an outline of brahmanism ca. 900 c.e

7/25/2019 An Outline of Brahmanism CA. 900 C.E. .

1/46

AO B

(. ..)Lectures delivered in All Souls College in Michaelmas Term

Alexis Sanderson,

Not to be quoted or copied without the permission of the author

roughout the first milleniumall religion in India was comprisedwithin Buddhi-sm, Jainism or Brahmanism (commonly called Hinduism).is tripartition is so wellknown that one can hardly begin an account of Indian religion without mentioningit. But further analysis must begin with the observation that it is no part of the cultureitself. All indigenous listings of the forms of religion include Buddhism and Jainism,

but none embraces the remainder under a single heading. In place of Brahmanismor Hinduism we never find less than three distinct entities: Vedism, Vai .s .navism, andaivism. Non-Buddhist, non-Jain participants in the culture did not define themselvesas Brahmanists or Hindus, but as Vaidikas, Vai.s .navas oraivas, meaning followers ofthe Veda, followers of the teachings of Vi.s .nu, and followers of the teachings ofivarespectively.us each of these three groups was identified by the fact that it has itsown corpus of sacred texts, just like the Buddhists and Jains, whose Sanskrit namesBauddha/Saugata and Jaina, mean those who follow the teachings of the Buddha andthose who follow the teachings of the Jina (Mahvra).e Veda which defines theVaidikas religion comprises the ruti, the direct testimony, namely the four Vedas (invarious recensions, each with Sa.mhit, Brhma.na and Upani.sads) and the Sm.rti, theindirect testimony, namely later scriptural works approved as reliable.e primary

works in this category are the Treatises on Religious Duty (the Dharmastras, alsocalled Sm.rtis). Apart from these are the History (itihsa.h [the Mahbhratam]), whichcontains the Bhagavadgtand other important religious materials, and the Pur.nas, avast and then ever-growing corpus of works covering the same ground and authorisingvarious devotional forms of religion. In the case of the Vai.s .navas andaivas it wascollections of Tantric works known variously as Tantras,gamas or Sa.mhits, whichthey attributed to their supreme deity, Vi.s .nu oriva. Of these I shall say more in duecourse.

It might be argued, therefore, that the category of Brahmanism or Hinduismshould be eliminated from our thought about pre-modern India. However, there are

affinities between Vedism, Vai.s .navism andaivism which, though not sufficient tocreate the sense of a shared religious identity in the face of Buddhism and Jainism,

justify us in grouping these three traditions under a single term in certain contexts.e Vai.s .navas andaivas adhered to texts which the Vaidikas rejected as outside theVeda (vedabhya.h), and in this respect they were as alien to them as the Buddhists


2/46

and Jains. Indeed the seventh-century Vaidika scholar Kumrila considered that the

aivas observances meant that they were more un-Vaidika than those groups. Sothere certainly was nothing corresponding to a larger Hinduism in the Vaidika pointof view. But the Vai.s .navas andaivas themselves sought to co-ordinate their religion

with the domain of Vaidika observance, a rapprochement never attempted by theBuddhists or the Jains. is they did by defining the Veda neither as invalid (inthe manner of the Buddhists and Jains) nor as the only valid source for knowledgeof religious duty (in the manner of the Vaidikas themselves), but as valid in its owndomain as the authority governing what they saw as the exoteric infrastructure of theirown esoteric forms of religion. One passed through the domain of Vaidika observancebefore one ascended through the ritual of initiation (dk.s) to become a Vai.s .nava oraaiva; and althoughaiva ascetics (yati .h, nai.s.thika.h), or at least certain classes of

them, were required thereafter to abandon all aspects of the religion from which theyhad entered, so that in effect they were as un-Vaidika as those who converted toBuddhism or Jainism, the majority ofaivas, particularly those who were not asceticsbut married men (g.rhastha.h), and almost all Vai.s .navas, whatever their mode of life,

were required to integrate key elements of Vaidika practice into the pattern of theirspecial observances and to do everything they could to embed themselves withinbrahminical society.

is affinity between Vedism, Vai.s .navism andaivism is also manifested throughrapprochement in the other direction, from the Vaidikas to the Vai.s .navas oraivas.For there were groups considering themselves to be Vaidika, who, while refusing theinitiation that would have put them beyond the Veda by making them Vai .s .navas

or aivas in the strong sense, incorporated Vaidika versions of Vai.s .nava or aivaworship into their regular observances and created the scriptural justification for thisdevelopment within the corpus of the Pur.nas.

ere are, then, sufficient grounds for continuing to accept the tripartition ofIndian religion into Buddhism, Jainism and a single category embracing Vedism,Vai.s .navism andaivism. However, this is not a classification which captures anythingof the nature of its elements. For that purpose the relevant distinctions are rather thefollowing:

. between Vedism and the rest,

. between (a) Vedism together with the Veda-integrating systems of the Vai.s .navas

and theaivas, and (b) the anti-Vaidika systems of the Buddhists and Jains,

. between social religion, found only in Vaidika observance, and personal re-ligion, both Vaidika and non-Vaidika; within social and personal religion


3/46

. between obligatory observances (nityakarma) and optional desiderative obser-

vances (kmyakarma),. between the classes of those qualified to participate, in terms of caste-status

(var .na.h, jti .h) and gender; and, within personal religion,

. between the religion of those in the world and the religion of anchorites,

. between religion for reward (bhoga.h) and religion for liberation (mok.sa.h),

. between the religion of men and the religion of women, and

. between esoteric, non-dualistic systems of observance (advaitcra.h), whichrequire transcendence of the social religions dualities (the pure and the impure,

the permitted and the forbidden), and exoteric, dualistic systems (dvaitcra.h),which do not.

e second of these five distinctions has been sufficiently clarified. efirstandthethird are related, because what set Vedism offfrom all the other systems was preciselythat it alone embraced both social and personal forms of religion. By social religionI mean primarily the life-cycle rites and other observances followingruti, Sm.r.ti andcustom (cra.h) which are performed by persons, or for persons, in their identity asmembers of a caste (jti .h) and household (g.rham); and secondarily the observanceof rules of endogamy, commensality, hierarchical interaction, avoidance of contact,and contraction of qualification to participate in the religion (adhikrasa.mkoca.h),

in accordance with notions of sanctity, ritual purity, and pollution, also the rulesgoverning the inheritance of property (dyavibhga.h), the rest of civil and criminallaw (vyavahra.h), and the institution of kingship (rjadharma.h), all of these beingauthorised by Sm.rti, or, as with the rules governing the interaction of castes (jti .h) andsubcastes in a particular region, by custom within the bounds of the rules laid down inSm.rti for the caste-classes (var .na.h). All this was seen to lie under the authority of theVeda, and that of the Veda alone. By personal religion I mean disciplines undertakenby persons believing that they are acting as individuals without, or without referenceto, any social identity. Vedism contains religion that is personal in this sense, and thenon-Vaidika systems in their classical forms saw themselves as comprising nothingelse.

Vai.s .navas,aivas, Buddhists, and Jains looked upon Vedisms social religion as aset of mundane observances (lokcra.h) or local customs (decra.h) irrelevant to theattainment of what they saw as the goals of religion proper. With the exception offunerary and post-funerary ritespeople who had left the mundane religion couldnot be returned to it in deaththey developed no social religion of their own. e


4/46

Vai.s .navas andaivas had Vedism for this purpose, while Buddhism and Jainism had

no need at all for social religion so long as they were the religions of celibate asceticsoutside society. When Buddhists and Jains did become fully socialised, to the extentthat they needed their own priests and life-cycle rites, as happened with the Digambara

Jains from the ninth century, with thevetmbara Jains from the fifteenth, and withthe Newar Buddhists of the Kathmandu valley from about the same period, theysimply took over the Vaidika rites and substituted non-Vaidika mantras.

As for the law, that always remained Vaidika, being administered by kings or theirrepresentatives with the help of brahmin experts who were to follow the ordinances ofthe Dharmastras taking into consideration any regional practices and conventionsof castes, religious orders, guilds and the like that did not contradict those ordinances.None of the non-Vaidika traditions aspired to produce anything in its place. It was

enough that they should be sufficiently influential in society and the court to renderinoperative the one Vaidika injunction that could disadvantage them, namely thestatement of the Dharmastra of Manu that kings should expel all non-Vaidikas(p.sa.n .dina.h) from their kingdoms.is rule was invoked at certain times, but onlyin the case of very extreme sects. While the strict religious view was that the onlyvalid precepts were those that were rooted in the Veda (vedamla.h), that drew theirauthority from that source, the socio-legal view was that a religion was valid if it waslong established and did not threaten the Vaidika order of castes (var .nadharma.h) andstages of life (ramadharma.h).

Let us now turn to Vedism in detail, and first to the Vedism of the vast majority,that is to say, the religion of Vaidikas living in the social world rather than that of

those outside it as ascetic renouncers. I shall outline what this comprised and thenconsider the question of who was qualified and required to observe it, and, in the lightof my answer to that question, assess the distribution of the social and the personalaspects of religion.

Within this segment of Indian religion were two distinct strata of practice, therauta and the Smrta, namely that which is taught in theruti and that which istaught in the Sm.rti.e first was an optional domain of special intensity open tothose already in the second.ose who entered it had thenceforth to maintain bothforms of the religion concurrently.e majority, perhaps the overwhelming majorityin our period, chose to remain with the Smrta alone.

rauta Vedism comprised nothing but a cycle of regular sacrifices into fire, in

which the off

eringswere the products of cultivation (havi .h: milk products, particularlyclarified butter, various grains, rice-beer, and various parts of goat or sheep), or, in theclimactic ritual of the cycle, the juice of the Soma plant (probably Ephedra originally,but in our period various substitutes). After the Agnydheya ritual, in which a man setup the three fires required forrauta sacrifice, he was bound to institute the following


5/46

recurrent sacrifices for the rest of his life, or until he renounced the world:

. the Agnihotra: a brief twice-daily offering of milk in the morning to Sryaand Prajpati and in the evening to Agni and Prajpati, an offering which, inthe case of a brahmin, could be done by the sacrificer himself and had to bedone by him on the days of the new and full moon, but was otherwise doneby a single officiant (.rtvik) in the presence of the sacrificer (yajamna.h) or his(principal) wife, his consort in religion (dharmapatn); then more complicatedrituals performed by a number of officiants in the presence of the sacrificer andthis wife, namely

. the Dare.s.ti and the Pr .namse.s.ti requiring four officiants: monthly offeringsof rice-cakes and butter on the day after the new moon day (dare.s.ti .h) to Agni

and the dual deity Indra-and-Agni, and on the day after the full moon day(pr .namse.s.ti .h) to Agni and Agni-and-Soma, with offerings to the sacrificerspaternal ancestors (pi .n .dapit.ryaja.h) on the new moon day itself,

. the graya.ne.s.ti, requiring four officiants: offerings of the first fruits of theharvest, of paddy in autumn and of barley in spring, to Indra-Agni, the All-Gods (Vive Dev.h) and Heaven-and-Earth (Dyvp.rthiv),

. the three seasonal Cturmsyas, requiring five officiants, which were performedat four-month intervals, the Vaivadeva in spring, the Varu .napraghsa in therains, and thekamedha in autumn, the last being concluded with offeringsto the sacrificers paternal ancestors (pit

.ryaja

.h),

. the Paubandha: the yearly or twice-yearly sacrifice of a goat, and finally

. the climactic Soma sacrifice, principally the Agni.s.toma form, requiring se-venteen officiants and occupying several days in spring each year, with () aSautrma.ni sacrifice to be performed with six officiants within the followingtwelve-month, in which rice-beer (sur), milk and animal sacrifices were offeredto the Avins, Sarasvat and Indra. e Soma sacrifice might be intensifiedby an Agnicayana: throughout the preceding year a huge and complex altar(agni .h) of bricks was built and consecrated to receive the sacrificial fire on itscentre before the Soma ceremonies began. Just as the Agnicayana is the highest

form ofrauta sacrifice, so its altar is the principal focus of metaphysical andsoteriological doctrine in therutis exegesis of its rites.

To be qualified to set up the three rauta fires and so to enter this cycle of sacrificesa man had already to have been practising the rituals laid down in the Smrta domain


6/46

of Vedism. He had to have set up the single Smrta or domestic fire (smrtgni .h,

g.rhygni .h) within his home and to have been offering the Smrta sacrifices in it.ese were much less elaborate affairs than therauta sacrifices and were done eitherby the sacrificer himself or with the help of two officiants, one (the kart) who dideverything for his client (yajamna.h) except certain crucial actions, and the other asilent supervisor (the brahm), whose function was to intervene with instructions ifnecessary.e rituals of Smrta Vedism were the following:

. a series of sacrifices running parallel to therauta programme up to and inclu-ding thegraya.nas and consisting of offerings in the domestic fire to the deitiesof those sacrifices; four seasonal sacrifices in the domestic fire on full-moon days(rva.n,vayuj,grahya.n and Phlgu .n)this is roughly parallel to therauta Cturmsyas; and similar sacrifices in connection with such events inthe agricultural calendar as ploughing, sowing seed and harvesting;

. a large number of non-calendrical rituals performed if the occasion for themshould arise, such as sacrifices to ward offplanetary influences (grahayaja.h,nak.satrayaja.h) and other dangers (ntihoma.h), or to sanctify constructions(prati.s.thhoma.h) such as houses, wells or bridges;

. regular and calendricalrddha rituals for honouring the patrilineal ancestors;

. the performance of the life-cycle rites (sa.mskra.h) for ones children, namely(in the procedure of the K.thakag.rhyastram, which was followed in Kashmir)

(a) the rite to ensure conception (garbhdhnam, bjavpanam);(b) the parting of the pregnant wifes hair (smantonnayanam) in the third

month after conception, believed to purify the womb for the benefit ofthe embryo;

(c) the rite to make the embryo male (pu.msavanam) in the sixth month orthe eighth;

(d) the rite to assist the mother to give birth when she is near her time(so.syantkarma);

(e) the sacrifice at birth (jtakarma),

(f) the name-giving (nmakara.nam) on the eleventh day after birth;(g) the rites of taking the infant outside the house to see the sun and the moon

(ni.skrama.nam: sryadaranam, candradaranam) in the third month;

(h) the rite of giving the first solid food (annapranam) in the sixth month;


7/46

(i) the tonsure (c .dkarma) towards the end of the third year;

(j) the investiture (upanayanam) that initiates a boys period of celibatestudentship in the Vedas (in any year from the seventh to the sixteenthyear for a brahmin, from the ninth to the twenty-second for a K.satriya,and from the eleventh to the twenty-fourth for a Vaiya);

(k) the five sacrifices that initiate the observances (vratam) adopted for thedifferent stages of the students study of the Veda (a. traividyakavratam; b.cturhot.rkavratam; c. pravargyavratam; d.ru .navratam; and e. aupani.sa-davratam);

(l) the ceremony of bathing and tonsure (godnam, kenta.h) that frees thestudent and enables him to marry;

(m) marriage itself (vivha.h), preceded in the case of a daughter by (a) ahair-parting rite (sa.toddhara.nam [=smantonnayanam]) before she is pro-mised, (b) a sacrifice to Indr.non the day that she is promised, and (c)a version of the celibacy-ending ceremony (godnam) but without theshaving of the hair; and, finally, (xix) the cremation of ones parents andothers in ones household (antye.s.ti .h);

. the Vaivadeva sacrifice, a daily or twice-daily ritual preceding the family meals,in which part of the prepared food is sacrificed in the fire to variousrauta godsincluding the All-Gods (vive dev.h)hence the name of the sacrifice andother portions are offered outside the fire first to the Beings (bhtni), namely

various minor deities around the hearth, in the kitchen implements, and in theparts of the house, and then to the ancestors (paternal and maternal). Finallyfood must be given to any uninvited guests (atithaya.h) or mendicant ascetics(bhik.sava.h);

. the daily recitation of some part of the Veda.is, the Vaivadeva sacrifice, andthe ancillary feeding of guests comprise the five Great Offerings (mahyaj.h):to the Beings (bhtayaja.h), the ancestors (pit.ryaja.h), the gods (devayaja.h),men (manu.syayaja.h), and the Veda (brahmayaja.h).e Vaivadeva consistsof the first three; and the fourth is the feeding of guests and ascetics that followsit.e fifth, the offering to the Veda, is this daily recitation of parts of theruti;

. sa.mdhyopsanam: the veneration (upsanam) of the sun at the three junctures(sa.mdhy) of the day preceded by a lustral immersion (snnam): the casting upof water to the sun (sryopasthnam) at dawn, noon and sunset accompaniedand followed by the recitation (japa.h) of the Gyatrmantra;


8/46

. tarpa.nam: the pouring of libations of water to the gods (dev.h), sages (.r.saya.h)

and ancestors (pitara.h);. devapj: the daily devotional worship (pj) of non-rauta Gods, principally

Vi.s .nu,iva, his consort Gaur, iva and Gaur as one (Gaurvara, Ardha-nrvara), the goddess Durg, Brahm, or Srya (the Sun).e worship couldbe carried out at a temple or, preferably, in a private shrine within the home,

where the deities were summoned into idols kept there for this purpose. Ifthere were no idols the deities could be summoned into a vase (kalaa.h) filled

with scented water, or on to a throne-diagram (ma.n .dalam) drawn on a raisedplatform (vedi .h) or the ground. If a man had neither idols nor shrine, he could

worship on the bank of a river after the purifying immersion (snnam) thatalways precedes this rite. In outline the Pjwas as follows. e worshippersummoned the deity (vhanam) by reciting verses that describe its iconic formand request it to be present.en, treating it as though it were an honouredguest that has just entered his home, he presented it with water for its feet(pdyam), water to drink (arghyamm), and water for purification (camanam),

washed it with various substances (snnam), collected the run-off(the waterfrom its feet [cara.nodakam]), anointed the idol (anulepanam), chanted a hymnof praise (stotram), gave it fresh garments and a sacred thread (upavtam),daubed it with a fragrant powder (gandha.h), gave it flowers, burning incenseand a lighted lamp, fanned it with a chowrie (cmara.h) while chanting a hymn,gave it an umbrella to shade it (chatram), held up a mirror (dara.h) to itsface, offered it a dish of sweetened yoghurt (madhuparka.h) and a handfu-ll of flowers (pu.spjali .h), circumambulated it, prostrated himself before it(a.s.tngapra.nma.h), gave the deity leave to depart (visarjanam), and then drankthe water from its feet.

. the same form of worship offered to the tutelary deity of the patrilineage(kuladevat), most commonly a localised form of the martial Goddess Durgor Ca.n .d.is deity could be included in the daily devapj just described;but the general practice was to worship it (oneself or through an officiant) onthe day of its annual festival, on the occasion of major life-cycle rites such asmarriage, and for personal favours such as the birth of a son or the averting ofdanger. Kings and others with the necessary resources might build their own

private temples for their lineage deities. However, for those with the appropriateTantric initiation (dk.s) the worship of the tutelary goddess would be Tantricrather than Vaidika.

e outline of the daily schedule of an observant Vaidika was as follows. Wakingbefore dawn he was to meditate on his chosen deity or deities (prta.hsmara.nam),


9/46

dress, leave his house to attend to the needs of nature (aucam), and then go to

a river, the water-tank of a temple (devakhtam), or a lake. Having tied up histopknot (ikhbandha.h), he was to squat with his arms between his knees, and holdingblades of darbha grass in each hand he was to drink water thrice from his right hand(camanam), wipe his mouth twice, touch his mouth nostrils, eyes, ears, navel, heart,head and arms with water (upasparanam), clean his teeth with a twig, and batheby immersing himself fully (snnam) while meditating on the mantra known as theEffacer of Sins (aghamar.sa.nasktam). Still standing in the water, he was to throw

water from his cupped hands into the air three times, first to the gods (facing east),then to the sages (facing north), and finally to the ancestors (facing south), come upout of the water, and wring out his garment on the bank (vastrap .danam) as a libationto those of his clan (gotram) without sons to do therddhas for them. He was then

to dress and do the dawn sandhyopsanam: having purified himself internally bydoing the breathing exercise (pr.nyma.h) while meditating on the Gyatrmantra,he was to sprinkle water on his feet, head and shoulders, drink water from his handagain, breath out his sins into a handful of water (ppapuru.sanirasanam), cast itfrom him, take water in his cupped hands, throw it up towards approaching sun(sryopasthnam), and then silently repeat the Gyatrmantra until the sun was rising(gyatrjapa.h). He was then to return home to offer the dawn sacrifice to his firesand/or fire (prtarhoma.h). During the morning he was to study or teach the Veda(the first and optional brahmayaja.h) and then attend to worldly matters.

Towards noon he was to repeat the sandhyopsanam with its preliminary bath,repeat portions of the Veda (brahmayaja.h), pour libations (tarpa.nam) from his

fingers to the gods, the ancestors and the sages, wring out his wet garment, thenreturn home to worship his chosen deity or deities in his shrine or do it there by theriver bank. When his devapjwas completed and his wife had announced that thefood for the midday meal was ready he was to offer the Vaivadeva, thus veneratingthe gods (devayaja.h), the Beings (bhtayaja.h) and his ancestors (pit.ryaja.h), feedany uninvited guests or mendicants (manu.syayaja.h), and then, after his family anddependents had eaten, take his meal himself, in silence and alone.e afternoon, likethe morning, was to be devoted to study or teaching and secular matters.

As the sun went down he was to do the third sandhyopsanam, make the eveningofferings to his fire or fires (sya.mhoma.h), perform the Vaivadeva (or have it offered

without mantras by his wife), eat again and retire to bed. His wife, who had risen

before him, was to retire after him, after doing obeisance to his feet. Reproduction wasa duty; but celibacy was advocated outside the sixteen days during which a womanwas considered able to conceive; and it was imposed within that period during thefirst four days (menstruation), the eleventh and the thirteenth, and at any time on thedays of the new and full moon, the eighth and fourteenth of each lunar fortnight, or


10/46

the many other days of the year marked by special observances (vratam).e sexual

act, when permitted, was to open and culminate with the husbands reciting a shortmantra.is was the first of the life-cycle rites (garbhdhnam) in case of conception.A man was to sleep with his head to the south, after reciting the Vedas Hymn toNight (rtrisktam) and meditating on his personal deity (i .s.tadevat).

In addition to the regular and incidental sacrifices mentioned above, which hadto be accommodated within this daily schedule, was a variable but generally largenumber of optional fortnightly, monthly or annual observances (vratam) in honourof non-rauta deities, marked by fasting (upavsa.h) on the preceding day, by worship(pj), long sacrifices into the fire (agnikrya, homa.h), and a vigil (jgara.nam) duringthe preceding evening and night, and by worship the next morning, followed bythe Vaivadeva and other regular duties. Notable are observances for Vi .s .nu on the

eleventh to twelfth of each fortnight, and foriva on the thirteenth to fourteenth ofthe dark fortnight; and, annually, observances for Sarasvat on the fourth to fifth of thedark fortnight of Vaikha, for Ga.nea on the thirteenth to fourteenth of the brightfortnight of Vaikha and the third to fourth of the bright fortnight of Bhdrapda,forSkanda on the fifth to sixth of the bright fortnight, for K.r.s .na on the seventh to eighthof the same fortnight and on the eleventh to twelfth of the bright fornight of Mgha,for Ananta on the thirteenth to fourteenth of the bright fortnight of Bhdrapda, forDurgon the seventh to eighth of the bright fortnight of Caitra at the beginning of theyear and on the same days invina at the end of the rains, followed in the second caseby worship of the Nine Durgs on the ninth (Mahnavam) and public ceremonies tomark the opening of the military season (Daaharor Vijayadaam), and foriva at

the end of winter on the thirteenth to fourteenth in the dark fortnight of Phlguna (theNight ofiva [ivartri .h]). In Kashmir this was the principal festival for the majorityof the population.e ceremonies around it extended from the eleventh to the newmoon. Homes were spring-cleaned on the days from the first to the eighth before it;on the eighth married women went to their parental households and returned to theirconjugal households on the ninth or tenth bearing gifts from their parents; lavishgifts were given to daughter-in-laws by fathers-in-law on the tenth and the authorityof elders joyously suspended in ritual gambling with cowrie shells between parentsand children.e preparatory rituals performed from the eleventh to the thirteenthin which the ferocious tutelary Bhairavas were worshipped, were followed by familyfeasts, though on the thirteenth the head of the household was fasting. At midday on

the fourteenth, after the nocturnal worship of

iva and his consort on the thirteenth,the main family feast occurred, in which the head of the household, who had beenfasting on the thirteenth, joined his family. After it women gathered in public to danceand sing together. On this day the potter, the barber, the carpenter and other drasthat provided services to the household would would visit their high-caste clients to


11/46

be rewarded with gifts. In the week after the concluding rites on the evening of the

new moon day people would send each other gifts of portions of the walnuts andbread consecrated during the main worship on the night of the thirteenth.ere were many other calendrical Pjs and rddhas occupying single days.

ere were also annual, seasonal or occasional pilgrimages (trthaytr), mostly tolocal shrines of deities on their holy days or fairs (utsava.h). Every deity had a day ordays of the month sacred to it and whatever the month the fair of a deity consideredto be its emanation, aspect or embodiment would occur on that day. e biggestfairs would coincide with the second day of the major observances just mentioned:for example, there were fairs in Kashmir at the shrine of the Goddess Mahrjonboth Durg.s.tams, and at the shrine ofrikon Mahnavam. Pilgrims would batheand do a simple Pjat the shrine, and present an offering of the deitys preferred

food (bali .h), vegetarian in the case ofiva, Vi.s .nu, Lak.sm, Srya, Ga.nea and thelike, usually fish and meat in the case of a goddess or Bhairava. ey might alsoperformrddhas and make donations; and they would sometimes undertake asceticobservances (vrata) at the temple in the hope of a cure or the grantingof a boon such asthe birth of a son, sleeping on the ground on Darbha grass (Saccharum cylindricum),fasting and singing hymns of praise to the deity.ere are many instances of this kindof desiderative practice in the Kathsaritsgara.h. rddhas, donations and asceticobservances are particularly recommended when the day sacred to the deity falls ona certain day of the week, or at some special time such as an eclipse of the son, anequinox or a solstice. On an ordinary month-day sacred to a deity devotees wouldcommonly visit the temple for daranam, simply to be seen by the god (d.r to

see) and to return with its blessing. It was also common practice for the infirm andthe world-weary to go or be taken to the precincts of a temple of the deity of theirdevotion in the belief that by dying there they would go to the world (loka.h) of theirgod. Near some shrines ofiva and/or the Goddess there were places to which onecould climb or be carried in order to throw oneself to a blessed death. Towards theend of the Kathsaritsgara.h the heros aged father, the King of Vatsa, chooses to diein this way: along with his two queens and his ministers he hurls himself from a cliffonivas sacred mountain Klajara (XVI,.-).

Pilgrimages to the holy river Ganges, to distant shrines of pan-Indian fame, orto a series of such sites, were sometimes undertaken, particularly to expiate a sin orto end ones life. In the Kathsaritsgara Surak.sita is sent by his king on pilgrimage

to the sacred sites of Kashmir and its holy river Vitast

to expiate the sin of killing abrahmin (VII,.-).en there was the meritorious activity of donation (dnam), the ritualised giving

of cows, gold, provisions, manuscripts of sacred texts, land or other valuables, oftenas the termination of observances such as those just mentioned, at the end of a pilgri-


12/46

mage, a seasonal Smrta sacrifice, a life-cycle rite, or at other auspicious times. Only

brahmins were fitting recipients and among them only those who were considered tobe rich both in learning (vidy) and austerity (tapa.h). Giving to such brahmins waslike making offerings to the gods or ancestors. It was a sacrificial act bringing about asupramundane benefit rather than an exercise of virtue.

e most meritorious of these acts was to make over lands in perpetuity.iswas a gift of kings, and it is clear from the countless records of such landgrants thatit was important to Indian sovereigns as a means of establishing the legitimacy oftheir power. It was also the principal means by which the influence of Vedism wasextended throughout India and beyond it to south-east Asia. Kings in regions beyondor once beyond the territory deemed fit for the performance of sacrifices (yajiyodea.h) sought to remove the taint or suspicion of barbarism by importing groups of

orthodox brahmins and granting them lands (agrahra.h).e higher the status of the receiver the greater the merit earned. On the otherhand to receive a gift was a mixed blessing; for it was believed that by doing so one tookon some of the givers sin, who thus gained merit at ones expense. Where a learnedand ascetic brahmin condescended to receive a single gift, such as an agrah ra.h froma king, the social disadvantage was surely insignificant. But to receive habitually orprofessionally was another matter. At the end of the Pjconcluding an observance(vratam), and in others besides, the worshipper needed brahmins to whom he couldgive each of the numerous vases (kala.h) that had located the secondary deities. Hehanded over each with additional gifts, addressing the recipient as the embodimentof the deity installed in it. He then gave the principal vase with a gold coin to the

officiating brahmin. To receive such gifts was inseparable from the role of a priest; andpriests, therefore, formed a separate and inferior community with whom non-priestlybrahmins would not intermarry.

Surely, then, the process would have been self-defeating, because any priest pre-pared to accept these gifts would soon have been demeaned to the point at which he

would no longer be a suitable recipient, since receivers must be superior to givers.edilemma is resolved by the notion that brahmins by their very birth, regardless of thedegree of their learning and asceticism, have a greater capacity to resist impurity thanothers, provided, of course, that they do not fail to go through the appropriate rites ofpassage and maintain their obligatory duties.ey are, as it were, born receivers.isinnate capacity is evident from the rules concerning the period for which a person is

considered impure after the death of a member of the patrilineage. In ascending orderof caste-classes (var.n.h) these periods were thirty days for adra, fifteen for a Vaiya,twelve for a K.satriya, and ten for a brahmin. Some have found this sequence illogical,arguing that brahmins should have the longest anddras the shortest periods, sincethe higher the level of purity to be maintained the longer it should take to return to it;


13/46

and it has been suggested, therefore, that the brahmins have simply turned the system

on its head for their own convenience. But the principle in these rules is rather that thehigher up the hierarchy of caste-classes one is born the greater ones capacity to resistpollution. One can throw this birthright away by falling below the lowest permittedlevel of orthopraxy, but one can also add to it by increasing ones sanctity throughexceptional learning in the Veda (vidy) and by taking onrauta duties in additionto the Smrta. For additional rules specify that the periods of thirty and ten days fordras and brahmins respectively apply only to those of average quality. For a piousdra it is only fifteen days. For a brahmin learned in the Veda it is five days, and forone who has also installed therauta fires it is only three. So the religious justificationfor giving land only to exemplary brahmins is that this produces the greatest benefitfor the giver with the smallest disadvantage to the receiver; and priests, who receive

gifts as a regular part of their profession, though they are demeaned to the pointthat other brahmins will not intermarry with them, do not fall to the point thattheir brahminness is in question. At the very edge of this arrangement are thosedespised brahmins who accepted gifts not as the gods of ordinary rites but as thedeceased in the postmortuary rite of the eleventh day which brings to an end to theperiod of the impurity cause by the death. Priests are inferior to non-priests, funerarypriests (yamabrhma.na.h) to non-funerary priests, and the priests of the eleventh day(pretabrhmana.h, mahbrhma.na.h) to all.

Finally, there were certain other meritorious and prestigious activities which thou-ghnotdnam in the technical sense were forms of giving or endowment: constructingpublic facilities (prati.s.th) such as wells, reservoirs, bridges, and hospices (ma.th.h),

planting trees and gardens, and most importantly establishing (prati.s.th) idols andtemples to enshrine them. e consecrated idol took the name of the donor (or after anindividual nominated by the donor) as the first half of its name, the second indicatingonly that it was aiva (-vara.h), a Goddess (-r .h), or a Vi.s .nu or Srya (-svm), andthis individualisation enabled it to be treated as a juristic person, so that it could ownproperty (devasvam, devottaram) in the form of lands and other valuables gifted tofund its worship, and defend its legal rights through an officer of its temple. Againthe majority of consecrations and associated endowments were the work of kings; butlesser individuals and corporations could add to the original endowment for specificpurposes.

e various forms of religious activity detailed above are divided by the learned

into sacrificial acts (i.s.tam) and non-sacrificial acts (p

rtam), where the former areall offerings in which the recipients are gods, ancestors or sages, and the latter aredonations, foundations and charitable endowments, in which the recipients are ve-nerable brahmins or deities in their worldly character as persons in law. All Vaidikareligion, then, is the relinquishing of something that is ones own (svadravyatyga.h)


14/46

in the faith that this will lead to mundane or supramundane rewards.e principal

act (pradhnam) in all these procedures is the surrender of ownership (utsarga.h) withthe formula is is no longer mine; it is Xs.us, as the Sm.rti of Yjavalkya putsit, all Vedism (dharma.h) is giving ones own to a proper recipient, in a pure place, atan auspicious time, with the right ritual, and with faith that this is the means to onesultimate good.

is definition is useful, because it shows how the apparently disparate domainsof the sacrificial (i.s.tam) and the non-sacrificial (prtam) were seen as two aspectsof a single technique. But I have another reason for mentioning it, and that is tobring forward a more fundamental disunity within the nature of the religion wehave reviewed. For the definition, while technically correct, is ideologically biased.It comes out of a formalistic perspective in which religion was entirely a matter of

conformity to rules. In this view the only transcendental truths conveyed by the Vedaare statements that a person of this or that category should do this or that ritual.e Veda, it was argued, conveys no information but what motivates or enablesconformity to these injunctions.us it teaches the existence of an eternal self withinthe sacrificer, because one needs to believe in such an entity if one is to sacrifice in thehope of benefits beyond the grave. But it does not teach the existence of a God whocreates the world and rewards souls for their religious acts.e supposed proofs of theexistence of such a being are invalid; and to teach it would be more than irrelevant ina religion of rites, for it would undermine that religion by implying that the successof rites depends on something more than their correct execution by autonomousagents.ere is the world, there are souls, and there is the Veda instructing souls in

their duty (dharma.h). All three are self-existent, uncreated and indestructible.eremay be passages in Vaidika texts that speak of an omniscient being who creates anddestroys the world in cycles through eternity, and recreates the Veda in every aeon,but these statements must be read in the context of the body of injunctions that is theVedas only literal truth: they must be construed as allegories of the causal processesthat operate in ritual.

is atheistic doctrine corresponded well to the mentality that informed theperformance of part of what I have described.e deities in the core of the Smrtarites, that is to say, the sacrifices in fire that run parallel to therauta, the sacrifices inthe life-cycle rites, and therddhas (all, in fact, that is taught in the early manualsof domestic ritual [the G.rhyastras]), are the same as those of the ancientrauta

sacrifices, and by the Christian era, and perhaps much earlier, these deities had lostall reality for the sacrificer outside the rituals in which he invoked them: they weremerely names uttered as part of meritorious performances. But the worship of thenon-rauta gods in the daily devapj, the many calendrical observances (vratam)and pilgrimages (ytr) in their honour, the periodic, incidental and desiderative


15/46

worship of patrilineage deities (kuladevat), and the meditations that begin and end

the day, all this was another matter. Here the deities were placated with devotion(bhakti .h) as the supreme being, omniscient and omnipotent, the granter of all boonsand the saviour of souls.

How, then, did those who propagated the ritualistic and atheistic view of Vedismreconcile this view with the existence of the theistic and devotional element in Smrtaobservance?e answer is that they did not. Devotionalism was an innovation whichthey rejected. It is not to be seen in the early manuals of domestic ritual but only intheir later Supplements (parii.s.tni), and it is nowhere mentioned in the earliest andmost authoritative of the Sm.rtis, the Dharmastra of Manu, which was composedaround the beginning of our era. To find it one must turn to the Pur.nas, thesecondary authorities within Sm.rti, and to the DharmastraofYjavalkya composed

some three hundred years later than Manus. ere was nothing, therefore, whichcompelled its acceptance. ose who held the atheistic view were simply conservativeswho confined their Vedism within its ancient boundaries, that is to say, the solemn anddomestic sacrifices with their common pantheon ofrauta deities. When indigenoussources wish to distinguish these two areas of Vedism they refer to the devotionalpractices as Paur.nika (taught in the Pur.nas) and reserve the term Vaidika (taughtin the Veda [proper]) for the undisputed practices of the older tradition.

For these conservatives, of course, the rauta sacrifices were of paramount im-portance.e Smrta rituals were merely an adjunct to the rauta, a lesser domainpreparatory to the more intense life of sacrifice culminating in the Agnicayana. eirvoice in our period is preserved in the learned literature of Vedic hermeneutics

known as the Mm.ms, whose prime concern is precisely the interpretation of theinjunctions bearing on these sacrifices. Its view of the Smrta domain is expressed inthe more conservative commentators on the Dharmastras of Manu and Yjavalkya,notably in those of the great Medhtithi and his near contemporary Vivarpa.

On the other hand were those for whom the zone of intensity in Vedism was nottherauta sacrifices but this element of theistic devotion to one or other of the non-rauta gods. e older stratum of their religion, therefore, tended to be confined to theSmrta; and the disjunction between the elements of this stratum and their devotionalobservances overlooked or minimised by the devices of framing and insertion.usthe ritual day was made to begin and end with meditations on ones personal deity; thePur.nas that authorised ones devotion could dominate the daily reading of scripture;

the repetition of the G

yatrmantra in the midday sandhyop

sanam could be followedby repetitions of the mantras used in ones devotional worship; offerings to ones deity

and his emanations could be inserted at the head of the offerings to the gods in theVaivadeva; and in the worship of men (manu.syayaja.h) that preceded ones meals,one could feed mendicants and others who shared ones devotion.is integration


16/46

could be reinforced by the doctrine that the Veda was created by ones chosen deity

and that conformity with its ordinances was therefore an expression of that devotion.But the inculcation of that attitude is particularly associated with the devotees ofVi.s .nu, who were generally the most eager to be seen as thoroughly Vaidika. edevotees ofiva, while insisting that their God is the author of all revelation, tendedrather to cultivate the awareness that the devotional and non-devotional aspects oftheir religion were distinct and hierarchically ordered. Devotion to iva, even withinVedism, always involved to some extent the sense that one was above and beyond thereligion of the herd.

Between the conservative Vaidikas who held aloof from devapjaltogether andthese devotees whose devapjwas directed exclusively to Vi.s .nu,iva or some othergod, there were others who minimised the disjunction between ritualism and devotio-

nalism by practising a form of devapjin which they worshipped several non-rautadeities at once: Brahm, Vi.s .nu and iva; Brahm, Vi.s .nu, iva and Srya; those,the Goddess (Gaur/Durg), and the patrilineages tutelary deity (kuladevat); or, ina series of five that became standard in later centuries, Vi .s .nu,iva, the Goddess,the Sun and Ga.nea. One of these gods was always pre-eminent, because one idolhad to be placed in the central position and worshipped before the others. Butany devotionalism that might subsist through this arrangement was policed by ananti-devotional theology based on the Upani.sadic division of theruti, a theology

which required the orthodox to see the deities as the interchangeable personificationsof an impersonal Absolute (brahma, paramtm). is position while having theappearance of a compromise between the non-devotional and devotional extremes,

was really the former colonising and ultimately assimilating the latter. So what we haveis a dichotomy between exclusive devotionalism on the one hand and conservativestrategies against it on the other. We will consider the social distribution of thedevotional and non-devotional forms of Vedism when we come to consider theparticipants in the religion.

Before doing that let us consider the anchoretic aspects of Vedism, that is tosay, the religion of those who had renounced life in the family. e contentiousissue of devotionalism apart, the religious life I have outlined was followed by theoverwhelming majority of Vaidikas. One would pass through the life-cycle rites upto the investiture (upanayanam) that qualified one to become a celibate studentof the Veda (a brahmacr), one would study the Vedas with the required ascetic

observances (vedavratam)in earnest and over many years, or nominally and in amatter of hoursmarry, and then spend the rest of ones life observing with greateror less strictness the religious rites I have described. But not all went right throughthis process, and there were some who went beyond it. In the first place there werenai.s.thik.h, men who after investiture spent their whole lives in the household of their


17/46

teacher in the pre-marital state of celibate studentship. In the second there were men

who married and observed the duties of a Vaidika householder, but later renouncedthis life for an anchoretic form of their religion.e texts speak of two such forms. In the first a man retired to the forest into a

community of hermits, taking his wife with him or leaving her in the care of his adultson, and living in a simple hut. As a forest-dweller (vnaprastha.h) he maintainedcelibacy but kept his fire or fires; and he would continue to offer the twice-dailyand twice-monthly sacrifices but with the difference that the offerings would now be

wild (ra.nya.h) rather than the products of farming (grmya.h) which he had offeredas a householder. He wore bark-garments, skins or rags, studied the sacred texts,refused to accept donations, ate a raw vegetarian diet in the intervals between exactingpenitential fasts, and subjected his person to punishing ascetic observances. When he

was too feeble to maintain his fires he was to give them up by inhaling their essence, sodepositing them within himself. He was then to sustain himself by seeking alms fromthe other forest-dwelling hermits in its vicinity or by going as a mendicant to a village.

When he was unable to do even this he was to renounce food and walk due northuntil he died (mahprasthnam), or hurl himself from a precipice (bh.rgupatanam).

In the alternative and more radical form of renunciation (sa.mnysa.h) a man wasto terminate his sacrificial life, either as a householder or as a forest-dwelling ascetic,by instituting arauta sacrifice to Prajpati (either non-Somic or the Somic sacrificeknown as the Puru.samedha) in which he was required to give all his property as theofficiants sacrificial fee (dak.si .n). He was then to internalise his fires and go forthto seek liberation (mok.sa.h), devoting himself to meditation (yoga.h) as a solitary

mendicant without fire, living at the foot of trees, indifferent to pleasure and pain,life and death. He was to shave his head, wear a loin-cloth (kaupnam) and an ochrerobe (ka.syam), and carry a water pot (kama.n .dalu .h), a piece of wool to filter hisdrinking water (pavitram), so as to avoid harming insects, a staffconsisting of a singlebamboo cane (ekada.n .da.h) of the same height as himself or a triple-staff(trida.n .da.h)made of three such canes, and an alms-bowl. He was enter a village for alms in theevening and he was to scan the ground as he walked to avoid crushing insects. If hekilled an insect in spite of this he was to bathe and purify himself by the exercise ofbreath-suppression (pr.nyma.h).

is is how the forms of anchoretic observance are described in the Dharmastras.However, there are reasons to think that practice was somewhat different during our

period. Later authorities list the first form, that of life in a forest hermitage, amongpractices allowed by scripture but not to be adopted in the present age of the world(kalivarjya.h); and it may well be that it was obsolete in our period. And there areother sources, some of which may go back to our period, which describe sa.mnysa.h,the second anchoretic observance, in terms that suggest the same conclusion; because


18/46

in place of a single form of practice representing the most extreme renunciation we

find a system of four stages in which a person would pass gradually from withdrawlwithin the family to living in a hut outside it but still taking food from it, to completerenunciation of social relations. It appears, then, that the second form of anchoreticobservance has extended itself back towards the world to fill the space vacated by theintermediate stage of the vnaprastha.h.

I suggest that the institution of the forest-dwelling hermit disappeared becauseit could not be be accommodated comfortably within the terms of the oppositionbetween the life of rites in the family and the life of knowledge outside it, thedichotomy which, as we shall see, was at the centre of Vedism in our period.eforest-dweller does not fit either category: he is outside the territory of the familybut he lives with his wife, albeit in celibacy, and continues to sacrifice. What we

have here is rather a distinction between sacrifice in the village and sacrifice outsideit; and this is an aspect of an ancient opposition between the village and the forest,the domestic and the wild, which is fundamental to the world-view preserved inthe earliest exegesis of the rules ofrauta sacrifice. I propose that when the lateropposition became established the earlier opposition lost meaning and the institutionof the forest-dwelling sacrificer which embodied it fell into disuse.

Another anachronism may be the practice of moving to full renunciation througha Somic Puru.samedha or an ordinaryrauta sacrifice to Prajpati. Later texts donot refer to it.ey say only that when a man has decided to renounce the worldhe should perform a series of eight postfuneraryrddha ceremonies, in the last of

which (thetmarddham) he indicates that he is dead to the world by including

himself among the recipient ancestors.e attitude of the renouncer in well capturedin this ritual and in the following passage taken from the supplement of a manualof domestic ceremonies (Bahv.rcag.rhyaparii.s.tam): en he says to his sons, friendsand kin, Nobody is mine and I am nobodys. He offers water from his hand saying,I have left behind all desire for sons, for wealth, or the world of men.en he shouldcut offhis hair-tuft (ikh) and cast his sacred thread on to the ground or into water.Having done this he should face east, raise his arms above his head, declare threetimes, I have renounced, and pour water from his cupped palms with the words Toall creatures I offer peace.

Concerning the relation between these four estates, those of the student (brahma-cr), the householder (g.rhastha.h), the forest-dwelling hermit (vnaprastha.h) and the

solitary renouncer (bhik.su .h, parivrjaka.h, yati .h), there were three di

fferent doctrines.At one extreme was the view that the last two were not permissible: once one had

married and installed a fire or fires one was obliged to maintain the cycle of sacrificesto the end of ones life. At the other extreme was the view that as soon as a man wasinspired by spiritual insight to throw offthe world he should do so.is insight was


19/46

most likely to arise towards the end of life, while a man was married and sacrificing

(in the home or in the forest), but it might arise while he was still a student, and inthat case there was no reason for him to marry. It could not arise before he becamea student, because its precondition was the study of the Upani.sads, to which, sincethey were part ofruti, he could have no access before his investiture. In the thirdview, which compromises between these extremes, renunciation is permissible butmust not be allowed to outweigh the life of the sacrificer to the extent that a man bepermitted to renounce without first having married. He must marry, set up a fire orfires, perform the required sacrifices, and only when he has reached the third quarterof his life, and provided that he has a son to continue the patrilineage, is he free toleave the home. He may go from the home into either form of renunciation, or hemay enter the first and then pass from it to the second.

Underlying these three views are different doctrines of the import of the Vedaand the means of liberation. We have seen that for the Mm.msaka ritualists (theprohibitors of renunciation) the Veda was purely injunctive. e religion in thatview was altogether a matter of prescribed observances. Liberation would not beachieved by the rituals themselves but it could not be achieved if those rituals onceundertaken were ever abandoned. To do that would be to incur sin, and since sinincurs retribution, that would preclude liberation, which arises only in the absence ofany such causes of further embodiment. Ritual must be maintained, therefore; andat the same time it must be reduced to those rituals which are obligatory (nitya.h).Desiderative (kmya.h) rituals, that is to say, rituals, or modifications of obligatoryrituals, performed as the means attaining benefits (sons, cattle, long life and the rest)

must be abandoned, since actions undertaken out of desire (kma.h) bear karmic fruit,just like sins; and for the same reason the obligatory rituals themselves must not beperformed for the sake of heaven (svarga.h) but simply because they are enjoinedor, according to others, because one wishes to avoid the sin of omission. e way toliberation, therefore, is to enter the estate of the married householder and remain thereuntil death, maintaining ones obligatory rituals in a spirit of perfect detachment.

ose who allowed renunciation from the stage of studentship claimed that theVeda taught two truths, a lower truth comprising the rules of the religion of rites (Ashould do X) and a higher truth taught in the Upani .sads (the Vednta[s]), whichbore not on duty but on ultimate reality (paramrtha.h). Liberation could be achievedonly by knowing this higher truth; and the nature of this knowledge was such that

once it had arisen the religion of rites and the estate of the married householderthat it required became irrelevant and dispensable. Faith (raddh) in ritual entails,as the M.m.msakas argued, belief in autonomous agents interacting with others inan external world; but the ultimate truth is that all plurality (bheda.h) is illusory, anappearance produced by a beginningless nescience (avidy). When a man realises that


20/46

there is nothing external to his identity, that what he really is a pure unstructured

consciousness, and that this is all there is, then the injunctive Veda no longer engageshim.e lower Veda, all the Veda, that is, except the Upani .sads, is not irrelevantbefore this knowledge arises; on the contrary, the unmotivated performance of therituals it enjoins (the spirituality of the Mm.msakas) was thought to prepare one forliberating knowledge by working on the stock of ones sins and thereby purifying themind until it was capable of the liberating insight that would eliminate the need forritual.

e Mm.msakas did not deny that the Upani.sads were part ofruti. But theydid deny that they taught anything that undermined the great corpus of which theyare a tiny part.ey accepted that they are the Vedas secret teachings (rahasyni)but preferred to construe them in harmony with the domain of rites. us they

admitted that the Upani.sads teach the Vaidika the nature of his soul, or enjoin theVaidika to understand its nature; but this nature, they insisted, must be that of a beingcapable of responding to injunctions and performing acts. For this reason passages inthe Upani.sads which appear to speak of the self in terms inconsistent with ritual, asinactive, blissful and unrelated (one), must not be taken literally.

ose who accepted renunciation but confined it to the last part of life did notdiffer substantially from those who allowed it from the stage of studentship.ey toodivided theruti into separate domains, one bearing on the religion of men in the

world and the other, the Upani.sads, containing the knowledge to be realised outsideit. For both there was Vedism for rewards (either mundane or heaven) and Vedism forliberation; both accepted that the latter is confined to the estate of the full renouncer;

and both agreed that the way from the religion of merit to that of the religion ofknowledge lies in behaving within the former as though one were already in the latter,that is to say, by reducing ones rituals to the obligatory and performing them withthe indifference to rewards that characterises the renouncer.

e Mm.msakas, though they stood against these two in their view of permissiblebehaviour, stood with them in this respect. For they too accepted that the way toliberation is to cultivate the values of the renouncer within the domain of rites. Itis just that they insisted that the goal can and must be achieved without leavingthat domain. Indeed the illusionism of the renunciationists and the ritualism of theMm.msakas have this in common, that both are views that turn away from worldlyreligion with its devotional and desiderative bias to construct domains of heightened

sanctity within

rauta Vedism. In this perspective the opposition is not between theMm.msakas and both groups of renunciationists but between the Mm.msakas andthe illusionist renunciationists on the one hand and the moderate renunciationists onthe other. For though, as we have seen, the moderate view shares their soteriologicalperspectives, it is unlike them in that it is concerned to accommodate the whole


21/46

religion, and in this the devotional and desiderative elements rejected by the rauta

purists are the principal substance, the primary realities.Let us now consider the participants in Vedism. In the first place Vedism wasfully accessible only to male members of the high-status castes, that is to say, tothose accepted as ranking within the three highest caste-classes (var .n.h): brahmins,K.satriyas and Vaiyas. is was because participation was empowered by performativeutterances (mantr.h) extracted from theruti, and neither women of any caste normen of the lower castes, were permitted to use them, because they were excludedfrom the ritual of investiture (upanayanam) that alone gave access to the ruti. Womenparticipated in the sacrificial rituals of the high castes only in as much as their husbandsneeded living spouses to be qualified to perform them. A wife might take the place ofher husband in the Agnihotra, the twice-daily offering of milk to the fire, but she then

had to have a brahmin officiant to assist her, because the rite involved mantras; and shecould offer the evening Vaivadeva, but only without mantras. Areas of religious lifeopen to women were all outside this sacrificial domain. Women performed variousancillary rites in marriage and other life-cycle rituals, rites about whose details thelearned sources are largely silent, mentioning them briefly in passing as customs ofthe caste (jtycra.h), the region (decra.h), or the family (kulcra.h). Apart fromthese their main religious activities were offering minor calendrical Pjs, such asthose to Ga.nea and the Moon (Candra), going to temples to make offerings andobtain blessings, and performing ascetic observances (vratam) for the benefit of theirfamilies, such as the fast undertaken in honour of Skanda on the bright sixth in thehope of conceiving a son.

A woman would hope to die before her husband, because while he lived she sharedhis religious status and would be cremated with the rites of one with sacrificial fire(sgnika.h,hitgni .h). If her husband died first she could preserve this status onlyby following him on to his pyre (anugamanam). She was an inauspicious being (a-mangal), whose duty was to live the rest of her days in celibacy and self-mortification.She was to shave her head, wear no ornaments, sleep on the ground, eat as little aspossible, and make daily offerings of water and sesame seeds to her departed husband.

A widower could remarry and set up another fire, or, if spiritually ready, renounce theworld.

Let us now consider further the nature and degree of the highcastes participationin Vedism. All brahmins, K.satriyas and Vaiyas who had gone through the life-cycle

rites and married within the rules of caste were qualified to participate actively inthe whole of Vedism, and it is widely believed that while rauta duties were optionalor postponable they had to commence their Smrta duties directly by setting up adomestic fire with embers from the fire used in the marriage ritual.is was not so.

A married man could set up the fire only when he had received his share of the joint


22/46

family property and was therefore able to establish a separate household with himself

at its head.e ideal was that this partition would happen along with his marriage,in which case he could carry out his obligations immediately. But it was more usualthat it should occur much later, either when his father reached old age or, mostcommonly, as the consequence of his death, on the day after the consequent periodof impurity.e texts on religious duty mention only immediate partition, or adddelayed partition as a less desirable alternative.ey do not tell us how people chose.But when reading prescriptive texts one is always bound to suspect that peoples choicein such cases, if choice it was, was the opposite of the preference expressed; and inthis case the suspicion is confirmed by the rules of inheritance (dyavibhga.h), whicheverywhere assume that the norm was not the two-generation elementary familyenvisaged or advocated by the ritual prescription but the three-generation extended

family of brothers living undivided with their wives under the tutelage of their fatheror, occasionally, a four-generation family with two generations of undivided marriedagnates.

A sacrificer, then, was not any married man (g.rhastha.h), but only one who wasthe head of his own household (g.rhapati .h, g.rhasvm, prabhu .h). In the case ofrautaobservances the restriction was even greater; for the head of a household could nottake this step before his eldest brother, nor if his father were still alive. In other

words, even if a son became the head of his own household through partition beforehis fathers death, he could not install therauta fires until his generation was theseniormost, and if there were to be more than one performer ofrauta sacrifices inhis generation the first had to be the eldest brother.

Now that we have distinguished between the heads of households and the majorityof married men, and realised that the latter, unless remerged, were normally fireless(niragnika.h, anhitgni .h), we must end this account of Vedism by considering howthe various elements of non-rauta Vaidika practice were distributed between the twolevels in the structure of the extended joint family.

e series of sacrifices running parallel to the rauta, the related calendricalsacrifices, the non-calendrical sacrifices to ward offdanger or sanctify constructions,the regular and calendricalrddha rituals, and the life-cycle rites, all these were thedomain of the head of the household. So was the worship of the non-rauta deities.It was he that performed the daily devapj; it was he that fasted and worshippedin the periodic observances (vratam); and it was he that worshipped the tutelary

goddess (kuladevat

). From this level of observance we must distinguish the twice-daily Vaivadeva sacrifice connected with the preparation and consumption of meals.is would be done by the head of the family if it was nuclear; but if it was extendedthen the head of the household alone would do it only if the family units within thehousehold were sharing a single hearth for cooking (ekapka.h). Some authorities did


23/46

not approve of undivided married brothers maintaining separate hearths; but even

they admitted that the practice existed. It was not unusual for a homestead to have asingle room for devapjand a single sacrificial fire, but as many kitchens as there weremarried couples. In that case a separate Vaivadeva would be prepared and offered ineach family unit.

Below this level were the daily obligations common to all the married men in thehousehold regardless of their status.ese were the sandhyopsanam, the pouring oflibations to the gods, sages and ancestors, and the reading of the Veda (brahmayaja.h,svdhyya.h). In the case of the undivided the sandhyopsanam would be done onlyat dawn and sunset, because the midday performance was an ancillary of the middayrituals of the patriarch alone.

e undivided, therefore, were latent persons in that they lacked both independe-

nce and the religious prestige that is bestowed by setting up the fires and worshippingthe gods.e purpose of the dawn and sunset sandhyopsanam is humble enough:it is said that the dawn rite removes the sins of the night and the sunset rite the sinsof the day. Having considered the participants and their roles I end this outline ofIndias basic and pervasive form of religion by asking what is social and what personal

within it. Evidently, if social religion is whatever rites and observances are performedby persons, or for persons, in their identity as members of a caste or family, andpersonal religion is any discipline that a person undertakes in the belief that he isacting as an individual rather than in his social identity, then the line between the twoin Vedism will lie exactly along that between the domain of rites and the domain ofrenunciation.

e religion of the renouncer is unambiguously personal. e effect of his knowle-dge may be to obliterate all sense of personality; but that is another matter: what, hewould ask, is personality outside relations with others?e relevant considerationsare that he seeks his knowledge by cutting himself offfrom all social relations andobligations, that society accepts that he has done so, and that his action is for himselfalone. If social religion exists here at all it is only to the extent that he can renounceonly from within that religion. It is a necessary, though by no means su fficientcondition for entering the path to liberation.

e religion of rites is entirely social in as much as rituals cannot be performedhere except by virtue of the agents position in the household and the position ofhis household in the hierarchy of castes. Optional rites, those undertaken for some

special benefit, are always social in this sense at least; and they are usually social in astronger sense, for while some may benefit the performer alone, the great majority areperformed to benefit the whole household, either directly, as when the purpose is to

ward offor counteract disease, famine or some other misfortune, or indirectly, sincein the joint extended family the material prosperity of the head of the household is


24/46

that of all the coparceners under his tutelage. Obligatory rites are social in the first

sense and another. For they are performed not merely by virtue of a social positionbut to create, express, or consolidate it. Indeed they give the appearance of beingperformed by individuals as the representatives or embodiments of their households:the patriarch acts for all. is is particularly obvious on such major calendrical festivalsas the Night ofiva(ivartri), when, as we have seen, the whole household is involvedin observances with and around the patriarch.

How far does this classification of Vedism and its purposes correspond to whatwe find in thought about the religion within the religion itself? Only incompletely.e learned recognised that the religion of the renouncer is personal, but there wasno room in their thinking for the idea of the social in the religion of rites beyond thenotion that it is carried out by individuals in society and as a function of their social

status.ey did not articulate the view that this religion is dominated by what is donefor social groups or, one might say, by those groups through their representatives. Arewe then imposing alien concepts on to the data and thereby distorting their meaning?I believe not. If the category of social action is missing from indigenous theory it isbecause thought about religion in our period is driven by the individualism of therenouncer. In that perspective there was no room for the concept of collective action:all action was that of monads, who acted for themselves and experienced the results oftheir actions without the intervention of any external factor. Society did not exist here:there was simply a hierarchy of ranks into which souls are incarnated in accordance

with the merit and demerit of their actions, and this hierarchy was set within a greaterhierarchy extending from hell-beings through animals and humans to gods. ose

who neglect religion descend for punishment to the hells before returning once morethrough an intermediate incarnation as an animal or untouchable to a qualified caste.ose who recognise that religion is the means of rising after death and so observe itsordinances are rewarded by incarnation in the world of Brahm. But that too is onlytemporary: because merit is finite, so is its reward. e way out of this beginninglessand potentially endless system of rewards and punishments is to break away from thedesire that drives it: to cease to act for good or bad and to renounce the world. So seenin the perspective of this renunciationist doctrine the opposition between the world ofrites and the world of renunciation is one between two kinds of personal religion, onecognitive and detached for liberation and the other active and desiderative (kmya)for the world of Brahm(brahmaloka.h).

In our period this perspective was so thoroughly established that even the anti-renunciationist ritualists could think only in its terms. And it was not just that theyperceived the sacrificer as an unrelated monadic agent.ey also superimposed thevalues of the renouncer on to ritual action itself. ey rejected the renunciationistview that the world of rites is entirely desiderative and therefore incapable of leading


25/46

to liberation. Instead, they transposed this dichotomy into the world of rites by

making it correspond to the non-soteriological distinction between obligatory andoptional observances.ey claimed that only optional rites perpetuate the cycle ofincarnations (sa.msra.h), because they are motivated and must therefore be rewarded.But obligatory rites, provided they are performed without desire for any result, in therenouncers spirit of detached inaction, will not propel the soul towards retribution.Indeed performing them in this spirit is, as we have seen, the only way to liberation.ere is, therefore, a higher life of ritual, which has the same effect as renunciationand renders it unnecessary.

e older view, which sacrificed the social agent, but which, since it had no reasonto challenge the institution of the renouncer, had no reason to conceal the worldly,desiderative character of the world of rites, is clearly stated in the the Dharmastra

of Manu and in the more or less coeval Mahbhratam, though the Mm.msakaperspective of later centuries has tended to gloss over it in the commentarial literatureof our period. However, it was also very much alive in the devotional Pur .nas,particularly in the main authorities for the devotees ofiva, theivadharm.h andivadharmottaram.e religion of men in the world is the principal subject ofthose still unedited texts, and they present it entirely as a desiderative religion ofmerit (pu .nyam).ose who undertake the rites and observances it prescribes oradvocates will achieve a better incarnation, either in this world, typically as a righteous(dhrmika.h) monarch or learned brahmin, or inivas heaven. And the practices ofascetics are often presented as bestowing no higher reward than the latter. In thisrespect they are to be classified from the Dharmastric point of view as equivalent

to practices in the estate of the forest-dwelling hermit rather than in that of the fullrenouncer, since those practices being within the domain of rites or at least not clearlyout of it are said by Manu to lead one only to the world of Brahm .

ese devotional Pur.nas, then, have escaped the influence ofrauta exegesissufficiently not to have disguised the realities of religious motivation. Might they alsoshow a great awareness of the social nature of religious action? I believe they do.For in their prescriptions of observances they declare very frequently not only that the

worshipper or donor will go to the world ofiva but also that the rewards of his actionwill be shared by his family, dependents, ascendants and descendants. A commentatorcommitted to the perspective of high brahmanical exegesis would automaticallyobserve that these promises are not meant literally but are only rhetorical praise

(arthav

da.h). But I prefer to see here an unmodified awareness of the social realityof religion in the world. Not only because the ivadharmottaram itself forbids thereader to apply the Mm.mss distinction between the literal and the rhetorical(vidhi .h and arthavda.h)it insists that everything in it is literally true, includingits promises of rewards, but also because there is so much else in our texts that


26/46

contradicts the individualistic, renunciationist perspective that dominates systematic

thought in Vedism and yet is in harmony with widely held beliefs; and now I amspeaking not just of the texts of devotion toiva but of the whole body of Vaidikareligion.us, for example, the influence of the renunciationist is perhaps nowheremore fundamental than in the doctrine of karma. e essence of this doctrine in(renunciationist)Vedism, aivism, Vai.s .navism, Buddhismand later, exegeticalJainismis that it is intention that makes an action into a force leading to retribution for thedoer. Unintentional actions are mere happenings with no karmic consequences. Yetthe practice of Vedism contradicts this rational doctrine. e case of evil deaths(durmara.nam) will suffice to illustrate this. If one is killed by an untouchable, by asnake, by a brahmin, by lightning, by a tusked animal, by drowning, or by beingcrushed by a vehicle, or if one dies as the result of an infectious disease, or at the

wrong time of the month, it is believed that one will go to hell, and the proof of thisbelief is the existence of a ritual (the nrya.nabali .h) designed to rescue one after ayear has passed.is makes no sense at all in terms of the theory of intention, andsome of the commentators have shown that they were aware of this by making feebleattempts to moderate the contradiction by arguing that one goes to hell only if onehas provoked ones slayer.ey do not tell us how this is condition will apply in thecase of lightning, disease or the day of the month.

It is not enough, then, to look to the renunciationist, anti-devotional and ritualistdoctrines of Vedism for an explanation of its beliefs.eir source is outside the worldof rites and although they have been widely assimilated and are therefore a real partof the religious thought of men in the world they cannot serve to clarify the greater

part of what those men in the world were doing. For that one must look instead tothe implications of the actions and interactions that make up that practice. Whenone does so one finds that implied beliefs contradict explicit doctrine at every turn.

And because they do so the exegetes whom we might expect to clarify the principlesthat govern this world of practice are largely silent when we need them. Instead ofexplanations they give us only discussions of rules, options and their application. And

when beliefs are expressed in the texts, then, if these make no sense in terms of theofficial doctrines, the exegetes dismiss them as rhetorical commentary (arthavda.h) onthe injunctions (vidhi .h).is device is the very essence of high brahmanical thought,for that holds either that the Veda is purely injunctive, so that it contains no doctrinesat all, or that the doctrines it contains have no bearing on active religion. One is

off

ered doing without knowing or knowing without doing.is dualism, however, was an active part of the religion we are describing, atleast among brahmins.e ritual agent was at once, or rather in different framesof reference, both a creature of society and an individual holding himself alooffrom it.is ambiguity is well revealed in the structure of qualification within the


27/46

household.e higher an individual rose within that hierarchy, the more interactive

and representative of others his observances became. He took on more and moreresponsibilities, which involved him in an ever widening network of patronage: as aperformer ofrauta sacrifices he might be not only the head of his own householdbut also the senior representative of a ku.tumba, a group of such households; andas a paterfamiliasconsider the example ofivartrihe interacted not only withthe transcendental gods, the tutelary goddess, and the ancestors of his patrilineagebut also with the families of his wife and daughters-in-law, and with the variousdras on whose ritual and other services his household depended. At the sametime, however, he was to see himself as rising through these stages in the oppositedirection, from involvement in the mundane towards transcendence. When in themidst of the feasting and rejoicing household the head withdrew himself to fast and

worship he communicated not only that he was acting on behalf of all but alsothat he was doing so by virtue of his greater capacity to transcend the mundane, tomerge himself withiva through worship. Transcendence through seniority is also

written into the non-devotional Vaidika system. When the patriarch ascended fromthe level of Smrta sacrifices to that of therauta sacrifices he raised himself beyondthe domestic domain into a higher individuality and sanctity. He was an ascetic whileconsecrated (dk.sita.h) for the duration of the sacrifice, and through the Agnicayanasymbolised his translation from the world of desires to that of eternity. rough thePuru.samedha he could leave that world for ever, or else renounce symbolically and,like the head of the household rejoining his family for the midday feast after the Nightofiva, return with heightened sanctity to the mundane world.

I turn now to the followers of non-Vaidika forms of religion, and first to thosethat did not reject the Veda outright, that is to say, to the aivas and Vai.s .navas.As we have seen, both included the Veda within the domain of valid scripture, butsaw the religion it ordained as a lower, mundane observance which the favoured fewcould transcend through initiation into theaiva or Vai.s .nava observances taught ina higher division of revelation. As for their attitude to each other, the aivas sawthe Vai.s .nava scriptures as revelations intermediate between the Vaidika and theirown, whereas the Vai.s .navas generally rejected the aiva scriptures as works createdby God to delude sinners (mohastr.ni).e Vai.s .nava style of social self-assertionhas been to claim the status of superior insiders.ey saw themselves, and wished tobe seen, as a God-blessed lite more scrupulous in the avoidance of impurity than

the Vaidikas themselves and beyond them spiritually through their access to Vi.s .nushigher teachings.eaivas on the other hand preserved throughout our period amuch less ambiguous commitment to the values of transcendence. is orientationis in keeping with that of Rudra in therauta world. For that deity, with whom theaivas identified theiriva, is the deity of the outside, a powerful and ruthless being


28/46

who must be placated at the edge of the rauta sacrifice to protect the ordered world it

symbolises. In stark contrast is the position of Vi.s .nu in the ruti. ere he is identifiedwith the sacrifice itself, as we are frequently reminded by his devotees in support oftheir claim that they are in harmony with Vedism. And though he transcends themanifest world he also pervades it, so that in the Vai.s .nava metaphysics that translatesthese symbols the whole universe is Vi.s .nu embodied. Inaivism, except in certainlate forms not relevant in this period,iva is a purely spiritual being above the world,the world itself is of an entirely non-spiritual nature, and the way toiva, and therebyto liberation, is by gradually ascending a ladder of worlds or ontological levels untilone reaches its summit.

aivism and Vai.s .navism exhibited their transcendence of the mundane religion indifferent ways; and, as we shall see, there were differences withinaivism itself; for it

comprised a number of distinct systems of observance, and it was in the manner anddegree of overt transcendence that they differed most significantly. However, bothVai.s .navism and these various forms ofaivism had certain features in common:

. In all one entered through a ritual of initiation (dk.s) which consubstantiatedone with the deity and bestowed a new and higher identity marked by aninitiatory name (dk.snma), which replaced or, in the case of householders,

was superimposed on that received in the social religion. ere, when a brahmininstituted a ritual, he was identified by his officiant in the opening formulaof intent (sa.mkalpa.h) firstly under the name he had received in the post-natalnaming ceremony (nmakara.nam) and secondly as a member of a certain Gotraof brahmins, that is to say, of a group comprising the married men patrilineally

descended from a mythical founding sage (.r.si .h), or, more commonly, a divisionof that group, together with their wives, sons and unmarried daughters. AK.satriya or a Vaiya was assigned the Gotra of the brahmin who was hishereditary family priest (purohita.h); and all dras were assigned to the KyapaGotra, as were brahmins whose Gotra had been forgotten. Like the namereceived in the rite of passage this Gotra membership derived its meaning fromthe most unambiguously social level of the exoteric religion; for its primaryfunction lay in the rules governing marriage alliances: to be qualified throughmarriage to participate in the religion as a member of a Gotra a brahmin hadto have married a girl born into a Gotra other than that of his patrilineageand also, according to some Vaidikas (the Mdhyandinas), other than that

of his mothers fathers patrilineage (mtmahagotram).e same applied toK.satriyas and Vaiyas through the Gotras of their priests. Fordras, of course,this primary function was inoperative: if not, then since all dras are of oneGotra nodra could marry adra, which would contradict the fact of casteendogamy. Adra was assigned a Gotra only to enable him to share the formal


29/46

consequence of Gotra membership, namely qualification to institute Vaidika

rituals. Ones name and Gotra, then, were ones identity as a performer of ritesin the exoteric religion.e Tantric initiates transcendence of this identity wasindicated by giving him a name belonging to an aspect of the deity, so that he

was now to be seen more or less literally as its embodiment; and in his funeraryrites (antye.s.ti .h) and subsequentrddhas he was identified by means of thisname and as a member of his deitys supramundane Gotra. us, for example, abrahmin beneficiary Goplaarman of the Kashmirian Ka.n.tha-DhaumyyanaGotra would be identified after Tantric aiva initiation as a member of theiva Gotra with a name such as Aghoraiva, Aghora being an aspect ofiva,or as a member of the Bhairavagotra with a name such as Aghorabhairava or

Aghoraakti.e suffixed title -arman indicated that he was a brahmin. His

initiatory name ought to have given no indication of his caste-class, if, as wasinsisted, initiation had lifted him above the level at which such distinctionsare relevant. In practice, however, it might do so; and whether it did or didnot depended on a traditions strategy of relating to the exoteric religion.usnames in -bhairava and -akti gave no indication of caste and were given inthe more radical forms of Tantricaivism, whereas names in -iva indicatedthat the initiate is a brahmin, and were given in the more exotericaiva systemknown as the Siddhnta.

. rough initiation all believed that they had gained access to the means ofattaining supernatural powers (siddhi .h) and/or liberation far beyond the reachof the practitioners of Vedism.Indeed in most systems a greater part of initiation

was the ritual destruction of all ones past actions and their capacity to give riseto effects, and these included not only ones sins but also ones merits (pu .nyam).

What one had achieved through ones previous religiosity in the Vaidika domainwas merely an obstacle now.e heavens of Brahm, Vi.s .nu oriva to whichones accumulated merit would have sent one at death were all part of the worldof limited and transient experience to be transcended through initiation.

. An important and distinctive element in Paur.nika representations of the non-rautagods is mythology.

an outline of brahmanism ca. 900 c.e

Documents