Ancestor Worship
Ancestor Worship
Ancestor Worship
1. Introduction
Post-mortem ancestor worship (śrāddha) is an essential component of a
brahmanical householder’s ritual routine. As has been discussed above,
Śaivism sought to extend its reach among the population of householders,
adopting core features of the brahmanical ritual repertoire even if they cre-
ated doctrinally awkward positions. This aspect was most notably reflected
in the context of funerary rites. It is also seen in the Śaiva adapting of śrād-
dha rituals. Here, as in the case of Śaiva antyeṣṭi, Śaiva ritualists had to
accept and work around a fundamental premise that was doctrinally prob-
lematic, namely, that during śrāddha rituals the soul is in essence being
venerated as an entity which passes through various ancestral levels after
death, and not as an entity that has attained ultimate liberation.1
This chapter contains a survey of early sources on Śaiva rites and ex-
plores their rationale. Notable is that amongst the early Śaiva Siddhānta
scriptural sources, only the Kiraṇa (Kir) contains prescriptions for Śaiva
rites for post-mortem ancestor worship. All the others avoid the topic,
which is likely an indication that their respective tantric circles were not yet
as integrated into the brahmanical mainstream while householder initiates
consulted conventional śrāddha priests for their services if required.
Sources that do contain descriptions of śrāddha rites date from the tenth
century onwards and mainly belong to the genre of ritual manuals. This
means that by this time tantric priests had more systematically started to
encroach on this sphere of rituals. Another post-tenth-century source is the
eclectic Śaiva scripture Bṛhatkālottara (BK). However, as mentioned
above, here the śrāddha section, together with a chapter on funerary proce-
1
See pp. 33ff.
178 LIBERATING THE LIBERATED
2
See Sanderson 2001, pp. 17 and 38–41 and p. 83. In his article, Sanderson points to
traces of specifically Vaiṣṇava formulations that were overlooked by the redactors of the BK
in the process of revising the passage to fit the Śaiva context.
3
The earliest known epigraphic attestation of a royal śivaśrāddha is in a South Indian
inscription of Rājarāja I (South Indian Inscriptions XII, No. 144 [A.R. No. 444 of 1918],
Tiruvenkadu, Shiyali Taluk, Tanjore District). It registers in detail the several gifts made to
the temple. In the third year of the reign of Rājarāja I, i.e. 987 CE, the queen of Uttama-
Chol, Sorabbaiyar Tribhuvanamadeviyar, presented a gold flower in lieu of a gift of land for
the śivaśrāddha of her husband. I am grateful to Alexis Sanderson for having pointed this
epigraph out to me.
ANCESTOR WORSHIP 179
4
This formulation of the three debts is commonly found in Dharmaśāstric literature.
Note that earlier texts define the debt to the ancestors also as the obligation to produce off-
spring, an obligation clearly linked to the maintenance of the patrilineage. For a detailed
discussion of the development of the concept of debts, see Sayers 2008, pp. 62–80.
5
See Kane 1953, pp. 380–383, for a short survey.
6
The aṣṭakas are defined as the eighth day of the dark half of three months in the year.
180 LIBERATING THE LIBERATED
2.1.1. Pārvaṇaśrāddha
Occasions
The days on which pārvaṇaśrāddha are to be performed are enjoined in the
YājS as: (a) the amāvāsyā, the night of the new moon; (b) the aṣṭakas, the
eighth day of the dark fortnight of the months Pauṣa, Māgha and Phālguṇa
according to the pūrṇimānta calendar, or Mārgaśīrṣa, Pauṣa and Māgha
7
Cf. part of the Mitākṣarā ad YājS 1.217–218: tac ca dvividhaṃ pārvaṇam ekoddiṣṭam
iti. tatra tripuruṣoddeśena yat kriyate tat pārvaṇam, ekapuruṣoddeśena kriyamāṇam
ekoddiṣṭam. punaś ca trividhaṃ nityaṃ naimittikaṃ kāmyaṃ ceti. tatra nityaṃ niyata-
nimittopādhau coditam aharahar amāvāsyāṣṭakādiṣu. aniyataṃ niyatopādhau coditaṃ
naimittikaṃ yathā putrajanmādiṣu. phalakāmanopādhau vihitaṃ kāmyaṃ yathā
svargādikāmanāyāṃ kṛttikādinakṣatreṣu. punaś ca pañcavidham aharahaḥśrāddhaṃ
pārvaṇaṃ vṛddhiśrāddham ekoddiṣṭaṃ sapiṇḍīkaraṇam ceti.
A kind of śrāddha that is omitted from Vijñāneśvara’s commentary is the śrāddha per-
formed at certain sacred sites, such as Gayā. These receive separate treatments in the
Purāṇas.
ANCESTOR WORSHIP 181
according to the amānta reckoning;8 (c) certain days during the dark half of
the lunar month (kṛṣṇapakṣa); (d) the winter and summer solstices; (e) the
two equinoxes; (f) the sun’s entry into a new sign of the zodiac
(sūryasaṃkrama); (g) the vyatīpātayoga, that is, when the sun and the
moon are equidistant from the solstice and on opposite sides of it; (h) days
on which certain conjunctions occur, such as the gajacchāyā, when the
moon is in the asterism Maghā and the sun in asterism hasta during the
thirteenth lunar day (tithi) of the dark half of the month; (h) and days on
which there is a solar or lunar eclipse.9 This list as presented in the YājS is
exhaustive, but over the course of time and according to different local
traditions, it is unlikely that all of these days were observed. For example,
according to this list, śrāddha would have to be performed at some point
during the dark fortnight of every month, a prescription found in early
sources such as the Śāṅkhāyanagṛhyasūtra.10 In contrast, however, later
sources indicate that the pārvaṇaśrāddha is to be performed – as is still
commonly done – only during a particular fortnight, namely the so-called
pitṛpakṣa (“the fortnight of the ancestors”) in the month of Āśvina (accord-
ing to the pūrṇimānta reckoning) or of Bhādrapada (according to the amān-
ta reckoning). It is this later practice that we find reflected in the Śaiva
sources.
Procedure
Invitation: On the day before the śrāddha is to take place, the person spon-
soring the rite, referred to as the yajamāna, should formally invite pious
and learned śrāddha Brahmins who are free of any defect11 to take part in
8
There are two calendrical systems in the Indian tradition: the pūrṇimānta system, in
which a lunar month begins with the day after the full moon (pūrṇima), and the amānta
system, in which a lunar month starts on the day of the new moon.
9
YājS 1.217–218: amāvāsyāṣṭakā vṛddhiḥ kṛṣṇapakṣo ’yanadvayam | dravyaṃ
brāhmaṇasaṃpattir viṣuvat sūryasaṃkramaḥ || vyatīpāto gajacchāyā grahaṇaṃ can-
drasūryayoḥ | śrāddhaṃ prati ruciś caiva śrāddhakālāḥ prakīrtitāḥ. The passage also in-
cludes the incidental and optional times for śrāddha, such as auspicious occasions like the
birth of a son, if one simply has the means for performing a śrāddha, when a Brahmin ar-
rives, or if one has the desire to perform a śrāddha.
10
Cf. Sayers 2008, p. 149.
11
The choice of appropriate invitees is crucial to accomplishing the rite. The YājS first
describes suitable candidates in 1.219–221, followed by a list in verses 1.222–224 of defects
182 LIBERATING THE LIBERATED
that must be avoided, including physical deformation, inappropriate birth, or the lack of
propriety and good conduct.
12
YājS 1.225: nimantrayeta pūrvedyur brāhmaṇān ātmavāñ śuciḥ | taiś cāpi saṃyatair
bhāvyaṃ manovākkāyakarmabhiḥ.
13
YājS 1.228: dvau daive prāk trayaḥ pitrye udag ekaikam eva vā | mātāmahānām apy
evaṃ tantraṃ vā vaiśvadevikam. According to Vijñāneśvara’s commentary on this verse, in
the absence of five suitable invitees, the yajamāna is to invite one Brahmin for the
Viśvedevas and another for the ancestors: pakṣāntaram āha ekaikam eva vā. vaiśvadeve
pitrye ca ekam ekam upaveśayet. Later in the commentary it reads: tantraśabdaḥ samudāya-
vācakaḥ iti. yadā tu dvāv eva brāhmaṇau labdhau tadā tu vaiśvadeve pātraṃ prakalpya
ubhayatraikaikaṃ brāhmaṇaṃ niyuñjyāt.
14
YājS 1.226: aparāhṇe samabhyarcya svāgatenāgatāṃs tu tān | pavitrapāṇir ācāntān
āsaneṣūpaveśayet. See also YājS 1.227c–228b: paristṛte śucau deśe dakṣiṇāpravaṇe tathā ||
dvau daive prāk trayaḥ pitrye udag ekaikam eva vā.
15
YājS 1.229: pāṇiprakṣālanaṃ dattvā viṣṭarārthaṃ kuśān api | āvāhayed anujñāto
viśve devāsa ity ṛcā.
ANCESTOR WORSHIP 183
16
YājS 1.230–232b: yavair anvavakīryātha bhājane sapavitrake | śan no devyā payaḥ
kṣiptvā yavosīti yavāṃs tathā || yā divyā iti mantreṇa hasteṣv arghyaṃ vinikṣipet | dattvoda-
kaṃ gandhamālyaṃ dhūpadānaṃ sadīpakam || tathācchādanānaṃ ca karaśaucārtham
ambu ca.
17
YājS 1.232c–233: apasavyaṃ tataḥ kṛtvā pitṝṇām apradakṣiṇam | dviguṇāṃs tu kuśān
dattvā hy uśantas tvety ṛcā pitṝn | āvāhya tadanujñāto japed āyantu nas tataḥ.
18
YājS 1.234: apahatā iti tilān vikīrya ca samantataḥ | yavārthās tu tilaiḥ kāryāḥ
kuryād arghyādi pūrvavat.
19
YājS 1.235: dattvārghyaṃ saṃsravāṃs teṣāṃ pātre kṛtvā vidhānataḥ | pitṛbhyaḥ
sthānam asīti nyubjaṃ pātraṃ karoty adhaḥ.
20
It is further specified that this rite should be performed in the manner of the pitṛyajña,
the śrauta rite of offering piṇḍas to the ancestors. Cf. Sayers 2008, pp. 92ff.; Kane 1953,
p. 201; see also Caland 1893.
21
Kane (1953, p. 444) has identified the full mantra from the Āpastambamantrapāṭha.
184 LIBERATING THE LIBERATED
should then take the hands of each Brahmin and have them stir the food
with their thumb, while he recites the mantra viṣṇur vicakrame.22
Offering the food to the Brahmins: The yajamāna is then to recite the
group of three mantras, the gāyatrī along with the mantra oṃ bhū bhuvaḥ
svāhā pronounced beforehand (savyāhṛtikām), and the triplet madhu vātā.23
He then is to invite the Brahmins to consume the suitable food as they
please, which they are to eat silently until they are satiated. Then he is to
recite purifying mantras and the same verses as before the meal. He then is
to pick up the food with the words, “May you be satiated”, and with their
permission take the rest of the food and scatter it around on the floor. Then
the Brahmins are each given some water for rinsing their mouths. After
this, the yajamāna is to gather all the leftovers and mix them with sesame
seeds while facing south.24
Offering of piṇḍas: Next, he should place the rice balls (piṇḍas) for the
ancestors close to the area of the leftovers; the manner of offering them is
to be done in the same way as during the Vedic sacrificial procedure for
ancestors, the śrauta pitṛyajña.25
Dismissing the ancestors: At the end he is to give the Brahmins the
ācamana and address them with svasti (“May it be well!”), and present
them “unfailing water”26 and their ritual fee (dakṣiṇā). He should then for-
mally request the Brahmins to pronounce the words astu svadhā (“May
there be svadhā”), which they then do. After this the officiant is to sprinkle
22
YājS 1.236–238: agnau kariṣyann ādāya pṛcchaty annaṃ ghṛtaplutam | kuruṣvety abhya-
nujñāto hutvāgnau pitṛyajñavat || hutaśeṣaṃ pradadyāt tu bhājaneṣu samāhitaḥ | yathālābhopa-
panneṣu raupyeṣu ca viśeṣataḥ || dattvānnaṃ pṛthivīpātram iti pātrābhimantraṇam | kṛtvedaṃ
viṣṇur ity anne dvijāṅguṣṭhaṃ niveśayet.
The mantra to be recited while the food is being stirred by the Brahmins is Ṛgveda
1.22.17 (see also Kane 1953, p. 444).
23
See Kane 1953, p. 444, who gives the verse references Ṛgveda 1.90.6–8, Vājasaneya-
saṃhitā 13.27‒29 and Taittirīya Brāhmaṇa 4.2.9.3 for the mantra starting with the phrase
madhu vātā.
24
YājS 1.239–242b: savyāhṛtikāṃ gāyatrīṃ madhu vātā iti tryṛcam | japtvā yathāsu-
khaṃ vācyaṃ bhuñjīraṃs te ’pi vāgyatāḥ || annam iṣṭaṃ haviṣyaṃ ca dadyād akrodhano
’tvaraḥ | ā tṛptes tu pavitrāṇi japtvā pūrvajapaṃ tathā || annam ādāya tṛptāḥ stha śeṣaṃ
caivānumānya ca | tad annaṃ vikired bhūmau dadyāc cāpaḥ sakṛt sakṛt || sarvam annam
upādāya satilaṃ dakṣiṇāmukhaḥ.
25
On the topic of piṇḍa offerings, see also Buss 2006.
26
In this procedure, the yajamāna offers them some water and formally requests the
Brahmins to declare it unfailing, which they then proceed to do.
ANCESTOR WORSHIP 185
the floor saying, “May the Viśvedevas be pleased.” Then the Brahmins are
to pronounce auspicious prayers for the family of the yajamāna, who
should bow and dismiss each of the ancestors starting with the father.
Dismissing the Brahmins: The yajamāna now turns the vessel in which
he collected the guest water (arghya) for the ancestors right side up and
ceremonially dismisses the Brahmins, following them to the boundary of
his land where he should circumambulate them a last time.
Concluding procedures: When the yajamāna returns home he is to eat
the remains of the food that was served for the ancestors.27 The piṇḍa balls
should be disposed of by feeding them to a cow, a ram or a Brahmin, or by
offering them to the fire or water.28 Both the yajamāna and the Brahmins
are to remain celibate that night.29
27
YājS 1.243b–249b: dadyād ācamanaṃ tataḥ | svastivācyaṃ tataḥ kuryād akṣayyoda-
kam eva ca || dattvā tu dakṣiṇāṃ śaktyā svadhākāram udāharet | vācyatām ity anujñātaḥ
prakṛtebhyaḥ svadhocyatām || brūyur astu svadhety ukte bhūmau siñcet tato jalam |
viśvedevāś ca prīyantāṃ vipraiś coktam idaṃ japet || dātāro no ’bhivardhantāṃ vedāḥ
santatir eva ca | śraddhā ca no mā vyagamad bahu deyaṃ ca no ’stu || *ity uktvoktvā (em.
Sanderson; ityoktoktvā Ed.) priyā vācaḥ praṇipatya visarjayet | vāje vāja iti prītaḥ
pitṛpūrvaṃ visarjanam || yasmiṃs te saṃsravāḥ pūrvam arghyapātre niveśitāḥ | pitṛpātraṃ
tad uttānaṃ kṛtvā viprān visarjayet || pradakṣiṇam anuvrajya bhuñjīta pitṛsevitam.
28
YājS 1.257: piṇḍās tu gojaviprebhyo dadyād agnau jale ’pi vā | prakṣipet satsu vi-
preṣu dvijocchiṣṭaṃ na mārjayet.
29
YājS 1.249cd: brahmacārī bhavet tāṃ tu rajanīṃ brāhmaṇaiḥ saha.
30
For an account of the offerings during the ten days following a death, see Kane 1953,
p. 262, and Mueller 1992, pp. 135–136.
186 LIBERATING THE LIBERATED
31
See Kane 1953, pp. 262–263.
32
Mitākṣarā on YājS 1.252: navaśrāddheṣu yac chiṣṭaṃ gṛhe paryuṣitaṃ ca yat | dam-
patyor bhuktaśiṣṭaṃ ca na bhuñjīta kadā cana.
33
Mitākṣarā on YājS 1.252: prathame ’hni tṛtīye ’hni pañcame saptame tathā | na-
vamaikādaśe caiva etan navaśrāddham ucyate.
34
The Gāruḍapurāṇa teaches further where these rites are to be performed. The first is
to be held at the place the person died, the second at the place where the funeral procession
stopped for a rest, and from the third onwards at the cremation site where the bones were
collected during the asthisaṃcayana. See Gāruḍapurāṇa 2.5.67–70: navaśrāddhasya te
kālaṃ vakṣyāmi śṛṇu kāsyapa | maraṇāhni mṛtisthāne śrāddhaṃ pakṣin prakalpayet ||
dvitīyañ ca tato mārge viśrāmo yatra kāritaḥ | tataḥ sañcayanasthāne tṛtīyaṃ śrāddham
ucyate || pañcame saptame tadvad aṣṭame navame tathā | daśamaikādaśe caiva nava
śrāddhāni vai khaga || śrāddhāni nava caitāni tṛtīyā ṣoḍaśī smṛtā | ekoddiṣṭavidhānena
kāryāṇi manujais tathā. See also Kane 1953, p. 263.
ANCESTOR WORSHIP 187
ble” (akṣayyam astu),35 and the verbal endings are changed from the plural
to the singular in the formulas used to address the one Brahmin represent-
ing the deceased person.
After the series of ekoddiṣṭaśrāddhas, the sapiṇḍīkaraṇa is held, in
which the deceased, who is still a ghost (preta), is incorporated into the
ancestral line. According to the YājS, the basic procedure is again that of
the pārvaṇaśrāddha, with an additional rite in which four vessels are pre-
pared with water mixed with sesame seeds and perfume. Three of these
vessels represent each of the ancestors. The fourth represents the newly
deceased; from this vessel water is poured into each of those of the ances-
tors, signifying the deceased person’s incorporation into the ancestral line.36
However, there are two significant features not mentioned in this text that
are standard in many other accounts. First, in addition to the two Brahmins
representing the Viśvedevas and the three representing the ancestors, an-
other Brahmin is invited to represent the deceased. Secondly, there is an-
other, later more common rite marking the incorporation of the deceased
into the ancestral line, in which four rice balls (piṇḍas) are first offered to
the three ancestors and the recently deceased, and then the piṇḍa for the
recently deceased is divided into three parts, each of which is then com-
bined with one of the piṇḍas for the ancestors.37
The timing of the ekoddiṣṭaśrāddhas and the sapiṇḍīkaraṇa are interre-
lated, since the latter can be performed only after the series of navaśrād-
dhas and the sixteen ekoddiṣṭaśrāddhas have been completed. In the brah-
manical literature there is a great deal of divergence as to when these
should be held.38 The standard view expressed in the sources is that they are
to be held throughout the year after death, according to the YājS, every
35
YājS 1.252–252: ekoddiṣṭaṃ daivahīnam ekārghyaikapavitrakam | āvāhanāgnau-
karaṇarahitaṃ hy apasavyavat || upatiṣṭhatām akṣayyasthāne vipravisarjane | abhiramya-
tām iti vaded brūyus te ’bhiratāḥ sma ha ||.
36
YājS 1.253–254b: gandhodakatilair yuktaṃ kuryāt pātracatuṣṭayam | arghyārthaṃ
pitṛpātreṣu pretapātraṃ prasiñcayet || ye samānād iti dvābhyāṃ śeṣaṃ pūrvavad ācaret.
37
See, for example, Kane 1953, pp. 521–523; Sayers 2008, pp. 152–154; and Buss
2006. Even though this rite is part of many standard accounts of the sapiṇḍīkaraṇa, both the
Agnipurāṇa (2.163.26–27) and the Nāradamahāpurāṇa (51.136b–138c) incorporate this
passage of the YājS and therefore also omit this rite. See Michaels 2005 for a comprehen-
sive account, including video footage, of a present-day version performed in Nepal.
38
See Kane 1953, pp. 517–519 for a summary of the brahmanical view on this matter.
188 LIBERATING THE LIBERATED
month on the anniversary of the death.39 However, the most common view
is that found in Vijñāneśvara’s commentary, which rules that sixteen
ekoddiṣṭaśrāddhas are to be performed: on the twelfth day after the death,
after three fortnights, after six months, once every month, and on the first
anniversary of the death. 40 Alternatively, to be able to perform the
sapiṇḍīkaraṇa before a year has elapsed, it is also possible to compress the
performance of the ekoddiṣṭaśrāddhas into a shorter period.41 The literature
gives different days for the sapiṇḍīkaraṇa, but most commonly – if the
series of ekoddiṣṭaśrāddhas are compressed in time – it is to be performed
during the first month after the death. Vijñāneśvara, for example, quotes the
Āśvalāyanagṛhyasūtra, where it is ruled that the sapiṇḍīkaraṇa can be per-
formed either a year following the death or after twelve days. If the latter
case is chosen, the entire series of ekoddiṣṭaśrāddhas is to be performed on
the eleventh day following the death. However, the YājS points out that if
the day of the sapiṇḍīkaraṇa takes place earlier than the end of the year, the
daily or monthly offerings of food and water to the deceased person none-
theless must be carried out for the whole year.42 The issue is thus very
complex, and this is just a brief example of the various regulations. For our
aim of comparing these procedures to the Śaiva śrāddha rituals, however, it
is important to note that even within the brahmanical sphere there was no
39
Cf. YājS 1.256ab: mṛte ’hani tu kartavyaṃ pratimāsaṃ tu vatsaram.
40
See Mitākṣarā on YājS 1.255: śoḍaśa śrāddhāni ca. dvādaśāhe tripakṣe ca ṣaṇmāse
māsi cābdike | śrāddhāni ṣoḍaśaitāni saṃsmṛtāni manīṣibhiḥ.
41
Mitākṣarā on YājS 1.255: yadā prāk saṃvatsarāt sapiṇḍīkaraṇaṃ tadā ṣoḍaśa
śrāddhāni kṛtvā sapiṇḍīkaraṇaṃ kāryam. Later in the commentary, Vijñāneśvara mentions
that according to some, compressing the ekoddiṣṭaśrāddhas is done in times of need: yadā tv
āpatkalpatvena prāk sapiṇḍīkaraṇāt pretaśrāddhāni karoti tad ekoddiṣṭavidhānena kuryāt.
Neither the YājS nor Vijñāneśvara specify how these ekoddiṣṭaśrāddhas are to be ar-
ranged in this case. The Kashmirian Dvādaśāhavidhiḥ (Srinagar ORL MS 1677
Pārthivapūjā, etc.), in which the sixteen ekoddiṣṭaśrāddhas are compressed into twelve days
from the eleventh day after the death, which marks the end of impurity, through to the twen-
ty-second, by performing one of the monthly ekoddiṣṭas every day, and adding the
ekoddiṣṭaśrāddha of the third fortnight on the second day, that of the fifth fortnight
ekoddiṣṭaśrāddha on the third day, that of the day before the end of the sixth month on the
sixth day, and that of the day before the end of the year on the twelfth.
42
YājS 1.255: arvāk sapiṇḍīkaraṇaṃ yasya saṃvatsarād bhavet | tasyāpy annaṃ soda-
kumbhaṃ dadyāt saṃvatsaraṃ dvije. Mītākṣara ad loc.: saṃvatsarād arvāk sapiṇḍīkaraṇaṃ
yasya kṛtaṃ tasya taduddeśena pratidivasaṃ pratimāsaṃ vā yāvat saṃvatsaraṃ śak-
tyanusāreṇānnam udakumbhasahitaṃ brāhmaṇāya dadyāt.
ANCESTOR WORSHIP 189
43
YājS 1.242c–243a: ucchiṣṭasannidhau piṇḍān dadyād vai pitṛyajñavat || mātāmahānām
apy evaṃ.
44
YājS 1.253–254cd: etat sapiṇḍīkaraṇam ekoddiṣṭaṃ striyā api.
45
Mitākṣarā ad YājS 1.254: striyāpi mātur api kartavyam. evaṃ vadatā pārvaṇe
mātṛśrāddhaṃ pṛthak kartavyam ity uktaṃ bhavati.
46
Parts of this section have been adapted from Mirnig 2013, pp. 285–288.
47
Sanderson 1995, pp. 34–36.
190 LIBERATING THE LIBERATED
And further:
He should perform the śrāddha because the initiation ritual (dīkṣā) ends
with it. [Thus] the śivaśrāddha that consists of the five brahmamantras
has been taught in brief.51
As discussed above,53 the claim that śrāddha rituals are somehow necessary
to achieve the initiate’s liberation is highly problematic from a doctrinal
48
See, e.g., Brunner-Lachaux 1979, p. 626.
49
The mantras for addressing the ancestors are defined in the SP and KKD as consisting
of the oṃ syllable, the name of the ancestral deity in the dative form, and namaḥ at the end.
Trilocana clarifies this by adding the heart mantra and the object offered to the ancestor in
the accusative, as for example oṃ hāṃ īśāya pitre idam āsanaṃ namaḥ. See Brunner-
Lachaux, p. 647, [17b].
50
Kir 61.34ab. For the text and annotated translation of the passage, see the appendices.
51
Kir 61.41. For the text and annotated translation of the passage, see the appendices.
52
JR ŚP 105ab. For the text and annotated translation of the passage, see the appendices.
53
See pp. 33ff.
ANCESTOR WORSHIP 191
point of view. Even if one accepts the already problematic view that Śaiva
cremation has merely an expiatory function, ensuring that the soul is not
held back by transgressions performed since its nirvāṇadīkṣā, any further
rites for guiding the soul through various divine manifestations should no
longer be necessary. Theoretically, their performance should be virtually
superfluous. In view of this, the tenth century Kashmirian non-
Saiddhāntika Śaiva commentator Abhinavagupta was not hesitant in stating
that the Śaiva śrāddha is potentially useless. While he includes an entire
chapter of Śaiva śrāddha rituals in his Tantrāloka (TĀ), he nevertheless
explicitly states that śrāddha serves no purpose for the truly enlightened,
linking this to the case of a jñānin, that is, an initiate of the highest kind
who has attained liberation through gnosis:
Therefore, O Garuḍa, one should perform this rite [of post-mortem an-
cestor worship] even when [the corpse] cannot be found. For how can
anyone become free of debt as long as [the śrāddha] is not performed?55
54
TĀ 25.10.
55
Kir 61.39. For the text and annoated translation of the passage, see the appendices.
192 LIBERATING THE LIBERATED
Next, I shall teach the śrāddha ritual, which removes the debt to the
teachers and so forth. And [the following] has been taught: The initiate
should pay [his] debt to the deities by fully performing rites of worship,
oblations and the like; to the sages, too, [he pays his debt by] being of
unwavering conduct at all times; and to [his] gurus and the like he
should [pay the debt] by performing the rite of śrāddha.56
56
JR ŚP 1–2. For the text and annotated translation of the passage, see the appendices.
57
On this, see also Sanderson 1995.
ANCESTOR WORSHIP 193
58
In the Kir and the KKD, the śrāddha Brahmins are replaced by ācāryas for the ances-
tors and sādhakas for the Viśvedevas, who are now named Rudra and Ananta; see Kir 61.5–
6 (see appendices). See also KKD T 370, p. 239: tatra pitrārthaṃ deśikān viśvedevārthaṃ
sādhakau […] gṛhṇīyāt. The SP specifies that ascetics are to take the roles of Īśa, Sadāśiva
and Śānta; SP 11.3–4b: liṅgino brāhmaṇādyāś ca śrāddhīyāḥ śivadīkṣitāḥ | īśaḥ sadāśivaḥ
śāntaḥ śivaśrāddhe tapasvinaḥ || rudraś cānantanāmā ca viśvedevau sthitau dvayoḥ. In the
absence of suitable candidates, the śrāddha can be performed alternatively with two initiates
or one, in whom the respective deities are installed on the head, shoulders and arms, just as
in the brahmanical model. SP 11.6c–8b (in Brunner-Lachaux 1977): pañcābhāve tu sarvatra
yujyate puruṣadvayam || śiraḥskandhavibhāgena trayam ekatra kalpayet | bāhumūlavibhedena
viśvedevau dvitīyake || ekatra bāhuyugmena saha vā pañca kalpayet.
59
Cf. YājS 1.229–232b.
60
Cf. YājS 1.232c–234.
194 LIBERATING THE LIBERATED
For detailed accounts of the Śaiva śrāddha rituals, the Kir and JR are
representative, since structurally their procedures correspond to the brah-
manical model described above, with only minor variations.61 These do not,
however, necessarily reflect modifications specific to the Śaiva setting;
they are similar in kind to variations also found in brahmanical sources,
such as differences in the number and shapes of maṇḍalas to be drawn for
receiving the śrāddha priests.62 The only notable Śaiva-specific addition is
found in the JR. Here, a worship of the liṅga, the cult object in which Śiva
is commonly venerated, is integrated into the ritual sequence, taking place
after the invitees have been seated and before the rice balls (piṇḍa) have
been offered to the ancestors.
An important category of brahmanical śrāddha rituals that is largely
missing in the early Śaiva authoritative accounts is that of prescriptions for
61
Other clear variations, such as differences in the order of ritual actions, will be dis-
cussed at the relevant places in the annotated translations.
62
This concerns the procedure for welcoming the śrāddha priests, an occasion for which
maṇḍalas are drawn in the eastern part of the ritual area where the invitees are ceremonially
welcomed and presented with a mixture of earth and water to wash their feet, and water to
perform the ācamana. We find the following variations: Kir 61.14c–15b enjoins that two
maṇḍalas, one round and one square, be drawn for this reception in the south and north,
respectively. In the first round maṇḍala, a mixture of water and earth for washing the feet is
offered; in the second square maṇḍala, ācamana water is presented to the invitees. In con-
trast, in the early Śaiva Siddhānta manuals four maṇḍalas are enjoined at this point; two
square maṇḍalas for Ananta and Rudra – the Śaiva equivalent of the Viśvedevas – and two
round maṇḍalas for the divine Śiva manifestations representing the initiated ancestors. This
step of drawing maṇḍalas to receive the śrāddha priests is omitted in the brahmanical YājS
and Vijñāneśvara’s commentary, but appears in other authoritative brahmanical sources,
such as the Nāradīyapurāṇa. Here it is enjoined that maṇḍalas be drawn for receiving the
guests, with their shape depending on the yajamāna’s caste: if the yajamāna is a Brahmin, a
square maṇḍala is drawn; if he is a Kṣatriya, a triangular maṇḍala; if he is a Vaiśya, a circu-
lar maṇḍala; and if he is a Śūdra, no maṇḍala is drawn, but the ground is prepared by sprin-
kling it with consecrated water. Nāradīyapurāṇa 1.28.32c–33: śrāddhārthaṃ samanujñātaḥ
kārayen maṇḍaladvayam || caturaśraṃ brāhmaṇasya trikoṇaṃ kṣatriyasya vai | vaiśyasya
vartulaṃ jñeyaṃ śūdrasyābhyukṣaṇaṃ bhavet. See also Kane 1953, pp. 456–457.
Another example of a ritual feature that varies is the standard brahmanical procedure of
the invitees stirring food with their left thumb in the counter-clockwise direction (i.e. the
direction signalling an inauspicious occasion). This is entirely absent in the Kir, JR and BK
quoted therein. It is, however, found in the SP (11.26 nyubjaṃ savyaṃ tadaṅguṣṭham
amṛtāyānnasaṃgatam | vāmaṃ mṛtyujitaṃ vāpi parito bhrāmayet paṭhan) and the KKD
(Brunner-Lachaux 1977, p. 655, [24b]: teṣāṃ savyam aṅguṣṭhaṃ svahastena saṃgṛhya,
nyubjam annasaṅgataṃ kṛtvā, mṛtyuñjayaṃ vāmadevaṃ vā paṭhan paribhrāmya…). A
difference here, however, is that the Vedic mantra viṣṇur vicakrame recited over the food
has been replaced by the tantric Śaiva mantras mṛtyuñjaya or vāmadeva.
ANCESTOR WORSHIP 195
63
KKD (T 370, p. 244; T 403, p. 174): mātṛśrāddhaviṣaye bhojanavarjyam anyat sar-
vaṃ *pūrvoktena (T 370; pūrvānte T 403) kuryāt.
64
This is not the case for the śrāddha rituals preserved in the manuals in use until recent
times by Śaiva priests in Kashmir. There, the female ancestors, the mātṛs, are incorporated
as the three highest of the eight Śāktis surrounding Manonmanī, namely, Balavikaraṇī,
Balapramathanī and Bhūtadamanī at the highest level of the throne of Śiva. Nevertheless,
while they have thus been given an independent spiritual identity, they are situated at a
cosmic level below that of Īśvara, Sadāśiva and Śiva. See Sanderson 1995, p. 35.
65
Kir 61.37: ekoddiṣṭaṃ yad atroktaṃ kāryam āvāhanaṃ vinā | navaśrāddhādikaśrād-
dhaṃ proktam etat samāsataḥ.
196 LIBERATING THE LIBERATED
66
See p. 198.
67
See Sanderson 1995, p. 35.
68
See NGMPP B24/57, ff. 224r2–227v5.
69
See p. 83.
ANCESTOR WORSHIP 197
According to the JR, during the first ten days after a death, daily offerings
of rice pudding are cooked and offered either in the house of the deceased
or at the waterside in the presence of an invited ācārya.70 The yajamāna
receives the invitee and performs worship of Śiva, followed by a simplified
worship of Śiva in the fire, using only the heart mantra without its being
installed. A piṇḍa is then offered to the recently deceased, with a formula
constructed with a different mantra each day that incorporates the deceased
person’s name and ends with svadhā. For the first five days the officiant is
to use the five aṅgamantras starting with the heart mantra (hṛdayamantra)
and ending with the netramantra,71 and for the last five days, the five face
mantras starting with sadyojāta.72 After offering the piṇḍa, the officiant is
to pour some food mixed with sour milk, milk and clarified butter into the
Brahmins’ hands or a vessel made out of silver or gold. After sprinkling the
piṇḍa and dismissing the recently deceased, he is then to pour a full obla-
tion73 and hand a jar filled with water and sesame seeds together with
clothes and gold to the officiant. At the end he is to dispose of the piṇḍa
into water or fire, and finish with the concluding rites at the site. After this
he should return home for a communal meal with other initiates, at the con-
clusion of which he is to pronounce the aghoramantra, scatter white mus-
tard seeds and prepare his bed.
4.1.1.2. Ekoddiṣṭaśrāddha
70
JR ŚP 42c–63. For the text and translation of the passage, see the appendices.
71
Since there are actually six aṅgamantras, the injunction to recite those starting with
the hṛdaya and ending with the netra implies that the weapon mantra, which would follow
next, is not included.
72
JR ŚP 63. For the text and translation of the passage, see the appendices. The five
aṅgamantras are the hṛdaya, śiras, śikhā, kavaca, and netra; the five face mantras are
sadyojāta, vāmadeva, aghora, tatpuruṣa, and īśāna. See Brunner 1963, p. 331, Appendix VI
and s.v. aṅga and aṅgamantra in TAK 1.
73
See p. 145, fn. 72.
74
JR ŚP 64–79. For the text and translation of the passage, see the appendices.
198 LIBERATING THE LIBERATED
or two weeks during the first year after a death.75 As in common brahmani-
cal practice, this kind of śrāddha is distinguished from the pārvaṇaśrāddha
by the fact that only one person is invited for the ceremony to represent the
recently deceased. The invocation of the ancestral deities is consequently
omitted.
The ācārya is invited in. Seated facing north, he conducts the piṇḍa of-
fering followed by an oblation into the fire. The yajamāna is then to offer
the ācārya some food, and with the leftovers together with the arghya wa-
ter, make a bali offering to appease the spirits (bhūta) near the ritual site.
At the end he offers the guru some water, sprinkles the ground and pays the
guru his ritual fee (dakṣiṇā). The guru then performs a worship of Śiva.
The rite is concluded with a full oblation76 poured into the fire while recit-
ing the śivamantra.
75
The text in JR ŚP 65–66 literally says that the ekoddiṣṭaśrāddha is to be performed on
a monthly basis (pratimāsika). However, in light of standard practice, it is unlikely that
Jñānaśiva meant this to imply that they were to be performed only once a month, but rather
fifteen or sixteen times during the first year. Further, as we have seen, another term com-
monly used to refer to these ekoddiṣṭaśrāddhas is māsānumāsikaśrāddha, which may have
led to the expression pratimāsika in the current passage. In this case, however, the reading is
awkward, so its interpretation remains uncertain. A similarly vague instruction is also found
later in the passage, in JR ŚP 79a, where it is enjoined that the ekoddiṣṭas are to be per-
formed every month through the thirteenth.
76
See p. 145, fn. 72.
ANCESTOR WORSHIP 199
370 and T 403 contain slightly different chapter divisions and insertions at
this point. It is therefore possible that some passages may not be part of the
original work, but have been added by later redactors. For instance, we find
quotations from the Kāmikāgama in this section. These are clearly later
insertions, since the Kāmikāgama belongs to the second wave of Saiddhān-
tika scriptural texts that emerged only after Aghoraśiva’s lifetime.77 Fur-
ther, a colophon marking the end of the funeral rites is found in these tran-
scripts twice: at the end of the account of the nirvāṇāntyeṣṭi as well as at
the end of the account of the pārvaṇaśrāddha. This introduces uncertainty
as to where the end of the section on funerary rites was originally con-
ceived. With these hesitations in mind, we can nonetheless note that Ag-
horaśiva included some rituals in his manual that appear to have been
unique to the southern ritual context, similar to the cūrṇotsava discussed
above. 78 Analogous to this case, it is therefore conceivable that the
pāṣāṇasthāpanavidhi and rudrabali procedures, which we only find in the
KKD, are in fact original. Further, the detailed and explicit instructions at
this point leading up to the regular śrāddha rites would fit Aghoraśiva’s
general approach of providing elaborate instructions amounting to a com-
prehensive practical guide that could be used independently. In general, it
appears that the KKD propagates a Śaiva system that was independent from
the brahmanical and covers all rituals in detail. Therefore to include an
outline of the procedures set out in these passages seems useful, despite the
uncertainties regarding the text.
4.1.2.1. Pāṣāṇasthāpanavidhi79
77
For the place of works such as the Kāmikāgama in Śaiva literature, see, e.g., Sander-
son 2014, p. 89.
78
See p. 119.
79
KKD T 370, pp. 232–234 and T 403, pp. 160–62.
80
KKD T 370, p. 233: …maraṇāhāt prabhṛti daśāhaparyantam evaṃ vidhāya…
200 LIBERATING THE LIBERATED
cation at the funeral site on the day of cremation, or whether it signifies the
collection of rites involving the daily offerings of water and food to the
deceased over the entire ten days after the death (the daśakriyā). Hence, it
is not clear from the instructions whether the pāṣāṇasthāpanavidhi is to be
performed in place of the brahmanical daśakriyā or as an addition to it.81
On the first day, immediately after the tarpaṇa, the officiant is to build a
small construction called a pitṛgṛha (house for the ancestors) either in-
doors, in a courtyard, or in the open air. In this pitṛgṛha he installs a throne
for worship and next to it, digs a hole in the ground of the same size,
whereupon he is to smear both with cow dung.82 Next, he is to make a bun-
dle of either seven blades of darbha grass or five blades of kuśa grass, and
on it installs the oṃ syllable as a throne. On this he invokes the body of the
deceased with his initiatory name followed by the exclamation svāhā, and
then invokes the soul using the ātmamantra83 with the name of his clan
(gotra) followed by namaḥ. He then greets the deceased person’s soul with
81
Of the available post-12th-century South Indian sources, only the Dīkṣādarśa (T 153,
pp. 589–594) contains this procedure, which it attributes to the Kāmikāgama. However, the
account is much shorter and most of the passage is either corrupt or missing and is thus not
of much help for reconstructing this procedure. It appears that the redactor himself was
unsure about the original contents. Further, it is unclear what the Dīkṣādarśa’s redactor
understood the pāṣāṇasthāpana to be, since the surviving text does not mention a stone, and
the section contains much more information than just the rite described in the KKD. In the
same chapter, the Dīkṣādarśa also quotes the Kāmikāgama concerning rulings about the
days of impurity for the various castes, and the timing of the ekoddiṣṭaśrāddhas and the
navaśrāddhas.
82
KKD (T 370, pp. 231–232; T 403, p. 160): *atha (T 403; om. T 370) prathamadivasa-
tarpaṇānantaram eva gṛhe vā catvare vā bahir vā pitṛgṛhaṃ vidhāya, *tasmin (T 403;
tasmāt T 370) mekhalātrayayuktaṃ pīṭham *uttarābhimukhaṃ vidhivad (T 370; om. T 403)
vidhāya, pīṭhasaṃnidhau tanmānena gartaṃ *nikhanya (T 403; nighanya T 370), tad ubha-
yaṃ gomayenopalipya…
According to T 370, the officiant is to place seven grains into a jar and invoke into it
Rudra, Īśvara or Sadāśiva, presiding over the various pādas according to the level of initia-
tion the deceased person has received and thus his spiritual level: pīṭhamadhye
…*saptadhānyoparikumbhaṃ saṃsthāpya tasmin dīkṣānurūpaṃ tattatpadādhiṣṭhāyakān
rudreśvarasadāśivāṃś cāvāhya…
83
According to Tantrikābhidhānakośa Vol. I (p. 185), the ātmamantra is a synonym for
the ātmabīja (= haṃ); the ātmamantra is hence oṃ haṃ ātmane namaḥ. If the same applies
here, one might speculate that the mantra in this context is to be recited in the form of oṃ
haṃ ātmane amukagotrāya namaḥ.
ANCESTOR WORSHIP 201
offerings of water for washing its feet, water for the ritual sipping (āca-
mana) and guest water, together with clothes, flowers, incense and lamps.84
From this point, the two transcripts provide different procedures. In
T 370, the officiant is to dig a pit, install a jar in it and then place stones
there. He then invokes the soul of the deceased onto the stones, sprinkles
the pit with water and makes offerings such as scented powder. Then a
woman who has taken a bath and is free of impurity should cook some rice
with water or the guest water in a new pot. The officiant should then sweet-
en this with molasses and present this offering together with betel nuts.
There are different options for when the offerings are to be presented,
namely, during the morning, at dawn, or both.85 In T 403, the order is re-
versed. First the food offerings are prepared and then a pit, into which,
instead of several stones, a single brick (iṣṭakā) is placed. The brick is
where the deceased person’s soul is invoked.86
84
KKD (T 370, p. 233; T 403, p. 160): …paścāt (T 370; om. T 403) *saptadarbhair (T
370; saptadarbhena) vā pañcabhiḥ kuśair vā prādeśamānena nirmitaṃ (T 403; nirmita T
370) kūrcaṃ saṃsthāpya, tadupari *praṇavenāsanaṃ (T 370; praṇavāsane T 403) saṃkal-
pya, tasmin mṛtasya dīkṣānāmnā *svāhāntena (T 403; svāhoktena T 370) mūrtiṃ
samāvāhya, *tathaiva (T 403; om. T 370) tadgotranāmayuktenātmamantreṇa na-
montenātmānaṃ sampūjya, pādyācamanārghyāṇi dattvā, *vastrapuṣpair (T 403; vastra-
puṣpa T 370) alaṃkṛtya, dhūpadīpau dattvā…
85
The times for the different offerings are not clear from the text. While they should be
made in the morning and at dawn, the food and piṇḍa are either to be offered in the morning
and at dawn, or twice in the morning. In either case, they are to be offered with a formula
containing the heart mantra, the initiation name, and svāhā at the end.
86
KKD (T 370, p. 233; T 403, pp. 160–161): …{<-T 370a/T 403b->} *tato garte
śarāvaṃ saṃsthāpya (T 370; pīṭhaṃ vinā garte śarāvaṃ saṃsthāpya T 403), *tanmadhye
śilāḥ saṃsthāpya, tāsu *mṛtasyātmānam (conj.; mṛtāmātmānam T 370, T 403) āvāhya (T
370; tasmin *iṣṭakāṃ [em.; iṣṭakaṃ codd.] saṃsthāpya tasmin mṛtam ātmānam āvāhya T
403) *ābhiṣicya gandhādibhiḥ saṃpūjya (T 370; āvāhya sampūjya piṇḍatrayaṃ dattvā T
403) {<-T 370a/T 403b->}, {<-T 370b/T 403a->} *paścāt (T 370; om. T 403) *taṇḍulam
udakena (em. Sanderson; taṇḍulā ukena T 370; taṇḍulādikena T 403) vā tadarghena vā
*prītyarthaṃ (T 370; tatprītyarthan T 403) navapātre *āśaucarahitayā (conj.;
āśaucārahitayā T 370; agnau ca rahitāyā T 403) snātayā pākaṃ *kārayitvā (T 403;
karayitvā T 370), gulaghṛtatilayuktaṃ kṛtvā, †*sopadaṃśaṃ (T 403; sopadeśaṃ T 370)
saṃnivedya (T 370; nivedya T 403)† tāṃbūlaṃ dattvā evaṃ *prātaḥsāyāhṇayor (conj.;
sāyāhṇayor T 370; prates sarayāṇabhayo T 403) nivedya piṇḍaṃ dattvā athavā, pūrvāhṇe
vāpi kāladvayaṃ pratikuryāt. hṛdbījayuktena tasya *dīkṣānāmayuktena svāhāntena (T 370;
dīkṣnāmnā svāhāntaṃ T 403) sarvaṃ nivedya athavā *prātar arghyaṃ (T 403; prātarghyan
T 370) dattvā, aparāhṇe *nivedya piṇḍaṃ dattvā (T 370; naivedyaṃ dattvā athavā T 403)
{<-T 370b/T 403a->}…
202 LIBERATING THE LIBERATED
The two transcripts then converge again. On the tenth day, after having of-
fered the libations in the manner stated, the officiant, wearing the sacred thread
over the right shoulder and holding a blade of kuśa grass, should go to a body
of water, lay down darbha grass and install the mantras on his hands in the
reverse order.87 After this, the text becomes unclear and seems to be corrupt.
From the available passage we can infer that the officiant is to offer libations to
the deceased, whereby the deceased is addressed in the three divine ancestral
manifestations that he will successively take on in the future.88
At the end of this, the officiant performs the sakalīkaraṇa and returns
home, where he is to worship Śiva and offer another piṇḍa and some obla-
tions, and then dismiss the deity. Then the officiant is to dispose of every-
thing by taking the throne and stone together with the bundle of grass and
the flowers on a bier, cover it with pieces of cloth and with incense, lamps
and music, and take it to a pure site at a body of water. Then, facing north,
he is to throw the stone, the piṇḍas and the throne over his shoulder into the
water.89 Finally, he should take a purifying bath and return home, where he
performs some more concluding rites such as Śiva worship.90
87
KKD (T 370, p. 233; T 403, p. 161–162): … daśāhe ’pi pūrvavat saṃtarpya *jalatīre
(T 370; jale T 403) darbhān *āstīrya (T 403; āstārya T 370) apasavyopavītī *pavitrapāṇir
(T 403; pavitrapāṇī T 370) viparītakrameṇa karanyāsaṃ kṛtvā…
88
The passage provides an incomplete list of mantras with which offerings to the de-
ceased are to be made. While the available transcripts of the text preserve only the option for
a deceased samayin, a list is given of Śiva manifestations that mixes up the terminology to
be applied in the cases of a samayin and of a full initiate. It prescribes, in ascending order,
the deceased person’s future manifestations as Skanda, Caṇḍa and, instead of Gaṇeśvara,
Śānta, who is the highest manifestation of Śiva and is only used at the great-grandfather
level for the full initiate. KKD (T 370, pp. 233–234; T 403 p. 161): …samayinaś cet oṃ hāṃ
śuddhātman skando bhava *svadhā (T 403; svāha T 370) *oṃ śuddhātman caṇḍo bhava
svadhā (T 403; om. T 370) oṃ hāṃ śuddhātman śānto bhava *svadhā (T 403; svāha T 370)
*†tulaṃ† kuśodakaṃ svadhā, oṃ vāsodakaṃ svadhā, oṃ tilodakaṃ svadhā (T 403; om. T
370) iti saṃtarpya…
89
KKD (T 370, p. 234, and T 403, pp. 161–162): kūrcapuṣpasamāyuktaṃ pāṣāṇaṃ ca
tathaiva ca || pīṭhaṃ *cā sahasoddhṛtya (T 370; cāhasoddhṛya T 403) *yānam (T 370;
dhāmnā T 403) āropya śobhanam | vastrair ācchādya bahuśo dhūpadīpasamanvitam ||
sarvātodyasamāyuktaṃ sarvaśobha*samanvitam (T 370; -samantritam T 403) | utthāpya
*tena yānena (T 403; tena toyena T 370) puṇyatīrthe vinikṣipet || *udaṅmukhas (T 403;
udaṅmukhan T 370) tatas tasmin *śiṣyas (T 403; śiṣyan T 370) tattīrthamadhyame |
*pāṣāṇaṃ (T 403; pāṣāṇa T 370) piṇḍapīṭhe ca paścādbhāge vinikṣipet.
90
While there are some concluding instructions for when the officiant returns home
from the water site, it is not clear from the witnesses available to me what these entail. First,
some kind of worship of yogins takes place, followed by asking the deity for forgiveness.
ANCESTOR WORSHIP 203
4.1.2.2. Rudrabali
The rudrabali is a rite prescribed for the tenth day after a death. It consti-
tutes a bali offering to the eleven Rudras. It is not apparent from the text
what the function and application of this rite are, and I have as yet not en-
countered any procedure in the brahmanical literature or in the pre-twelfth-
century Saiddhāntika sources that correspond to it. Further, the two tran-
scripts contain differing instructions at the beginning and end of this sec-
tion that imply different functions for the rite. This suggests that there may
have been confusion about this rite even amongst the redactors. Transcript
T 370 treats the rudrabali as if it were an additional rite after the
pāṣāṇasthāpanavidhi. Transcript T 403, on the other hand, appears to treat
this rite as a procedure performed in cases of inauspicious death (dur-
maraṇa).91 The rudrabali would thus correspond to the sections on rites for
inauspicious deaths found in the JR as well as, in an earlier section, in the
KKD itself. However, these rites are embedded within the cremation pro-
cedure itself; they are not to be performed on the tenth day after it.
The instructions for the rite are as follows: On the tenth day, the officiant
is to go to a pure site close to a body of water such as a river and prepare
himself for the rite, holding a blade of kuśa grass and carrying out the
prāṇāyāma. On top of a sthaṇḍila he should draw a lotus diagram with eight
petals, and place a jar (kumbha) in the middle of it into which he is to invoke
Rudra.92 Then he is to install eleven jars, placing them in the anti-clockwise
Next, T 370 seems to envisage a communal meal, T 403, some other kind of worship: KKD
(T 370, p. 234; T 403, p. 162): tataḥ snātvā gṛhaṃ gatvā sarveṣāṃ yoginām api | pūjāṃ
kṛtvā kṣamāpyātha *svayaṃ bhṛtyaiś ca bhojayet (conj.; svayaṃ bhūtyaiś ca bhojayet T 370;
svayamantivyavat pūjayet T 403).
91
This passage is highly corrupt and so I have been unable to reconstruct it satisfactori-
ly. After the list of the Rudras, in a series in the instrumental case, the text seems to list ways
of dying that the manual earlier described as unfortunate deaths. At the end of the account,
the passage seems to provide instructions concerning the timing of the rudrabali. KKD, T 403,
pp. 163–164: …oṃ hāṃ devo haro rudraḥ śaṃkaro nīlalohita īśāno vijayo bhīmo devadevo
bhavodbhavaḥ, kapālī ceti vijñeyā rudrā ekādaśās tathā sarpājagaravāyvagnibhiḥ śastra-
vṛkṣaviṣatoyahastinā. vatsare ṣaṭtrimāse pakṣayos taddinaṃ tu vidhaṃ rudrabaliḥ.
92
KKD (T 370, p. 234; T 403, p.162): *daśame ’hani (T 370; daśamevāhni T 403)
*nadyādipuṇyajalatīram (T 370; nadyādipuṇyatīrtheram T 403) āsādya prāṇāyāmaṃ
vidhāya *sapavitro (T 370; pavitro T 403) rudrabalinimittaṃ saṃkalpya *śuddhabhūtale (T
403; śuddhatale T 370) *sthaṇḍilopary (T 370; sthaṇḍilopari lakṣaṇam T 403) aṣṭadala-
padmam ālikhya tanmadhye kumbhaṃ saṃsthāpya tasmin, *oṃ hāṃ (T 403; om. T 370)
204 LIBERATING THE LIBERATED
rudrāsanāya namaḥ, oṃ hāṃ rudramūrtaye namaḥ, oṃ hāṃ rudrāya namaḥ. ity abhy-
arcya…
93
This list of eleven Rudras seems unique to the KKD. It differs quite substantially from
the lists of the eleven Rudras in other Śaiva and brahmanical sources I have examined. For
instance, note that Hara is the second Rudra in the list of transcript T 403, but that T 370
features Śivottama; while Hara is more commonly found in lists of eleven Rudras,
Śivottama is never found.
94
KKD (T 370, p. 235; T 403, p. 163): … tata *ekasmin (T 370; ekasminn eva T 403)
pātre, *oṃ pitṛbhyaḥ (T 370; śivapitṛbhya svadhā T 403) svadhā, oṃ pitāmahebhyaḥ
svadhā, oṃ prapitāmahebhyaḥ svadhā, *oṃ vṛddhaprapitāmahebhyaḥ svadhā (T 370; om.
T 403), oṃ mātṛbhyaḥ svadhā, oṃ pitāmahībhyaḥ svadhā, oṃ prapitāmahībhyaḥ svadhā,
oṃ *vṛddhaprapitāmahībhyaḥ (T 370; vṛddhapramātāmahibhyaḥ T 403) svadhā, oṃ
sarvebhyaḥ pitṛbhyaḥ svadhā, *oṃ sarvebhyo mātṛbhyaḥ svadhā (T 370; om. T 403). ity
evaṃ baliṃ *dattvā (T 370; dāpayitvā T 403), *tat pātraṃ tṛṇasahitenāṃguṣṭhenāpra-
dakṣiṇaṃ (T 370; tat pātraguṇitakuśasahitenāṃguṣṭhenāpradakṣiṇan T 403) triḥ
paribhrāmya *tasmin (T 370; om. T 403) *tṛṇaṃ (T 370; kuśau T 403) visṛjya tat pātraṃ
*svadakṣiṇapārśvam adhomukhaṃ (T 370; svadakṣiṇe pārśve ’dhomukhaṃ T 403) *vidhāya
(T 370; om. T 403) *vikṣipya (T 370; vinikṣipya T 403) tat pātram uddhṛtya *tad annam
ekaṃ (T 370; tadekaleśaṃ T 403) gṛhītvāghrāya pāṇiṃ prakṣālya kuṃbhagurvādīn pra-
dakṣiṇaṃ kṛtvā namaskṛtya guro rakṣāṃ labdhvā *kumbhakalaśasthān (em. Szanto;
kumbhakalasthān T 403; kumbhakalastha T 370) devān udvāsya…
ANCESTOR WORSHIP 205
seems to enjoin that the officiant should worship the stones, which have had
the deceased person invoked into them, offer another piṇḍa, and dismiss the
deities. Transcript T 403 appears to prescribe that the stones be worshipped a
second time, but then continues with a corrupt passage listing again the elev-
en Rudras and what might be instructions for performing the rudrabali in the
case of unfortunate deaths.
4.1.2.3. Ekoddiṣṭaśrāddha
95
KKD (T 370, p. 238; T 403, p. 168): evaṃ pratimāsaṃ tripakṣe ṣāṇmāsānte vatsarānte
ca ekoddiṣṭāni pañcadaśa kartavyāni… This list of fifteen śrāddhas is a little surprising, since
in standard śrāddha accounts, sixteen ekoddiṣṭaśrāddhas are prescribed.
96
KKD (T 370, p. 238; T 403, p. 168): maraṇāhāt samārabhya vatsarāntaṃ pratidinaṃ
*sodakakumbhāni bhojanāni (T 370; asenadakakuṃbhabhojanadīni T 403) dātavyāni,
vatsarānte sapiṇḍīkaraśrāddhaṃ kartavyam…
97
KKD (T 370, p. 238; T 403, p. 168): yad vā vatsarāntam ātmano *’vasthādyasaṃbhavād
(conj. Sanderson; sthāpanāyasaṃbhāvayan T 403; vasthādyāsaṃbhāvan T 370) dvādaśāhādiṣu
māsikaśrāddhasodakakumbhabhojanāni dattvā sapiṇḍīkaraṇaṃ kuryāt.
98
KKD (T 370, p. 235; T 403, p. 164): tatra pūrvedyur deśikaṃ *sādhakaṃ (T 403; sākaṃ
T 370) *putrakam (T 370; putram T 403) alābhe samayinaṃ *voktalakṣaṇaṃ (T 403; yathok-
talakṣaṇaṃ T 370)…
99
KKD (T 370, p. 235; T 403, p. 164): gṛhājire gomayopalipte dakṣiṇottaraṃ sthānaṃ
karāntaraṃ karamānaṃ *vṛttamaṇḍaladvayaṃ (T 370; vṛttamaṇḍala T 403) kṛtvā…
206 LIBERATING THE LIBERATED
statements addressing the deceased are in the singular; (e) and the invoca-
tion and offerings are made to the deceased using his initiation name in the
mantra.
At the end of the first ekoddiṣṭaśrāddha on the eleventh day, the
yajamāna may offer twenty-four presents to the invitee for the deceased.101
4.1.2.4. Sapiṇḍīkaraṇa
100
KKD (T 370 p. 237; T 403 p. 163): *bhūmiṣṭhasavyajānur dakṣiṇābhimukho (T 370;
bhūmiṣu savyañjanānuddakṣiṇābhimukho T 403) mūlayuktadīkṣānāmnā *svadhāntaṃ (T
403; om. T 370) piṇḍaṃ dattvā…
101
The list in the KKD of the twenty-four presents to be given to the invitee represent-
ing the deceased contains the following: (1) a purifier (pavitra), (2) a sacred thread, (3) a
water jar, (4) a rosary (akṣa- here for akṣasūtra), (5) sandals, (6) a staff, (7) a cloth worn to
conceal the private parts, (8) a seat, (9) a yoga band, (10) a parasol, (11) a turban, (12) an
upper garment, (13) pieces of cloth, (14) a golden ring, (15) a golden vessel, (16) clarified
butter, (17) sesame oil, (18) a cow, (19) land, (20) food, (21) a jacket, (22) a bed, (23) grain,
and (24) female and male slaves. KKD (T 370, pp. 237–238; T 403, p. 167): pavitraṃ yaj-
ñasūtraṃ ca kamaṇḍalv akṣapāduke | daṇḍa<ṃ> *kaupīnam āsanaṃ (conj.;
kaupīnabhasmāṅgaṃ T 370; kaupīnabhasmāga T 403) yogapaṭṭātapatrakam || uṣṇīṣam
uttarīyaṃ ca vastrahemāṅgulīyakam |*hemapātraṃ (T 370; hemapadmaṃ T 403) ghṛtaṃ
tailaṃ gau bhūmir bhojanāṃgakam || śayanīyaṃ *sarvadhānyāni (T 370; sarvadhānādi T
403) dāsīdāsopahārakā | caturviṃśati dānaṃ syād *ekoddiṣṭe (T 403; ekoddiṣṭeṣu T 370)
saṃmatam. The conjecture to emend to kaupīnam āsanaṃ is made on the basis of a passage
in the SP (1. 57–59b), which gives a list of offerings to be presented to Īśāna during the rite
of pavitrarohaṇa; this contains a similar list of utensils to be given to an ascetic:
daṇḍāksasūtrakaupīnabhikṣāpātrāṇi rūpiṇā | kajjalaṃ kuṅkumaṃ tailaṃ śalākāṃ keśa-
śodhanam || tāmbūlaṃ darpaṇaṃ dadyād uttare rocanām api | āsanaṃ pāduke pātraṃ
yogapaṭṭātapatrakam || aiśānyām īśamantreṇa dadyād īśānatuṣṭaye. This part of the text has
also been incorporated into the Agnipurāṇa; another parallel is found in the KKD.
The practice of offering gifts to Brahmins on behalf of the deceased mirrors brahmanical
practice. Cf. Kane 1953, pp. 534–535; and Müller 1992, p. 151.
ANCESTOR WORSHIP 207
102
KKD (T 370, p. 239; T 403, p. 168): tatra pitrarthaṃ *deśikān (T 370; deśikāt T 403)
viśvadevārthaṃ sādhakau tadabhāve *dīkṣitān samayasthān (conj.; dīkṣitān samassthān T 370;
adīkṣitān samayasthān T 403) vā gṛhṇīyāt […] nimittāya *yaddeśikādiṣu (T 403; devaśikādiṣv
T 370) ekaṃ gṛhṇīyāt.
103
After the section prescribing the ceremonial reception of the ancestors, the KKD en-
joins the following (T 370, p. 240; T 403, p. 170): nimittam api tathaiva pādaṃ prakṣālyāca-
mya…
104
KKD (T 370, p. 240; T 403, p. 170): nimittam *atithyabhyāgatau (em.; atithyabhāgatau T
370; atithyatyagatau T 403) ca pūrvānanān *āsaneṣu (T 370; vāsaneṣu T 403)
vāmadevenopaveśya…
105
KKD (T 370, p. 240; T 403, p. 170): nimittārghyaṃ pitrarghyavat kṛtvā…
106
KKD (T 370, p. 240; T 403, p. 170): sapiṇḍīkaraṇaśrāddhaṃ kariṣya iti prārthya ku-
ruṣveti tair anumatas…
107
Hence, after worshipping the ancestors following their invocation, the KKD (T 370,
p. 241; T 403, p. 171) teaches that the recently deceased receives the same kind of worship:
evaṃ nimittam api saṃpūjya…
108
KKD (T 370, p. 242–243; T 403, p. 173): pitrarthaṃ* piṇḍatrayaṃ (T 370; traya T
403) *mātṛpakṣasyāpi (T 370; mātravarasyāpi T 403) trayaṃ nimittāyaikaṃ dattvā … That
the procedures for all the aforementioned rites are the same as those of the pārvaṇaśrāddha
is inferred from the absence of any other specifications.
208 LIBERATING THE LIBERATED
109
KKD (T 370, p. 243; T 403, p. 174): *tataś (T 370; tatra T 403) *cocchiṣṭam (T 370;
taccocchiṣṭaṃ T 403) *uddhṛtya (T 370; om. T 403) saṃmārjya *saṃprokṣya (T 370; om. T
403) *pādau (T 403; pābhyā T 370) prakṣālyācamya *sakalīkṛtya (T 370; sakalīkṛtvā T
403) *sāmānyārghyaṃ kṛtvā (T 403; om. T 370) *gomayopalipte (T 403; gomayopalipta T
370) maṇḍale pātracatuṣṭayaṃ *viśuddhaṃ (T 370; śuddhaṃ T 403) *vinyasya (T 370;
vinyaset T 403) *teṣāṃ (T 370; teṣu T 403) gandhodakṣīratiladarbhāgragandhapuṣpāṇi
*kṣipya (T 403; vikṣipya T 370) īśaṃ sadāśivaṃ śāntaṃ *caturthe (T 370; om. T 403)
*mūlamantrayuktadīkṣānāmnā yuktaṃ (conj.; mūlamantrayuktadīkṣānāṃ ayukta T 370; +
layuktadīkṣānāmnā yuktaṃ T 403) nimittaṃ cottarādidakṣiṇāntaṃ *saṃpūjyārghyadhūpādi
(T 370; saṃpūjyārghyadhūpadīpādi T 403) dattvā nimittapātrajalaṃ mūleneśādipātreṣu
yojayet. tatas tatpātrāṇi *visṛjyācāmet (T 370; visṛjyācamya T 403).
110
For the term rudrāṃśa denoting lay devotees of Śiva in this context, see Sanderson
2003/2004, pp. 354–355, fn. 16.
111
See pp.72–73.
ANCESTOR WORSHIP 209
ple, the ancestors are identified, in ascending order, with Gaṇeśa (the chief
of Śiva’s attendants), Skanda (the son of Śiva) and Rudra (a manifestation
of Śiva belonging to a lower level in the Śaiva tantric framework), collec-
tively a group of deities representing the key figures of popular lay Śaiva
mythology as one finds in the purāṇa genre. The Viśvedevas, in turn, are
replaced with Caṇḍa and Mahākāla, the two attendants of Śiva.112 The ben-
efit gained from performing the rudraśrāddha for the yajamāna who com-
missioned the rite is declared to be attaining union with Rudra,113 thus a
spiritual level corresponding to the level of lay Śaivism. In the case of the
laukikaśrāddha, during the ritual the ancestors are equated with Brahmā,
Viṣṇu and Rudra; the Viśvedevas, with Sūrya and Yama; 114 and the
yajamāna is declared to attain the world of Brahmā as a reward.115
112
Kir 61.7d–8. For the text and translation of the passage, see the appendices. Caṇḍa
and Mahākāla, besides being names of attendants of Śiva, are also two names for Śiva him-
self when he appears in his wrathful manifestation. However, given the current context, it is
more likely that the former is intended.
113
Kir 61.34b–35a. For the text and translation of the passage, see the appendices.
114
Kir 61.10. For the text and translation of the passage, see the appendices.
115
Kir 61.36. For the text and translation of the passage, see the appendices.
116
In the JR we may also have an indication that tantric priests could use the manual’s
instructions to perform post-mortem rites for a clientele beyond the initiate community. In
one half-verse (JR ŚP 42ab), this text appears to teach that the śrāddha rituals as outlined
can also be performed for those who wish to attain a different heaven or spiritual level; in
this case, the śrāddha rites are simply performed using the names of whichever deities the
client wishes. However, this interpretation must be considered with caution since the reading
and interpretation are not clear. Further, there is no other passage to this effect in the JR.
Nevertheless, given that we find such possibilities in both the SP and KKD, it would not be
surprising to find such an option also included in the JR, even if added by later redactors.
117
SP 11.3b–6b: īśaḥ sadāśivaḥ śāntaḥ śivaśrāddhe tapasvinaḥ || rudraś cānantanāmā
ca viśvedevau sthitau dvayoḥ | dīkṣitānām amī devā rudrāṃśānām atho ’nyathā || tatra
nandimahākālau viśvedevau gaṇau dvayoḥ | skandacaṇḍagaṇādhīśā viśvedevau yamāruṇau.
210 LIBERATING THE LIBERATED
However, by the time of the SP’s composition, there seems to have been
some ambiguity about who these rudrāṃśas were, a complication also re-
flected in today’s scholarly interpretation of the term. It has been demon-
strated that at the time of the Kir’s composition, rudrāṃśa referred to Śaiva
lay devotees.118 Regarding the use of the term in the SP, however, various
scholars have interpreted it as meaning “a part (aṃśa) of Rudra” and thus
referring to lower level initiates, that is, samayins.119 Indeed, it is not sur-
prising if the meaning of a technical term, especially one designating classi-
fications within the community of adherents, changed over time based on
the structures envisaged by the respective authors. With the material we
have at hand, however, it is not possible to determine with any certainty
how this term was understood at the time of Somaśambhu, the author of the
SP.
A slightly clearer picture of the intended clients emerges in Aghoraśiva’s
manual KKD. Adapting the three-level śrāddha model, in addition to the
śivaśrāddha, he also teaches two more modes. They too follow the same
procedure, but modify the deities equated with the ancestors and Viśvedevas.
Here, the second level of śrāddha corresponds to the rudraśrāddha, but uses
a slightly different set of Śaiva-related purāṇic deities: used for the ancestral
deities are Skanda, Caṇḍa and Gaṇādhīśa (= Gaṇeśa) – none actually a mani-
festation of Śiva, but rather his son and attendants – and for the Viśvedevas,
Nandin and Mahākāla, further attendants of Śiva. However, unlike the Kir
and the SP, this level of śrāddha is explicitly linked to a deceased samayin.120
By extension it is perhaps also intended for recipients of a viśeṣāntyeṣṭi, those
who had received a slightly higher level of initiation during their lifetime but
not yet ultimate liberation. As in the case of the laukikaśrāddha above, the
third level of śrāddha in the KKD is explicitly linked to orthodox brahmani-
118
Sanderson 2003/2004, pp. 354–355, fn. 16.
119
Brunner-Lachaux 1977, p. 630, fn. 21. This interpretation is partly based on Nirmala-
maṇi’s commentary on the KKD.
120
KKD (T 370, p. 244; T 403, p. 174): *samayinaḥ (em.; samayinaṃ T 370; mamayina
T 403) *skandacaṇḍagaṇādhīśāḥ (em.; skandañ caṇḍa° T 403; kaṣṭaś caṇḍa° T 370)
*pitṛpitāmahaprapitāmahāḥ (T 370; om. T 403) pitaro nandimahākālau *viśvedevau (conj.;
viśvadevau codd.) laukike *brahmaviṣṇvīśā (em.; brahmaviṣṇvīśa T 370; brahmāviṣṇuvīśāḥ
T 403) *viśvedevau (conj.; viśvadevau codd.) yamavaruṇau…
ANCESTOR WORSHIP 211
cal practitioners, with Brahmā, Viṣṇu and Īśa as the ancestral deities, and
Yama and Varuṇa as the Viśvedevas.121
Śaiva lay devotees are not mentioned separately here. There appear to
be two possible explanations for this. First, it might be speculated that in
the society being described by Aghoraśiva, lay Śaivas were offered the
same type of post-mortem ancestor worship as orthodox Brahmins. Second-
ly, it is also plausible that lower-level initiates and lay devotees had simply
collapsed into a single group by this point in time. After all, performing a
full śivaśrāddha for a samayin and invoking superior forms of Śiva mani-
festations would have been problematic, and it is more likely that such a
practitioner was assigned to a lower register. While this problem is not
mentioned directly in the Kir or SP, in the KKD Aghoraśiva does not shy
away from addressing these points. In fact, rites for samayins were clearly
an important issue for him, perhaps suggesting that this lower-level initia-
tion was quite common amongst the population at the time and that the
boundaries between these and ordinary lay devotees had become blurred.
With this in mind, one might also consider whether a similar phenomenon
was in fact also the case for the Kir and the SP, since neither of them speci-
fies what kind of śrāddha a samayin is to receive. This omission is not sur-
prising in the context of the Kir, which does not make any such distinction in
its discussion of cremation either. But the SP is concerned about preserving a
distinction between putrakas, who are to receive a full initiatory cremation,
and samayins, who receive a simple cremation without initiation of the de-
ceased person’s soul.122 It would thus not be surprising if different modes of
śrāddha rituals were intended for these two groups. If so, also here the group
of samayins and Śaiva lay devotees has been collapsed into a single group
and spiritual level for the śrāddha rites. For lack of more conclusive evi-
dence, however, these considerations remain speculative.
121
Ibid.
122
See p. 97.
212 LIBERATING THE LIBERATED
123
Sanderson 1995.
124
In this sense, in the SP it is declared that the rite also benefits the yajamāna and his
family; SP 11.36: śivaśrāddhe kṛte devā †manuṣyāḥ† pitaro ’surāḥ | tṛptāḥ śivānubhāvena
vardhayanti kuladvayam. This verse is also quoted in the KKD (T 370, p. 245; T 403, pp.
175–176). Here it seems that pāda b is corrupt, since living people (manuṣya) are not satiat-
ed during the ritual; the mention of Asuras in the śivaśrāddha is also puzzling. However,
this corruption may have arisen through confusion concerning the brahmanical formulation,
which clearly served as the model. There, the pitṛs (ancestors) are equated with the sons of
the group of śrāddha deities, namely the Vasus, Rudras and Āditis, and are opposed to the
pitṛs of humans. See also YājS 1. 269–270 (which is incorporated into the Agnipurāṇa
[2.163.40c–42b] and the Nāradamahāpurāṇa [51.152c–154b]): vasurudrāditisutāḥ pitaraḥ
śrāddhadevatāḥ | prīṇayanti manuṣyāṇāṃ pitṝn śrāddhena tarpitāḥ || āyuḥ prajāṃ dhanaṃ
vidyāṃ svargaṃ mokṣaṃ sukhāni ca | prayacchanti tathā rājyaṃ prītā nṛṇāṃ pitāmahāḥ.
“The śrāddha deities, which are the ancestors and sons of the Vasus, Rudras and Āditis,
pleased with the śrāddha offerings, please the ancestors of men. [And] the [individual]
ancestors (pitāmahāḥ), who are gratified, bestow long life, offspring, wealth, knowledge,
heaven, liberation, pleasures and royalty on men.” In his commentary on these verses,
Vijñāneśvara interprets that the divine ancestors and the directly named human ancestors are
to be treated as different categories throughout the rite, but that they are nevertheless part of
the same entity.
Kir 61.38 also declares the great benefit of performing a śrāddha. This is simply the re-
sult of performing the ritual well. This is achieved despite the fact that a human can do
nothing to benefit supreme Śiva: śivasya paramātmānaḥ kin nāma kriyate naraiḥ | kriyā
sampūrṇatām eti tad uddiśya phalaṃ mahat. For the apparatus and translation, see the ap-
pendices.
214 LIBERATING THE LIBERATED
tory practices. Here, the injunctions were reformulated to define the śrād-
dha as the debt to the “gurus, etc.” (gurvādi), rather than to the ancestors
(pitṛ).125 In the initiatory context, this could mean that the family lineage
has been replaced by the spiritual lineage, and thus performing a Śaiva
tantric śrāddha simply represents an additional ritual obligation for the
Śaiva ritualist: if a householder, he would perform a śivaśrāddha for his
spiritual lineage and an ordinary śrāddha for his familial context. The
forms of rudraśrāddha and laukikaśrāddha would then simply be addition-
al services that Śaiva tantric śrāddha priests could offer, thereby reaching a
wider audience.
It may also be considered that for the śivaśrāddha, which is after all
primarily taught, the choice is not problematic if both the deceased and the
yajamāna are initiates, since the former is then worshipped in his supreme
divine state suiting his initiatory status and the latter is performing his post-
initiatory duties, the accomplishment of which is believed to bring about
his ultimate liberation at death. In this light, it is tempting to consider the
hypothesis that in some areas, only the head of a joint household underwent
initiation.126 If this theory is correct, this would mean that an elder son
would be initiated before his father’s death. This, in turn, would qualify
him to undertake the performance of a Śaiva śrāddha. In fact, particularly
the JR seems to support this model, since it also explains how to coordinate
Śaiva postmortem rites with mundane practice.127 This would make sense if
we consider that some members of a household headed by a Śaiva initiate
would not necessarily be initiated and thus would follow mundane practices
in terms of post-mortem offerings and impurity regulations on the occasion
of death.
6. Conclusion
These considerations raise more questions than they answer. Nonetheless,
the texts do reflect the complex situation in which Śaiva ritualists found
themselves as they expanded and adapted their ritual repertoire to include
125
See p. 192.
126
See p. 17.
127
See p. 171.
ANCESTOR WORSHIP 215
more communities. After all, especially in the sphere of śrāddha rituals, the
Śaivas tantric priests were encroaching upon a major ritual area in the
brahmanical world, a ritual cycle deeply rooted in its socio-religious struc-
tures and intrinsically connected with familial values and the image of an-
cestral deities. This involved notions that were difficult to project onto a
world focused on the spiritual liberation of the individual. Especially at this
level of mainstream practice, it is likely that how Śaiva rituals were inte-
grated varied widely in response to well-established practices in different
geographical and societal contexts. The doctrinal repercussions seem to
have been too complex to have been completely resolved and rationalized
in Śaiva terms. While not enough material is available to make more defi-
nite assessments about actual practices, it is nevertheless clear that the
sources prescribing śrāddha rites transcend the closed world of the initiated
elite. Indeed, this development, as well as the fact that various types of
Śaiva cremations were firmly established by this time, points to a profes-
sionalized group of tantric funerary and śrāddha priests having emerged.
While this group was likely to have been perceived as unwanted competi-
tion by its brahmanical counterparts, its existence is testimony to the suc-
cess of Śaiva tantric communities in gaining firm footing in brahmanical
mainstream society.