın�ıbh�uy�apatad
devi vitast�ay�a jale tatah_=bhagav�an api tatraiv�anvapatad balibhir vr
_tah_(The next three
lines added in the margin of B:)=�aj~n�apayat tad�a devo gan_�am_s c�urn
_ayateti ca=
m�ın�ıbh�ut�a yogin�ıs ca s�ulena gan_asattam�ah
_=evam astv iti te sarve gan
_�as tam
_yogin�ı-
gan_am=m�ın�ıbh�utam
_tad�a s�ulais c�urn
_ay�am �asur �ayudhaih
_=kuntaih
_prah�arito *hy atra
(conj. : yatra AB) yogin�ıgan_a uttamah
_=tasm�at kunt�ıprah�aro ’yam
_gr�amo para-
map�avanah_=s�ulaprotas tad�a devi yogin�ın�am
_gan_o mah�an=punar dev�ım
_samutth�apya
darsay�am �asa bhairavam=dr_s_t_v�a dev�ım
_tad�a devah
_punar utth�apit�am
_pur�a=jag�ama
paramam_
hars_am_
samutphullavilocanah_‘A man wins the merit of giving ten thou-
sand million cows who bathes at the confluence of the Harat�ırtha and the Vitast�awhere of old the bands of Yogin�ıs during the worship on the occasion of Sivar�atrisacrificed the Goddess together with wine and after worshipping great-souled Bhai-rava gave [her] to him as the Bali offering. But when the God of Gods saw that he
became extremely agitated. Seeing his agitation, O Great Goddess, the band ofYogin�ıs quickly tried to escape. When he saw this the Lord raised his trident and ranafter them surrounded by his Gan
_as. O Goddess, when the band of Yogin�ıs saw this
they turned into fish and dived into the waters of the Vitast�a. The Lord dived in afterthem accompanied by his mighty [Gan
_as]. He then ordered the Gan
_as to use their
tridents to pierce them and they did so. This most sanctifying settlement of Kunt�ı-prah�ara has its name because it was here that the supreme band of Yogin�ıs was
attacked [!-prah�ara] with pikes [!Kunt�ı-]. Then, O goddess, once the great bandof Yogin�ıs had been impaled on the tridents they restored the Goddess and showedher to Bhairava. When he saw before him the Goddess restored he became extremely
happy, his eyes wide [with joy].’144 Tantr�alokaviveka ad 28.7 (vipatprat�ık�ara
_h pramodo ’dbhutadarsana
_m yogi-
n�ımelaka_h ‘the countering of a disaster, rejoicing, seeing a marvel, mingling with
the Yogin�ıs’) concerning Sivar�atri: vipada_h svasaktyapah�ar�adir�up�ay�a
_h. pramodo h�a-
ritasya punarl�abh�adin�a: adbhutasya visvak_sobh�ade
_h: anena ca vipatprat�ık�ar�adin�a ca-
tu_s_tayena sivar�atrisa
_mj~nakam api naimittika
_m sa
_mg
_rh�ıtam ‘ ‘‘. . . of a disaster’’, e.
g. the removal of one’s Sakti. ‘‘Rejoicing’’, e.g. as a result of getting back whathad been taken away. ‘‘A marvel’’, e.g. when the whole world shakes. By [men-tioning] these four beginning with the countering of a disaster he means to includethe occasional ceremony known as�Sivar�atri’.
288 ALEXIS SANDERSON
THE ICONOGRAPHY OF BRAHM�A
The significant features of the Netratantra’s image of Brahm�a (13.33–34b) are that
it is four-faced and four-armed, with an ascetic’s staff, a rosary, an ascetic’s water-vessel, and the gesture of protection as its hand-attributes, and that it is accompa-nied by personifications of the four Vedas, two standing on either side of it.145
This combination of hand-attributes is found only in the Netratantra, other Saivasources that teach four-armed Brahm�as having only two or three of the four,146
and the distinctive presence of the personified Vedas is a detail found in no other
145 Netra 13:33–34b : lambak�urca_h sutej�as ca ha
_ms�ar�u
_dhas caturbhuja
_h=dan
_ _d�ak
_sa-
s�utrahastas ca *kaman_ _dalvabhayaprada
_h (N: kaman
_ _dalvabhaye dadhat Ed.)=vedais
caturbhir sa_myukta
_h sarvasiddhiphalaprada
_h. Ks
_emar�aja explains ad loc. that the
four Vedas are embodied and standing beside Brahm�a: vedair iti s�ak�arai_h p�arsva-
sthai_h.
146 The images as prescribed in early Saiva Pratis_t_h�atantras, scriptures concerned
only with the consecration of images and related matters, are four-armed but thehand-attributes are different. The Devy�amata (f. 69r4–v1) has the rosary and water-
vessel but the two sacrificial ladles (sruk and sruva_h) rather than the gesture and
ascetic’s staff. The Pi _ngal�amata (f. 23r4–6) and Mohac�urottara (f. 8r6–8) have onlythree of the Netra’s four hand-attributes: the rosary, the ascetic’s water-vessel, andthe staff. Instead of the fourth, the gesture of protection, the latter prescribes ‘‘ba-
rhis grass, butter, etc.’’ (barhir�ajy�adikam). The former mentions only the first three.The general scripture Kira
_na mentions the rosary and water-vessel and perhaps the
staff but not the fourth attribute (Pat_ala 52: brahmar�upa
_m prakartavya
_m catu-
rvaktra_m caturbhujam=sak�urca pi _nganetra sy�aj ja
_t�a *ytrya
_msay [for da
_m_da?] kama-
_n_dalum=s�ak
_sas�utra vratastha
_m tu ha
_msaga
_m v�abjaga
_m tu v�a).
289SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Saiva authority known to me. Now this unusual iconography corresponds exactly
with that of the surviving Kashmirian images of this god.147
LINGUISTIC EVIDENCE
Further evidence of the Netratantra’s origin in Kashmir is its use of the termkh�arkhodah
_(/kh�arkhodakah
_),148 which according to the commentator Ks
_emar�aja
denotes a supernatural device employed by an enemy for such effects as killing orexpulsion.149
147 I am aware of four such Kashmirian images: (1) a late seventh-century bronzein the Museum fur Indische Kunst, Berlin (Pal, 1975, pl. 3.): four-armed with the
Netra’s hand-attributes, attended by four small figures rightly identified by Pal as the
four Vedas, but single-faced; (2) a black stone Brahm�a in the Ganapathy�ara temple in
Srinagar dedicated to Sultan Sikandar (r. 1389–1413) (I thank Dr. John Siudmak for
sending me a photograph of this image): four-faced and four-armed with the same hand-
attributes, attended by four small figures, two on each side, their heads lower than
Brahm�a’s knees; (3) a Brahm�a in a relief of Brahm�a, Siva and Vis_n_u at N�adih�e l:
four-faced (three represented), four-armed (only the staff can be made out), with
four diminutive attendants (Siudmak 1993, p. 638, pl. 50.1 and p. 640, assigning itto the classical K�arkot
_a style of the eighth and first half of the ninth century);
and (4) an image in a private collection assigned to the late seventh or earlyeighth century (Siudmak 1993, p. 640–42, pl. 50.3; 1994, pl. 125). This is very
similar to the Berlin bronze. The outer right holds a staff (dan_ _da_h) and the inner
left shows the abhayamudr�a. The other two arms are lost. It has been broken offacross the thighs but we still have the heads and necks of two small figures on
the right. These are evidently two of the four Vedas (Siudmak 1993, p. 640).148 Netra 18.4ab: paraprayukt�a nasyanti *k
_rty�akh�arkhodak�ani caðN : k
_rty�akh�ark-
hodak�adaya_h Ed.) ‘Kr
_ty�as, Kh�arkhodakas and the like employed [against a person]
are destroyed’; 18.88b: k_rty�akh�arkhodap�ı
_dita
_h ‘tormented by a Kr
_ty�a or a
Kh�arkhoda’; 19.132bcd, 134a: kh�arkhod�as tasya v�a grah�a_h=s�akinyo vividh�a yak
_s�a_h
pis�ac�a r�ak_sas�as tath�a= . . . sarvam
_na prabhavet tatra ‘Kh�arkhodas, Planets, S�akin�ıs,
the various kinds of Yaks_a, Pis�aca and R�aks
_asa, . . . none of these can have power
over him in that [country]’.149 Netroddyota ad 19.132b: kh�arkhod�a
_h paraprayukt�a yantr�a
_h; and ad 18.4b:
m_rty�ucc�a
_tan�adik
_rd yantram
_kh�arkhoda
_h. A yantram/yantra
_h is a Mantra-inscribed
diagram written in various colours and with various inks on cloth, birchbark, thehides of various animals and the like, wrapped up and then employed in variousways (by being worn as an amulet, by being buried in a cremation ground, and so
on) for purposes such as warding off ills, harming an enemy, or forcing a personto submit to the user’s will. Cf. Ks
_emar�aja’s definition of a yantracakram as a ser-
ies of Mantras written in a particular spatial arrangement (ad 20.59c): yantra-
cakram_visi
_s_tasam
_nivesalikhito mantrasam�uha
_h.
290 ALEXIS SANDERSON
The word, which is of Iranian origin,150 appears in Sanskrit sources in a
number of variants; and these form two categories according to whether the rprecedes the second consonant, as in the Netratantra, or the third. The latterposition is the original, since it is that which we see in the Iranian source as evi-
denced by Avestan kaxvar��a- (m.), kaxvar
e
i��ı- (f.) denoting a kind of malevo-lent spirit, probably associated with sorcery.151 This is the source of the formskhakkhorda-, khahkhorda-, kh�akkhorda- and khakhorda- seen in early northwest-
ern and Central-Asian Sanskrit sources152 and in the G�andh�ar�ı (Kroraina Pra-krit) of the Kharos
_t_h�ı documents of the third century AD from Niya in
Xinjiang,153 and of the k�akhorda- that appears in Mah�ay�ana-Buddhist works.154
Over against these we have the form kh�arkhoda- seen in the Netratantra and avariant kh�arkhot
_a, in which the r has migrated from the final to the second con-
sonant. It is only in this form that the word occurs in non-Buddhist sources;and I have found it outside the Netratantra only in works that were composed
or redacted by Kashmirians. We see it in the R�ajatara _ngin_�ı of Kalhan
_a,155 in the
Kashmirian part of the Jayadrathay�amala156 and the related Trida�sad_�amara,157
150 Burrow (1935, p. 780).151 Bartholomae (1961), s.v., pointing to the fact that the Armenian loan-word
kaxard means ‘sorcerer, wizard’. These beings, male and female, are mentioned inYasna 61 of the Avesta among the creatures of the ‘‘hostile spirit’’ Angra Mainyu
(Pahl. Ahriman).152 See Hoernle (1892, pp. 356, 368–69); Hoernle (1893, p. 25).153 Burrow (1935, pp. 780–81) concerning the punishment khakhordastriyana ‘of
witches’ and khakhordi stri ‘a witch’, reading rda for rna in the light of the Iraniansource word.
154 See, e.g., Amoghap�asakalpar�aja f. 3v: k�akhordacchedan�ı sastren_a; f. 48v:
k�akhordacchedana . . . k�akhord�a vinasyanti; Suvarn_abh�asottamas�utra p.3, l.2: k�akho-
rdad�arun_agrahe; p. 107, l. 8: sarvak�akhordavet�a�d�a_h; Bhai_sajyagurus�utra, pp. 13–14:k�akhordavet�al�anuprayogena j�ıvit�antar�ayam
_sar�ıravin�asam
_v�a kartuk�am�a
_h; Mah�am�a-
y�ur�ı p. 57: k_rty�akarman
_ak�akhordakiran
_a-.
155 R�ajatara _ngin_�ı 4.94: khy�ata
_h kh�arkhodavidyay�a=ni
_hsam
_bhrama
_h stambhayitum
_deva divyakriy�am alam; 5.239: kh�arkhodavedinam=r�amadev�ahvayam_
bandhum abhic�a-ram ak�arayat.
156 Jayadrathay�amala, S:a_tka 3, f. 70v6 (9.41c–42b): evam
_vidh�am
_yantran�ase k
_rty�a-
kh�arkhodamardane=cintayet parames�an�ım abhic�arupramardane; f. 72r5 (10.2ab): para-mantragr�asakaram
_k_rty�akh�arkhodagha
_t_tanam; S:a
_tka 4, f. 3v5 (2.49ab): k
_rty�a*
kh�arkhodadalan�ı (em. : kh�akhodalan�ı Cod.); f. 7v6 (2.74ab): k_rty�akh�arkhodav-
ighnaugham_
bandhan�ad dhvam_sayi
_syati; f. 14v7 (2.235a): k
_rty�akh�arkhodadaman�ı;
f. 16v6 (2.297ab): bh�utavet�aladaman�ı k_rty�akh�arkhodamardan�ı.
157 Tridas�ad�amara-Pratya _ngir�akalpa f. 11v5: mantrav�adas tu *kh�arkhodam_(conj.:
kh�arkhoda Cod.) *vi_sam_
(corr.: vi_sa Cod.) sth�avara*ja _ngamam (corr.: ja _ngama
Cod.) =garajvar�adayo devi anye ne_s�am anekasa
_h=paraprayukt�a nasyanti (cf. Netra
18.4ab: paraprayukt�a nasyanti k_rty�akh�arkhodak�adaya
_h). Since the subject-matter
and language of this text is closely allied to that of the Kashmirian part of theJayadrathay�amala, it is not improbable that it too was Kashmirian in origin orredacted from Kashmirian materials.
291SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
in the Br_hatk�alottara,158 in the exorcistic Ga
_nes�am�al�amantra of the Kashmirian
manual of �Saiva initiation,159 in a Kashmirian Ga_nesastotra attributed to the
�Adipur�a_na.160 and, as kh�arkhot
_a-, in the Haracaritacint�ama
_ni of R�aj�anaka Jaya-
dratha.161 Moreover it is only in Kashmiri that the word has survived into the
New Indo-Aryan languages.162
158 B_rhatk�alottara B f. 118v4–5 (Pavitr�arohan
_apa
_tala): sarvavighn�ani nasyanti
grah�a vai vy�adhayas tath�a=vin�ayakopagh�at�as ca *k_rty�akh�arkhodak�adaya
_h (em. :
k_rtyakh�akhodak�adaya
_h Cod.)
159 Kal�ad�ık_s�avidhi f. 3r6–9: OM� HUM� H�UM� NAMAH� KS�ETR�ADHIPATAYE SARV�ARTHASID-
DHID�AYA SARVADUH� KHAPRAsAMAN�AYA EHY EHI BHAGAVAN SARVAKH�ARKHOD�AN STA-MBHAYA 2 HRIM� H�UM� G�AM� NAMAH� SV�AH�A iti ga
_nesam�al�amantra
_h.
160 Ga_nesastotra v. 51, 53a: etat stotra
_m pavitra
_m tu ma _ngala
_m p�apan�asa-
nam=sastra*kh�arkhoda(em. : kharkhoda Ed.)vet�alayak_sarak
_sobhay�apaham=. . . tri-
sam_dhyam
_ya_h pa
_thet ‘He who at the three junctures of the day recites this hymn,
purifying, auspicious, that destroys [all] sins, that removes the danger of weapons,
Kh�arkhodas, Vet�alas, Yak_sas and Rak
_sases . . .’. That this hymn is Kashmirian is
made probable by its being assigned to the �Adipur�a_na, since that is one of the most
common loci of attribution for Kashmirian compositions seeking scriptural status.
It is made certain by two facts: (1) it refers repeatedly to Bh�ıma[sv�amin], the prin-cipal Ga
_nesa of Kashmir (v. 10: bh�ımam
_. . . kasm�ırav�asam; v. 17: sat�ısaraniv�asinam;
v. 36: kasm�ıre bh�ımar�upin_am) – Bh�ımasv�amin’s temple is in Srinagar near the foot
of H�araparvat (S�arik�aparvata, Pradyumnagiri) (see Stein 1961, vol. 2, p. 446) –
and (2) it mentions that Ga_nesa is seated upon two [couchant] lions (v. 4:
hariyugalanivi_s_tam_; v. 21: sim
_hayug�asana
_h), which is a distinctive feature of Ka-
shmirian Ga_nesa images (see Siudmak 1994, plates 73, 157 and 158; Reedy 1997,
K68, K86, K87, K89; Pal 2003, pl. 57 [‘‘Chamba, 10th century’’]).161 Haracaritacint�ama
_ni 2.125: k
_rty�akh�arkho
_tavet�al�a ye. There are also vari-
ants in Mah�ay�ana-Buddhist sources in which k�a takes the place of the initial kh�a:k�arkhoda- in a manuscript of the �Aryat�ar�an�am�a
_s_tottarasataka, v. 49:
_d�akinyo-
st�arak�ah_
pret�ah_
*skandonm�ad�a (conj.: skandom�ad�a Cod.) mah�agrah�a_h= ch�ay�apa-
sm�arak�as caiva *yak_sa(conj. : tak
_sa Cod.)k�arkhodak�adayah
_; and k�arkho
_ta- in the
edition of the Ma~njusr�ım�ulakalpa, p. 539, l. 8: sarvak�arkho_t�as chinn�a bhavanti.
162 See Kashmiri khokhu, khakha-bo_tu, kh�okha-bo
_tu and kh�okha-mo
_tu ‘bogey, bug-
bear, hobgoblin, ogre’ in Grierson 1915, p. 395b. The Kashmirian scholar who pre-pared the slips for these words used by Grierson, either Pa
_n_dit Govinda Kaula or
Pa_n_dit Mukunda R�ama S�astr�ı, gave Sanskrit kh�arkhodah
_as the meaning of these
terms. Turner (1966, p. 201, s .v. kharkh�oda) records no derivatives in any otherNIA language.
292 ALEXIS SANDERSON
I conclude from this iconographical and linguistic evidence that it is highly
improbable that the Netratantra was composed anywhere other than Kashmir.
THE DATE
As for the date of its composition, I have proposed above that the approximateouter limits are AD 700 and 850. The posterior limit is established by the fact that
the text teaches the Vaikun_t_ha form of Vis
_n_u without the face of Kapila at the rear.
It is therefore unlikely to have been composed later than the middle of the ninth cen-tury, since it was from that time that the four-faced form replaced the three-faced in
Kashmirian images.163
The prior limit cannot be placed before the end of the formative period of
Kashmirian iconography, that is to say, the late seventh century. For it is onlyafter that time that the Brahm�a icon taught in the Netratantra is seen in the stoneand bronze images of the region.
However, the iconography of the four-armed form of Siva taught in 13.29–30makes it probable that the work was composed towards the end of this period, c.800–850. The hands’ attributes are the trident paired with the gesture of protec-
tion, and the citron paired with the rosary.164 The pairing of the citron and therosary is seen outside Kashmir from an early date.165 But in Kashmir we see the
163 See nn. 133–134 above.164 Netra 13.30: caturbhujam
_mah�atm�anam
_s�ul�abhayasamanvitam=m�atulu _ngadha-
ram_devam ak
_sas�utradharam
_prabhum.
165 We see it in a fifth-century Li_nga with a bust of a three-headed Siva, which, Si-
udmak proposes, (1994, pl. 43) is from the cave known in Pashtu as Kashmir Smast(‘the cave to Kashmir’), located 25 km north of Shahbazgarhi on a mountain topbetween the Peshawar valley and Buner: the left hand holds the citron and the right
the rosary. The same two were probably in the front hands of the famous three-faced ‘‘Mahesam�urti’’ of the late sixth century in the Siva cave at Elephanta. The cit-ron is clear in the left hand and though the rosary is not visible in the damaged righthand of the posture, which is raised with out-turned palm, suggests its presence.
They are also seen in the two front hands of the three-faced bust of Mahesvara inthe inner sanctum (garbhag
_rham) of the Siva temple constructed in AD 637 at Ku-
suma in the Sirohi district of SW Rajasthan (Meister, 1988, pp. 208–214; pl. 437);
and a citron is held in the lower left hand of a six-armed, three-faced Mahesvaracarved in the centre of the wooden door frame of the Uttaresvara temple at Ter inthe Osmanabad District of Maharashtra (early 7th century?). The lower and middle
right hands are lost, but a rosary may well have been in the former. The other threesurviving hands hold a cobra (upper right), a lotus (middle left), and a mace sur-mounted by a Li_nga (upper right) (cf. Collins, 1988, p. 117).
293SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
ascetic’s water-vessel rather than the fruit in the early centuries.166 The citronappears in our Kashmirian images only from the ninth century.167
ABBREVIATIONS
166 This we see in Gandharan Siva images of the fourth and fifth centuries (Si-
udmak, 1994, plates 41–43 and 59) and in a number of Kashmirian images whosehand-attributes have survived: (1) the Fattegarh three-headed Mahesvara of thefifth to sixth centuries (Siudmak, 1994, pl. 39a,b ); and (2–3) two three-headedMahesvaras from P�andr~et
_han of the latter half of the seventh century (Siudmak,
1994, plates 85–86). This iconography continued after the introduction of theother, since we see it in a grey chlorite four-armed Siva and P�arvat�ı group now inthe Metropolitan Museum of Art (no. 1989.362) that is probably of the first half
of the ninth century (Siudmak, 1994, pl. 117).167 We see this in (1) the Siva with consort in the Kashmirian ‘brahmanical
triad’ (Pal, 1975, pl. 2: ‘9–10th century’): Siva is seated on Vr_s_a (his bull) with
P�arvat�ı on his left thigh and holds the rosary and citron in his inner right and lefthands. In his outer right he holds a trident, and in outer left a snake; (2) the two-armed single-faced Siva of a Kashmirian ekamukhali _ngam (Pal, 1975, pl. 5: ‘8–9th cen-
tury’) (= Reedy, K55); (3) a Kashmirian grey chlorite group of c. 850 (Siudmak,1994, pl. 116; Pal, 2003, pl. 67 [but dated 750–800]); and (4) the ‘Siva-P�arvat�ı’ image-set in the Gaur�ısa_nkara temple in Chamba (Pal, 1975, pl. 85: ‘10th century’).
ARE Annual Report on ½South Indian� Epigraphy. Archaeological Surveyof India, 1887
BL Bodleian Library, OxfordBORI Bhandarkar Oriental Research Institute, PuneCod. The reading of the manuscript
conj. My conjectural emendationcorr. My correctionEd. The reading of the published editionem. My emendation
Ep. The reported reading of an inscriptionK Khmer inscription, numbered as in Cœdes, 1966KSTS Kashmir Series of Texts and Studies
N The reading of the Nepalese Amr_tesatantra manuscript
NAK National Archives of Nepal, KathmanduNGMPP Nepal-German Manuscript Preservation Project
SII South Indian Inscriptions. Archaeological Survey of India, 1980–
294 ALEXIS SANDERSON
REFERENCES
Primary Sources: Manuscripts and Editions
Agnik�aryapaddhat. Paris, Biblioth�eque Nationale, MS Sanscrit 505 C; paper; S�arad�ascript (A); BL, MS Chandra Shum Shere f. 110; paper; S�arad�a script (B).
Atharvavedaparisis_t_a: The Parisis
_t_as of the Atharvaveda, ed. by Bolling, George Mel-
ville, and Julius von Negelein. Otto Harrassowitz, Leipzig, 1909.Amr
_tabhairav�arcanavidhi. In P�uj�ak�an
_d_a.
Amr_tas�uryap�uj�avidhi. NAK MS 1696/1713, NGMPP Reel No. A 435/29. Paper;
Newari script.Amr
_tas�ury�arcanavidhi. In P�uj�ak�an
_d_a.
Amr_t�ısabhairavabhat
_t_�arak�ahnikavidhi. In P�uj�ak�an
_d_a.
Amr_tesatantra. NAK MS 1-285, NGMPP Reel No. B 25/5. Palm-leaf; Nepalese
variant of proto-Bengali script; AD 1200. See Netra.Amr
_tesvarad�ıks
_�avidhi of Visvesvara. NAK MS 5–4867, NGMPP Reel No. A 231/
17. Paper; Newari script.
Amr_tesvarap�uj�a ofKingAbhayamalla (r. AD 1216–55).NAKMS5-4863,NGMPPReel
No. A 231/13. Paper;Devan�agar�ı script. Apograph ofNAKMS 1–1365.5 of AD 1216.Amr
_tesvarap�uj�agnik�aryavidh�ana. NAK MS, 3–380, NGMPP Reel No. A 49/11.
Paper; Newari script.Amoghap�asakalpar�aja. Transcr. by the Mikky�o Seiten Kenky�ukai. TranscribedSanskrit Text of the Amoghap�asakalpar�aja, Part I. Taish�o Daigaku S�og�o-Bukky�o-Kenky�ujo Nenpo 20, 1996, pp. 304–251 (ff. 1v–16v).
Arthas�astra, ed. by Kangle, R.P. Motilal Banarsidass, Delhi, 1969 (second edition).�Adipur�an
_a-Tithikr
_tya, ed. by Ikari, Yakuke & Takao Hayashi, A Study of the
N�ılamata. Aspects of Hinduism in Ancient Kashmir, pp. 83–136. by YasukeIkari, Kyoto, Institute for Research in Humanities, Kyoto University, 1994.
�Anande�svarap�uj�a. BL, Chandra Shum Shere MS e. 264 (‘Tantric Collectanea’), ff.58[1]v1–60[3]v6. Paper; S�arad�a script.
�Aryat�ar�an�am�as_t_ottarasataka. Manuscript in private hands, Kathmandu, NGMPP
Reel No. D 35/28 (‘Cakrasam_varap�uj�avidhih
_’), ff. 24v–31v2. Paper; Newari script.
Kath�asarits�agara of Somadeva, ed. byL�al S�astr�ı, Pan_d_it Jagad�ıs, Motilal Banarsidass,
Delhi, 1970.Karmak�an
_d_a: Karmak�an
_d_am;Caturtham
_Pustakam, ed. by Jyotirvid, Pan
_d_ita Kesava-
bhat_t_a (Pand
_ith K�eshe�v Bat
_a J�utish). Nirn
_aya S�agara Press, Bombay, 1936. Repro-
duced in Chandra,1984,pp.127–247.Kal�ad�ıks
_�avidhi of Manodaguru. BORI, Ms 157 of 1886–92. Paper; S�arad�a script.
K�at_hakagr
_hyas�utra with extracts from the commentaries of Devap�ala, �Adityadarsana,
and Br�ahman_abala, ed. by Caland, Wilhelm. Day�ananda Mah�avidy�alaya Sam
_skr_ta
Grantham�al�a 9. ResearchDepartment, D.A.V. College, Lahore, Vikrama, 1981.K�asm�ırikakarmak�an
_d_apaddhati. BL, MS Stein Or. c. 9. Birch-bark; S�arad�a script.
Kubjik�amata: The Kubjik�amatatantra: Kul�alik�amn�aya Version, ed. by Goudriaan,
T. & J.A. Schoterman. Brill, Leiden, 1988.Jayadrathay�amala, S
_at_ka 2. NAK MS 5-4650, NGMPP Reel No. A 153/2. Paper;
Devan�agar�ı script.Jayadrathay�amala, S
_at_ka 3. NAK MS 5-1975, NGMPP Reel No. A 152/9. Paper;
Newari script.
295SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Jayadrathay�amala, S_at_ka 4. NAK MS 1-1468, NGMPP Reel No. B 122/4. Paper;
Newari script.Jay�akhyasam
_hit�a, ed. by Krishnamacharyya, Embar. Gaekwad’s Oriental Series, 54.
Oriental Institute, Baroda, 1967.
Jay�akhyasam_hit�a, NAK MS 1-49, NGMPP Reel No. B 29/3 (‘Jay�aks
_arasam
_hit�a’).
Palm-leaf; Newari script; AD 1462.Tantr�aloka of Abhinavagupta with the commentary (-viveka) of R�aj�anaka Jayara-
tha, ed. by Mukund R�am S�astr�ı. KSTS 23, 28, 30, 35, 29, 41, 47, 59, 52, 57, 58,Bombay and Srinagar, 1918–38.
Tridasad_�amara-Pratya _ngir�akalpa. NAK MS 3-30, NGMPP Reel No. B 173/22
(‘Tridasad_�amar�apratya _ngir�avis
_ayakan�an�atantra’). Paper; Newari script; AD 1617/8.
D�utid_�amara. Section of this text on Sivar�atri quoted in the Nity�adisa _ngraha-
paddhati, f. 71v4–72v15.Devy�amata. NAK MS 1-279, NGMPP Reel No. A 41/15 (‘Nisv�as�akhyamah�a-tantra’). Palm-leaf; Nepalese Licchavi script.
Desopadesa: The De�sopadesa and Narmam�al�a of Ks_emendra, ed. by Kaul Sh�astr�ı,
Pan_d_it Madhus�udan. KSTS 40, Poona, 1923.
Nandiks_etram�ah�atmya of the Sarv�avat�ara. BL, MS Stein Or. e. 2 (v). Paper; S�arad�a
script.Naresvarapar�ıks
_�a of Sadyojyotis with the commentary (-prak�asa) of Bhat
_t_a
R�amakan_t_ha, ed. by Kaul S�astri, Madhus�udan. KSTS 45. Srinagar, 1926.
Narmam�al�a. See Desopadesa.Navar�atrap�uj�a. NAK MS 1–220, NGMPP Reel No. A 240/17. Paper; Newari
script; Newari and Sanskrit.Nity�adisam
_grahapaddhati of R�aj�anaka Taks
_akavarta. BORI MS. No. 76 of 1875–
76 (‘Bhr__ngesasam
_hit�a’). Paper; S�arad�a.
N�ılamata: N�ılamatapur�an_a, ed. by de Vreese, K. Brill, Leiden, 1936.
Netra and Netroddyota: Netratantra with the commentary (-uddyota) by Ks_emar�aja,
ed. by Kaul S�astr�ı, Madhus�udan. KSTS 46, 59. Bombay, 1926, 1939.168
Naimittikakarm�anusam_dh�ana of Brahmasambhu. Asiatic Society of Bengal, Cal-
cutta, MS G 4767. Palm-leaf; early Newari script; incomplete.Ny�ayama~njar�ı of Jayantabhat
_t_a, ed. by Varadacharya, K.S. 2 vols. University of
Mysore, Oriental Research Institute Series 116 and 139. Mysore, 1969 and 1983.
Pi _ngal�amata. NAK MS 3–376, NGMPP Reel No. A 42/2. Palm-leaf; Newariscript; AD 1173/4.
Picumata (=Brahmay�amala). NAK MS 3–370, NGMPP Reel No. A 42/6. Palm-
leaf; early Newari script; 12 January A:D: 1052 (Petech 1984, p. 44).P�uj�ak�an
_d_a. Cambridge, University Library, MS Add. 1412. Paper; Newari script.
Paippal�adavas�adis_at_karmapaddhati, compiled by Pan
_d_ita Um�ak�anta Pan
_d_�a. Bala-
sore: n.d.
Br_hatk�alottara. NAK MS 1–273, NGMPP Reel No. B 24/57 (‘K�alottaratantram’);palm-leaf; Nepalese variant of proto-Bengali script (A); NAK MS 1-89, NGMPPReel No. B24/59 (‘K�alottaratantram’); palm-leaf; Newari script (B).
168 Citations of Netra above give only the chapter and verse numbers of thisedition. The citations themselves have been edited by collating this edition with
the Nepalese MS (see here Am_rte�satantra).
296 ALEXIS SANDERSON
Br_hatsam
_hit�a of Var�ahamihira with the commentary (-vivr
_ti) of Bhat
_t_otpala, ed.
by Trip�at_h�ı, A.V. Sarasvat�ıbhavanagrantham�al�a 97, Varanasi, 1988.
Bhais_ajyagurus�utra, ed. by Dutt, Nalinaksha. Gilgit Manuscripts Vol: 1, pp. 1–32
Srinagar, 1939.
Ma~njusr�ım�ulakalpa: �Aryama~njusr�ım�ulakalpa, ed. by Vaidya, P.L. Buddhist SanskritTexts 18. Mithila Institute, Darbhanga, 1964. Reprinted in one volume from theedition of T. Ganapati S�astr�ı, Trivandrum Sanskrit Series 70, 76 and 84, Trivan-
drum, The Oriental Manuscript Library of the University of Travancore, 1920,1922 and 1925.
Mata _ngap�aramesvara, Kriy�ap�ada, Yogap�ada and Cary�ap�ada, with the commentary
(Mata _ngavr_tti) of Bhat
_t_a R�amakan
_t_ha up to Kriy�ap�ada 11.12b, ed. by Bhatt, N.
R. Publications de l’Institut francais d’Indologie No. 65. Institut francais d’Indol-ogie, Pondicherry, 1982.
Mayasam_graha. NAK MS 1-1537, NGMPP Reel No. A 31/18. Palm-leaf; Newari
script; incomplete.Mah�am�ay�ur�ı: �Arya-Mah�a-M�ay�ur�ı Vidy�a-R�aj~n�ı, ed. by Sh�uyo Takubo, Tokyo San-kibo, 1978.
M�alin�ıvijayav�artika. See Hanneder 1998.Moks
_op�aya;Vair�agyaprakaran
_a: Bh�askarakan
_t_ha’s Moks
_op�aya-t
_�ık�a, A Commentary
on the Earliest Available Recension of the Yogav�asis_t_ha: 1: Vair�agyaprakaran
_am,
revised edition by Jurgen Hanneder and Walter Slaje, Geisteskultur Indiens. Texteund StudienNo. 1, Aachen: Shaker Verlag, 2002.
Mohac�urottara. NAK MS 5-1977, NGMPP Reel No. A 182/2. Paper; Devan�agar�ı;copied from a palm-leaf manuscript in the NAK dated [Valabh�ısam
_vat] 806 [AD
1123/4].169
Y�aj~navalkyasmr_ti with the commentary (Mit�aks
_ar�a) of Vij~n�anesvara, ed. by Wasu-
dev Laxman_Sastrı Pan
_sıkar. Bombay, Pandurang Javajı, 1926.
R�ajatara _ngin_�ı of Kalhan
_a, ed. by Stein, M.A. Munshi Ram Manohar Lal, Delhi,
1960. Reprinted from the edition of 1892.Laks
_m�ıtantra, ed. by Krishnamacharyya, Pandit V. Adyar Library Series 87. A-
dyar Library and Research Centre, Adyar, Madras, 1959.Li _ngapur�an
_a with the commentary (Sivatos
_in_�ı) of Gan
_esas N�atu. ed. by Ga_ng�avis
_n_u.
Nag Publishers, Delhi, 1989 and 1996. Reprinted from the edition of the Venkate-
svara Steam Press, Bombay, 1924.Laug�aks
_igr_hyas�utra with the commentary -mantrabh�as
_yam of Devap�ala. ed. by Kaul
Sh�astr�ı, Pan_d_it Madhus�udan. 2 vols.KSTS 49 and 55, Bombay, 1928 and 1934.
169 The scribe of the apograph supposes that the apparent age of the palm-leafexemplar entails that the date 806 is not in the Nepali era but the Vikrama (f. 47v:
asy�adh�arabh�utasya t�a_dapatrapustakasya pr�ac�ınatay�a tatpratilipik
_rtasya pustak�a-
ntarasy�api trisat�abdap�urvapr�ac�ınat�adarsanena tadullikhita 806 samvatsaro nep�alasamvatsar�ad bhinno vaikram�adih
_sambh�avyate). But the resulting date, AD 749, is im-
plausibly early for this text. Dr. Divakar Acharya of the Mahendra Sanskrit Univer-sity, Kathmandu, who has recently located the manuscript, has reported that thescript is west-Indian in appearance and that the date is therefore probably to be cal-culated in the Valabhi era. This report was passed on to me by Dr. Dominic Goodall
in an e-mail message of April 23, 2004. Year 806 after the Valabh�ı era is AD 1123/24,an entirely plausible date.
297SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Vijayesvaram�ah�atmya attributed to the �Adipur�an_a. BL, MS Stein Or. d. 48 (viii).
Vitast�am�ah�atmya. BL, MS Stein Or. d. 55 [ii]; paper; S�arad�a script (A); BL, MSStein Or. d. 46; paper; S�arad�a script (B).
Vis_n_udharmottara: Vis
_n_udharmottarapur�an
_a, ed. by Kr
_s_n_ad�asa, Ks
_emar�aja. Nag
Publishers, Delhi, 1985. Reprinted from the edition of the Venkatesvara SteamPress, Bombay, 1912.
Vedakalpadruma, Pan_d_ita Kesavabhat
_t_a Jyotirvid (Pand
_ith K�eshe�v Bat
_a J�utish).
Nirn_aya S�agara Press, Bombay, 1921. Reproduced in Chandra, 1984, pp. 26–126.
Sarv�avat�ara. BL, MS Stein Or. d. 48(i). Paper; Kashmirian Devan�agar�ı script.Sivanirv�an
_avidhi. Karmak�an
_d_a, pp. 205–292. Reproduced in Chandra, 1984, pp.
185a–206d.Sivas�utravimarsin�ı of Ks
_emar�aja, ed. by Chatterji, Jagadisha Chandra. KSTS 1,
Srinagar, 1911.Saivavaisvadevavidhi. Karmak�an
_d_a, pp. 452–456. Reproduced in Chandra, 1984,
pp. 246d–247d.Siddh�antas�arapaddhati of Mah�ar�aj�adhir�aja Bhojadeva. NAK MS 1–1363, NGMPPReel No. B 28/29. Palm-leaf; early Newari script; copied in AD 1077/8.
Spandak�arik�a of Vasugupta with the commentary (-vr_tti) by Kallat
_a, ed. by Chatterji,
Jagadisha Chandra.KSTS 5, Srinagar, Sam_vat, 1970 [AD 1913/4].
Spandaprad�ıpik�a of Bh�agavatotpala, ed. by Kaviraja, Gopinatha. Tantrasa-_ngraha ½Part I � (Yogatantragrantham�al�a 3. Benares, Varanaseya Sanskrit Vishv-avidyalaya, 1970), pp. 83–128. Reprinted from the edition of V�amanas�astr�ıIsl�amapurakara (Vizianagaram Sanskrit Series 5. K�as�ı, Saka 1820 [AD 1898]).
Suvarn_abh�asottamas�utra, ed. by Nobel, Johannes (Suvarn
_aprabh�asottamas�utra : Das
Goldglanz-S�utra; ein Sanskrittext des Mah�ay�ana-Buddhismus nach dem Handschrif-ten und mit Hilfe der tibetischen und chinesischen €Ubertragungen). Harrassowitz,Leipzig, 1937.
Svacchanda: Svacchandatantra with the commentary (-uddyota) of Ks_emar�aja, ed.
by Kaul S�astr�ı, Madhus�udan. KSTS 31, 38, 44, 48, 51, 53, 56. Bombay, 1921–35.Haracaritacint�aman
_i of R�aj�anaka Jayadratha, ed. by Sivadatta & K�as�ın�atha
P�an_d_ura_nga Parab. K�avyam�al�a 61. The Nirn
_aya-s�agara Press, Bombay, 1897.
Hars_acarita of B�an
_abhat
_t_a with the commentary (-sam
_keta) of Sa_nkara, ed. by
Fuhrer, A. A. Bombay Sanskrit and Prakrit Series 66. Department Bombay of
Public Instruction, 1909.
Secondary Sources
Bartholomae (Christian), 1961: Altiranisches W€orterbuch. Berlin, Walter der Gruy-
ter and Co. Reprint of the first edition (1904).Brunner (Helene), 1974: Un Tantra du nord: le Netra Tantra. Bulletin de l ’ �Ecolefran�caise d ’Extreme-Orient 61, pp. 125–197.
Burrow (T), 1935: Iranian Words in Kharos_t_hi Documents–II. Bulletin of the
School of Oriental Studies 7, pp. 779–90.Chandra (Lokesh), 1984: Sanskrit Texts from Kashmir;Volume 7. Sata-pit
_aka Se-
ries, Indo-Asian Literatures, No. 333. New Delhi, Sharada Rani. A photo-
graphic reproduction of printed ritual texts.Chimpa (Lama), & Chattopadhyaya (Alaka), 1970: T�aran�atha’s History of Buddhismin India translated from the Tibetan. Simla, Institute of Advanced Studies.
Collins (Charles Dillard), 1988: The Iconography and Ritual of Siva at Elephanta.
Albany, State University of New York Press.
298 ALEXIS SANDERSON
Cœdes (George), 1937–66: Inscriptions du Cambodge. 8 vols. 1937 (vol. 1), 1942
(vol. 2), 1951 (vol. 3), 1952 (vol. 4), 1953 (vol. 5), 1954 (vol. 6), 1964 (vol. 7),
1966 (vol. 8). Paris, Ecole franc$ aise d’Extreme-Orient.
Cœdes (George), 1968: The Indianized States of Southeast Asia, ed. by Walter
F. Vella. Translated by Susan Brown Cowling. Honolulu, The University Press of
Hawaii. Originally published as Les Etats hindouises d ’Indochine et de l ’Indonesie(Paris, de Boccard, 1948).
Grierson (Sir George), assisted by Mah�amahop�adhy�aya Mukunda R�ama S�astr�ı.1915: A Dictionary of the K�ashm�ır�ı Language compiled partly from materials leftby the late Pan
_d_it �Isvara Kaula. Bibliotheca Indica, New Series, No. 1405. Hert-
ford, Asiatic Society of Bengal.Guesdon (Joseph), 1930: Dictionaire cambodgien-fran�cais. Paris, Librairie Plon.Hanneder (Jurgen), 1998a: Saiva Tantric Material in the Yogav�asis
_t_ha. Wiener Zeit-
schrift fur die Kunde Sudasiens und Archiv fur Indische Philosophie 42, pp. 67–76.Hanneder (Jurgen), 1998b: Abhinavagupta’s Philosophy of Revelation. M�alin�ısloka-v�arttika I; 1–399. Groningen Oriental Series 14. Groningen, Egbert Forsten.
Hanneder (Jurgen), 2003: Studies on the Moks_op�aya. Habilitation thesis, Faculty
of Oriental Studies, University of Hamburg.Hobart (Angela), Ramseyer (Urs) & Leeman (Albert), 1966: The People of Bali,Oxford, Blackwell.
Hoernle (A.R.), 1892: The Third Instalment of the Bower Manuscript. The IndianAntiquary 21, pp. 1–40.
Hoernle (A.R.), 1893: The Weber Manuscripts. Journal of the Asiatic Society of
Bengal 62, pp. 1–40.Huntington (Susan L.), 1985: The Art of Ancient India; Buddhist; Hindu; Jain. NewYork/Tokyo, Weatherhill.
Meister (Michael), 1988: C�ap�otkat_as of Bhillam�ala. Chapter 16 (pp. 207–214) of
Encyclopaedia of Indian Temple Architecture: North India: Foundations of NorthIndian Style; c: 250 B:C:� A:D: 1100, ed. by Michael Meister, M.A. Dhaky andKrishna Deva. Delhi, Oxford University Press.
Pal (Pratapaditya), 1975: Bronzes of Kashmir. Graz, Akademische Druck- u. Ver-lagsanstalt.
Pal (Pratapaditya), 2003. Himalayas : An Aesthetic Adventure. Chicago, The Art
Institute of Chicago.Petech (Luciano), 1958: Medieval History of Nepal ðc:750� 1480Þ. Serie OrientaleRoma 10. Rome, Istituto per il Medio e Estremo Oriente.
Phoen (Mak), 1987: The Cham Community in Cambodia from the Fifteenthto the Nineteenth Century, Proceedings of the Seminar on Champa, pp. 76–86.www.seacrc.org/media/pdfiles/ChamBook.pdf.
Reedy (Chandra L.), 1997: Himalayan Bronzes: Technology; Style; and Choices.
Newark, University of Delaware Press.Robson (Stuart (tr.)), 1995: Desawarn
_ana ðN�agarakr
_t�agamaÞ by Mpu Prapa~nca.
Verhandelingen van het Koninklijk Instituut voor Taal-, Land- en Volkenkunde
169. KITLV Press, Leiden.Sanderson (Alexis), 1990: ‘‘The Visualization of the Deities of the Trika,’’ L’ImageDivine: Culte et Meditation dans l ’Hindouisme, ed. by Andre Padoux. Paris, Edi-
tions du Centre National de la Recherche Scientifique, pp. 31–88.
Sanderson (Alexis), 1995: ‘‘Meaning in Tantric Ritual,’’ Essais sur le Rituel III :
299SAIVA OFFICIANTS – THE KING’S BRAHMANICAL CHAPLAIN
Colloque du Centenaire de la Section des Sciences religieuses de l ’Ecole Pratiquedes Hautes Etudes, ed. by A.-M. Blondeau & K. Schipper. Louvain and Paris,Peeters, pp. 15–95.
Sanderson (Alexis), 2001: ‘‘History Through Textual Criticism in the Study of Sai-
vism, the Pancar�atra and the Buddhist Yogin�ıtantras,’’ Les sources et le temps:Sources and Time: A Colloquium; Pondicherry; 11–13 January 1997, ed. by Fran-cois Grimal. Publications du departement d’Indologie 91. Pondicherry, Institut
francais de Pondichery/1cole francaise d’Extreme-Orient, pp. 1–47.
Sanderson (Alexis): ‘‘Remarks on the Text of the Kubjik�amatatantra.’’Indo-Iranian Journal 45, pp. 1–24.
Sanderson (Alexis), 2005: The Saiva Religion Among the Khmers, Part 1. Bulletin
de l ’Ecole francaise d ’Extreme-Orient 90–91 (2003–2004): 349–462.Sanderson (Alexis): forthcoming. Religion and the State: Initiating the Monarch in
Saivism and the Buddhist Way of Mantras. Heidelberg Ethno-Indological Series
I–II. Harrassowitz.Sastri (K.A. Nilakanta), 1984: The C�ol
_as. Madras, University of Madras. Reprint
of second, revised edition of 1955.
Siudmak (John E.C.), 1993: ‘‘Carved Rock Relief at N�adih�el in the Kashmir Valley.’’South Asian Archaeology 1991, pp. 637–646.
Siudmak (John E.C.), 1994: The Stylistic Development of the Sculpture of Kashmir.
D.Phil. thesis, Faculty of Oriental Studies, University of Oxford.Sivapriyananda (Swami), 1995: Mysore Royal Dasara. New Delhi, Abhinav Publi-cations. Photographs by Gajendra Singh Auwa.
Snellgrove (David L.) & Skorupski (Tadeusz), 1977: The Cultural Heritage ofLadakh. Warminster, Aris and Phillips.
Stein (Sir M.A.), 1961: Kalhan_a’s R�ajatara _ngin�ı: A Chronicle of the Kings of Kasm�ır:
Translated with an Introduction; Commentary and Notes. 2 vols. Delhi, Motilal
Banarsidass. Reprinted from the edition of 1900.Stuart-Fox (David J.), 2002: Pura Besakih: Temple;Religion and Society in Bali.Verhandelingen van het Koninklijk Instituut voor Taal-, Land en Volkenkunde
193. Leiden, KITLV Press.Tagare (Ganesh Vasudev), 1987: Historical Grammar of Apabhram
_sa. Delhi, Moti-
lal Banarsidass. Reprinted from the first edition of 1948.
Tod (James), 1920: Annals and Antiquities of Rajasthan or the Central and WesternRajput States of India. 3 vols. London, Humphrey Milford, Oxford UniversityPress, 1920. An annotated edition by William Crooke of the text first published
in 2 volumes in 1829 and 1832.Turner (R.L.), 1966:AComparative Dictionary of the Indo-Aryan Languages.London,Oxford University Press.
Vickery (Michael), 1998: Society, Economics and Politics in Pre-Angkor Cambodia.
The 7th–8th Centuries. Tokyo, The Toyo Bunko.
All Souls CollegeOxfordUK
300 ALEXIS SANDERSON