II

_,.,.,. .......
Texts and Traditions:
The Bengal PuralJ.as

In the older Vedic literature, the word purii1)a usually occurs in connec-
tion with itihiisa and originally it seems to have meant old narrative,
without any special significance as to the character of the narrative.' By
definition a Purii1)a is supposed to deal with five topics or paiica-
la/cyQT)a: the creation of the universe, re-creation after destruction, gen-
ealogy, periods of time (with Manu as the primal ancestor), and the
history of the solar and lunar dynasties. The Purii1)as, however, seldom
adhere to this definition which suggests that if the definition was based
on content the present Purii1)as rnus1 be thoroughly revised versions of
either the originals or at least what they were ideally intended to be. The
Purii1)as themselves state their number to be eighteen. Originally, like
the epics, they were narrated by siitas or bar1s. Thus in almost all the
Purii1Jas the siita Lomah~al)a or his son souti Ugra,Sravas appears as
the narrator. However, according to the orthodox tradition as recorded
in the Atharva Veda and BrhadiirQT)yaka Upani~ad, the Purii1)as were
of divine origin.2 Even the Puriil;las claim that the chief speaker gathered
his information, through Vyasa., from the creator himself; this has led to
the speculation that originally there was a single Purii1)a from which the
later ones were derived. Later the Puranic material was extensively
revised by briihma1)a priests who continued to add matters of religious
and social importance to the original core, so that the extant Purii1)as
have practically turned into Smrti codes.
 17te Bengal PuriJ.1as I 45

The PuriilJQ texts, as we have them now, were written at such diver-
gent periods that it is difficult to fix their cbrooology with even an
approximate degree of certainty. Although the word puriil;la occurs in
the Vedic literature, their actual existence can be traced from the Sutra
period. Thus references to the Puriil;las in the Dharma SU!ras of Gau-
tama and Apastamba and in the Mahiibhiirata seem to indicate that they
existed long before the Christian era. 3 R.C. Hazra has shown how dif-
ferent sections dealing with orthodox rites and customs in the manner
of the Smrtis were added on at widely varying periods.4 Hazra 's method
may be described as a close comparative analysis of the Pw:an.ic mate-
·a1 l
n.
This however is merely scratching the surface of the problem: in the
great majority of the cases, the Puriil)as have not come down to us with
only early incorporations; they have undergone continuous and substan-
tial re-editions. This re-editing was done in three different ways: by
adding fresh chapters, by substituting older ones with new, and by writ-
ing new works bearing old titles. But Hazra argues that all these prac-
tices had a· common feature in that they came to have units belonging
to different ages, and that fresh additions were not always fresh com-
positions but chapters and verses transferred from one to the other and
from the Smrtis to the Puriil;las.6 He proves the truth of his asswnption
by comparing almost verse by verse the contents of the eighteen Puriil)as
and other Smrti literature, and thus constructing a comparative chronol-
ogy of the different sections of the Puriil)as. 7
What necessitated this ceaseless revision? The Puriil)as themselves
assert that in order to keep pace with social change such periodic revi-
sions were required, 8 so that their importance as works of authority
might not decrease. Thus the texts grew into encyclopaedic works by
incorporating chapters not only on religious and social matters, but also
on law, politics, poetics, grammar, medicine, music and sculpture.9 If
the Puriil;las were quoted as authority in Medhatithi's commentary on
the Manusmrti as early as the ninth century, and if by the medieval
period they had acquired a status equal to that of the Smrtis,1° they
obviously must be continuously revised to accommodate new social
needs and incorporate matters which were traditionally considered to
have been exclusively the province of the Smrtis.
Hazra identified two main stages in the development of Puranic rites
and customs. He argues that ancient India saw the rise of various reli-
gious movements which he classified as Vedic (Srauta and Smiirta),
anti-Vedic (Buddhism, Jainism or Ajivikism), semi-Vedic (consisting
 46 I Religious Process

p~arily of Vai~vism, Saivism and Brahmlism), and non-Vedic

(Siktism). The recognition of the authority of the Vedas and the supe-
riority of the brahmtuJas are the two essential constituents of the Vedic
brahmanical religion. Hazra suggests that probably long before the time
of Gautama Buddha there were protests against the brahmanical doc-
trines and with the rise and popularity of Buddhism, Jainism and Ajivik-
ism these protests were institutionalized. Hazra adds that the original
character of Vai~vism and Saivism was also most probably non-
brahmanical, as both the Piiicaritras and the PiSupatas were irreverent
to the van:iiiiramadharma. Besides, Vedic religion got into further
trouble with the transfer of political authority from qatriyas to SUdras
under the Nandas, Mauryas and probably the Andhras. 11
As a result the brahmanical social order was disrupted to such an
extent that we perceive an obsessive fear of the Kali age in all the early
Pura1,1as. In that age people neglected the caste rules and the duties
enjoined by the J!.g, Siim and Yajur Vedas. The brahma1,1as gave up the
study of the Vedas and the performance of sacrifice and became wan-
dering mendicants. They lost their social and ritual supremacy and peo-
ple of all castes claimed equal status with them. The qatriyas and the
vaiSyas were almost extinct and the prevailing caste was the SUdra who
acquired political importance. Demoralization in the Kali age was par-
ticularly manifest' in the conduct of women, who became wicked and
self-willed, joined the Buddhist sarigha in large numbers and disre-
garded the four stages of life (caturiiirama), one of the indispensable
preconditions of tbe brahmanical way of living. 12
Thus every aspect of the fabric of the ideal brahmanical social struc-
ture was undermined. Consequently, an attempt was made to re-establish
the authority of the Vedas, the van:iiiJramadharma and the moral order
among women and lower castes. This attempt seems to have been made
by two sections in two different ways: by orthodox .brahma1,1as who
began to preach the performance of grhya rites through smrti works and
by the more numerous Smarta-Vai~~vas and Smiirta-Saivas who intro-
duced Smrti materials into the Mahabharata and Purar;ias for the propa-
. gation of their sectarian points of view against the heretical religions by
upholding the brahmanical social order.13
Hazra locates the second stage in the development of Puranic rites
and customs in the third to sixth centuries when Vedic religion was
threatened by invaders from the north-west. identified by Hazra as
Sakas, Yavanas, Bihlikas, and others. The early Pural,IQS are eloquent
about the evils brought about by these rulers. 14 From the brahmanical
 '11le Bengal Purtl)U / 47

point of view the socio-political disruption created by the rule of such

dynasties was somewhat remedied by the rise of the Guptas, but fiom
lbout the beginning of the fifth century the Vai'91vas and the Saivas
came under the influence ofTantrism. Hazra has shown bow an analysis
of the contents of the Pliicaritra Samhitiis and the Saiva Agamas deci-
sively proves that the Tantric cult attained popularity at a very early
period. The Tantras believed in equal opportunity of worship to all,
including women and SUdras, and the ritual practices recommended by
them were distinctly non-brahmanical in character. The spread of such
ideas obviously affected the existing status of brahmanism in such an
adverse manner that the autbon of the Puriil)as could Iio longer remain
satisfied with introducing only those S"'!fi topics which came within
the scope of the earlier Smrti Samhitiis; it was necessary to add chapters
on popular religiow practice such as piija (popular wdtship), vrata
(vow), homo (sacrifice), sandhyii (daily rites), utsarga (gift), tirtha (glo-
rification of holy places), and prat~fha (consecration of images). These
they tried to free from Tantric influence and infuse with Vedic rituals as
far as practicable. The occurrence of these topics in the comparatively
late Pura~ and the way these Puriil)as denounce the scriptures that
imbibed Tantric characteristics support this view. 1' However, the ele-
ment of Vedic ritual in the Pllriil)as diminished progressively with time
and was increasingly replaced by popular material. In fact by the begin-
ning of the ninth century some of these texts began to recognize even
the lllntras as an authority on religious matters. 16
This is how the Puranic tradition came into existence and eventually
crystallized through a process of continuous revision. Tradition, how-
ever, is, as Heestennan puts it, 'the way society formulates and deals
with the basic problems of human existence', and as these problems are
insoluble, they are 'attacked, formulated and dealt with each time anew
under a different aspect. Tradition therefore is and has to be bound up
with the ever-shifting present. Hc:nce lhe irritating flexibility and fluidity
of tradition.' 17 Romila Thapar has also emphasized the aspect of 'con-
temporaiy requirements' as a determinant of tradition and consequently
its renewability and the need to view it 'in its various phases' .11 Even
the eighteen Pural)llS must therefore have proved inadequate for
brahmanical requirements at some point of time: another group of lit-
erature, belonging to the same genre but with a different emphasis, came
into being. This was the Upapuriil)as, the fonnation of which, according
to Hazra, should be placed approximately between AD 650 and 800.1 9
This approximate date must not be taken to be the period of composition
48 I Religious Process
of the individual texts of the UpapuriUµJS, for all of them could not have
been written at the same time. Regarding their origin, the Siva-miihiit-
mya KJuuJ4a of the Kumra Puriir,la records a tradition that the sages
proclaimed the Upapuriif)as after listening to the eighteen Puriif)as from
Vyasa. 20 This tradition assigns to the UpapuriilJQS a date following that
of the PuriilJQS and thus an inferior status. Indeed, the Matsya Puriir,la
calls the Upapuriir,las mere sub-sections or supplements-upabheda-of
the principal PuriirµJS. 21 The Upapurlir}as, however, do not often loc;>k
upon this characterization with the same respect as the principal or what
were later called the Mahiipuriir,las, as a large number of the
UpapuriilJQS call themselves simply Puriif)a and do not try to attach
themselves for the sake of authority to any of the principal Purlir}as.
The non-inclusion of the term Upapuriir,la in the Amarako~a which
defines the Puriir,la in accordance with the paiicala/cya1Ja, and the men-
tion of the titles of the eighteen Puriir,las in the V~IJ" and Miirkat;u!eya
PuriitJas without any reference to the Upapuriir,las, shows that the group
of eighteen Puriit;1as bad been formed before the Upapuriif)as came into
existence. 22 After this group bad been completed, there came into focus
many sub-systems and offshoots of the Bhigavata Piiicaritra and Saiva
Pasupata systems, either directly, or by identifying the local deities with
one or the other of the prominent deities of the principal systems. In
addition, there were also other independent systems such as Saura and
Sikta, which began to rival those already established. These sub-systems
also had their Smiirta adherents who interpolated chapters in the
PuriilJQS already in vogue and in some cases wrote new and independent
Puranic works in order to propagate their own views. But as the follow-
ers of the principal Puriif)as believed that there could be no Puriir,la
beyond the eighteen, they were unwilling to assign to these new works
a status equal to that of the Mahiipuriir,las. On the other hand, these new
Puranic works had become too well known and popular to be completely
ignored. The problem was resolved by maintaining the numerical rigid-
ity of the eighteen •original' Puriit;1as and assigning to the Upapuriir,las
an independent albeit subservient position. 21
Following the tradition of the Mahiipuriir,las, orthodox opinion also
tried to limit the number of Upapuriif)as to eighteen, but while in the
enumeration of the Puriif)as there is almost complete agreement with
regard to the titles, this is by no means the case with the Upapuriif)as.
Hazra has compiled more than a hundred titles from the conflicting
lists of Puriif)as, Upapuriif)as and other texts, most of which are no
longer extant. 24
 11re Bengal Purii)as I 49
One of the major differences between the Mahapurar:ias and the
Upapuriil;las is that the latter are regionally identifiable. It bu been
claimed that even some of the additions to the Mahapuriil;las have a
local tinge so that the Brahma Puriif)a may represent the Orissa version
of the original work. just as the Padma may give that of ~kara, the
Agni that of Gayi, the Varaha Mathuri, the Jliimana ·Thine5var, the
Kiirma Vlrinasi, and the Matsya that of the brtihmar:ias of the Nar-
madi 25 In their entirety, however, none of these texts can be said to
have been composed in these regions and although they partially reflect
local interests, it is impossible to, precisely locate their place of origin.
The Upapurar:ias, on the other hand, are so overwhelmingly regional in
their concerns, catering as they did to local requirements, that they can
a
be identified with particular locale with a fair degree of certainty. Thus
Hazra points out that most of the available Upapuriif)as were written in
the areas peripheral to the brabmanical sphere of influence. For example
the Vi~r:iudharmottara was composed in Kashmir,26 the Ka/ilea in
Assam, 27 and a great majority in Bengal. Another curious fact to which
Hazra draws our attention is that while the Mahapura1.1as, because of
their status, have undergone innumerable modifications, the Upapuriif)as
have remained comparatively free from later redactors. They are thus
better preserved and among those extant there are some which are much
older than many of the MahapuratJas.21
While even the Mahapura1Jas did not entirely adhere to the classical
definition of the pafical~a1Ja, the Upapuraf)as, which were almost
exclusively adapted to suit the requirements of local cults and the reli-
gious needs of sects other than those already assimilated in the
Mahapurar:ias, are never in consequence particularly concerned about
these five topics. They especially neglect the genealogies of kings and
sages, and even when they mention them, it is only to acquire a stamp
of antiquity, and nothing is said about the dynasties of the Kali age.
What the Upapuriif)as do discuss are the religion and social structure,
particularly the assimilation of divergent values and practices through
the reformulation of certain fundamental principles of brahmanism in
the peripheral areas. They tell us about mythology, idol-worship, theism
and pantheism, festivals and ceremonies, ethics and superstitions; in
short, everything that constitutes the domain of religion in the widest
possible sense.29
In deciding the date and provenance of these texts, I have accepted
the suggestions of R.C. Hazra,30 which are as follows:
 SO I Religious Process

Name Locale Date

Brahmavaivarta unspecifiable, but so tenth to sixteenth

Puriif}a completely remodelled in centuiy31
Bengal, that it can be taken
to represent unambiguously
the specific regional context
of Bengal31
Kriyiiyoga3iira Beng1d33 not later than
(claims to belong eleventh ccntwy34
to the Pudma
PurO]Ja) 31
Brhannaradiya eastern Orissa or western 750 to 9003b
PuriirJa Bengal (recognized by the
early Smrti writers of
Bcngal)3s
Devi Puriif}a Bcngal37 cannot be placed
later than 85031
Kiilikii PuriiTJO either in Assam or that part tenth or the fust
of Bengal which is adjacent half of the
39
to Assam eleventh century40
Mahiibhiigavala eastern Bengal, close to tenth or eleventh
Puriif}a Assam41 century, not later
than twelfth
century42
Devibhiigavata Bengal43 eleventh or
Puriif}a twelfth ccntury44
Brhaddhanna Bengal4s latter half of the
Puriif}a thirteenth century4b

It may be. observed that except the Brahmavaivarta, all the other
· Puriir;as were written between the eighth and thirteenth centuries, and
most betw·een the eleventh and twelfth, so that the Bengal Puriir;as
may be broadly labelled as the product of the early medieval period
Indeed, by the late fifteenth century Puranic digests such as the
Puriir;a-sarvasva were being compiled in Bengal47 which shows. that
by then these Puriir;as had been accepted as nonnative texts and it
was considered worthwhile to produce abridged compendia containing
their essential message. '
Numerous manuscripts of these Puriir;as are preserved in various
 111e Bengal PuriQas I SI
collections all over the country and abroad Sanslaitists have prepared
'complete texts' from the extant manuscripts and all except the Devi
Puriit;ia, are available in more than one printed edition. I have used
the Valigavlsl Press (Calcutta) editions for this study for the following
reasons.
Only the Valigavlsl Press bas published all the Purii1.1as which have
been identified as belonging to Bengal. For those Puriif.las which have
several editions, the difference between the Valigavisi and the others is
minor and, for my purpose, negligible. Besides, the texts of the
ValigavisI editions are fuller in comparison with the others, except in
the case of the Devibhiigavata. The chief editions of the Devibhiigavata
are those published by the Samslqta-pustakilaya (Benares),
VeiikaJe5vara Press (Bombay), and ValigavisI Press. All have practically
the same text, but the Sarilslqta-pustakilaya edition contains a Devi- ·
bhiigavata-miihiitmya in five chapters, which extols the virtue of the text
and the merits of reading it, not very different from what the text claims
in bits and pieces throughout and more elaborately towards the end. The
Brahmavaivarta was published by the Sarasvati Press (Calcutta),
Valigavisi Press, VeilkaJesvara Press, and AnandASrama Press (Poona),
and the Kriyiiyogasara by the Vaiigavisi and VeilkaJesvara. These edi-
tions are not substantially different from one another. The VeilkaJe5vara
and the Valigavisi editions of the KiilikQ contain 93 and 90 chapters
respectively; in the VeilkaJesvara some chapters of the Vaiigavlsi are
split into two, but the text in both the editions is more or less the same.
The Gujarati Printing Press (Bombay) and the Vangavisi editions of the
Mahiibhiigavata have the same number of chapters, but the latter at
times contains some additional verses. The Brhaddharma bas been pub-
lished by the VaiigavisI Press and the Asiatic Society of Bengal (Cal-
cutta). These two editions differ in readings on numerous occasions but,
more importantly, the last seven chapters of the U1tarakha1J<fa, contained
in the Vaiigavlsi edition, are missing in the latter. These chapters, how-
ever, must not be taken as spurious, for they occur in almost all the
manuscripts of the text. The Brhanniiradiya has also been published by
the Valigavisi Press and the Asiatic Society of Bengal, but in the latter
edition the verses have not always been correctly numbered. For the
above reasons and for the sake of uniformity, I have relied on the
Valigavisi editions.. Recently, Navabhirata Publishers (Calcutta) have
started reprinting the Vailgavisi edition of some of these Puriinas, re-
edited by Srijiva Nyllyatirtha, which I have taken into account. But the
re-editing only corrects the editorial lapses of the previous edition.
52 I Religious Process

All the translations from the original texts are my own.

II

The Puriil;las were thus an instrument for the propagation ofbrahmanic:al

ideals of social reconstruction and sectarian interests, a medium for the
absorption oflocal cults and associated practices, and a vehicle for popu-
lar instruction on norms governing everyday existence. They combined
scripture and the social codes of the Smrtis in a manner acceptable to
most people. In short the Puriil;las performed the delicate task of
operating simultaneously at several levels, widening their scope to ac-
commodate local elements as much as possible and involve as many
people as permissible without compromising their principal objective of
establishing the brahmanical social order. In the process at times they
even ran the risk of appearing to be self-contradictory, without losing
their essential unity. The technique of accomplishing this task may be
described as the Puranic process. The cumulative effect of the compo-
sition of the Puriil;las and the popularization of Puranic religion was the
construction of the somewhat imprecise category of what is popularly
known today as Hinduism. The impressions of two English civil servants
of the early twentieth century regarding wha.t constitutes Hinduism will
serve to illustrate the point. L.S.S. O 'Malley, who edited the Bengal Dis-
trict Gazetteer in 1912, describes Hinduism thus:
Considered purely as a religion .. . Hinduism may be described as a conglom-
erate of cults and creeds. The non-Aryan tribes who were admitted to the fold
of Hinduism and the Hindus of Aryan descent reacted on one another, the former
adopting the rites and customs of their conquerors, while the latter assimilated
some of their less civilized cults and incorporated in their system the objects of
popular devotion. The higher and the lower forms of religion still coexist side
by side. At one end of the scale, therefore, is the cultured monotheist or the
eclectic pantheist for whom no mysticism is too subtle. Pantheists actually foi:m
a small minority, and the great majority of Hindus are theists believing in one
personal god, though they are at the same time polytheistic in their religious
observances. At the bottom of the scale is a great multitude of people in a low
state of religious development, some of whom have scarcely risen above mere
fetishism. 41

J.H. Hutton, who directed the Census of India in 1931, is more

particular:
If the view be accepted that the Hindu religion has its origin in pre-vedic times
and that in its later form it is the resuh of the reaction by the religion of the
 The Bengal PurlQas I 53
country to the inttusive beliefs of northern invaden, many features of Hinduism
will become at once more comprehensible, while the very striking difference
between the religion of the R igveda and that of the Dharmashastras will seem
narural .... This would explain Hinduism's amalgamation with and absorption
oflocal cults and its excessive multiformity, and is, moreover, in entire accord-
ance with the manner in which it still spreads at the present day, absorbing tribal
religions by virtue of its social prestige, by identification of local gods with its
own, by the experimental resort to Hindu priests, and by the social promotion
of pagan chiefs who are provided with suitable mythological pedigrees. Into the
early Hindu beliefs spread in this manner the religion of the Rigvcda has been
imposed and absorbed.49

Despite the explicit acceptance of the theory of Aryan invasion and

the expression of such value judgements on high and low forms of
religion that offend our understanding and sensibility, both men captured
the essence of Puranic Hinduism with remarkable accuracy. Although
neither of them actually uses the term Purii1Ja, what they have in mind
is evidently the Puranic reli.gion, and in fact, in Huttoo 's description one
can read Purii1Ja for 'the Dllarmashastras' without the slightest difficulty,
particularly when both these categories of texts were the product of the
same process and, in a sense, complementary to each other. Both O'Mal-
ley and Hutton view Puranic Hinduism as the result of continuous
interaction between Vedic and indigenous forms of religion, ever enlarg-
ing its scope and authority. Hence what Hutton calls the 'multiformity'
and O'Malley the 'scale' of Puranic Hinduism.
Puranic Hinduism is thus by its very nature and motivation assimi·
lative rather than exclusive; it accommodates rather than rejects. lrawati
Karve refers to this cultural process when she characterizes 'Indian
society' as a 'culture by accretion'. so According to her, in this scheme
of things, all phenomena are relative and the world of the humans as
well as the gods share in this relativity. Even human values, including
good and evil, have no absolute reality and have equal rights to prevail.
This is the source and justification of the continuous and simultaneous
existence of a multiplicity of behavioural patterns within society. This
conception of the universe is, however, so relativistic that it would lead
to anarchy- to a society in which there could be no agencies of control
other than the individual's own. Yet, Karve points out, Indian society
is governed by fairly strict rules of behaviour, the rules of the castes.
She suggests that the coexistence of groups with different norms of
behaviour and the sll'Uctures goveJDiAg the behaviour of coexisting
caste groups are explained by the theory of Brahman and karma
54 I Religious Process
respectively.51 Brahman, the Absolute, is above all atlributes and thus
all other categories belong to the world of partial truths; and Jeanna, as
the positive and negative worth of human action, is the regulating
agency of such actions of the members of the caste groups. Brahman,
as the origin of all beings, legitimates their right to exist while karma
ensures that they exist within the societal framework of discipline and
authority. Together they impart the requisite stability to the fluid struc-
ture of Puran.ic Hinduism and the society it presupposes.
All the Puriil}as acknowledge the centrality of Brahman and the
caste-oriented social structure governed by Jeanna as axiomatic in their
assimilative synthesis. Within the framework of these given constants
however, each Puriil}a has the manipulative space to choose a divine
manifestation of the Brahman as the Absolute, depending on its sectarian
leaning. Once this preliminary choice has been exercised, it can then
gather around its chosen deity all the diverse elements it seeks to
accommodate. At the philosophical level, a perfect example of the mul-
tiplicity of ~e Puran.ic ideation is evident in this passage from the
Brahmavaivarta, which recognizes ~a as the supreme deity:
The Lord has nine fonns ... . The six schools each assign to him a form, the
VaiH1avas and the Vedas one each, and the l'uriilJllS one also. In this way he
has nine forms. The Nyiiya and Saiikara call him indescribable; the Vaiiqika
calls him eternal; the Siimkhya calls him the luminous everlasting god; the
Mimiimsii calls him the form of all; the Vedanta, the cause of all; the Piitailjala,
the infinite; the Vedas, the essence of truth; the Pur&;ias, self-willed; and the
devotees [visualize him as one] having an external form. 52
Such adjustments in the realm of thought offered justification for ac-
commodation at a more mundane level.
The interesting point about the characterization of the views of the
various schools in this passage is that they are not merely imprecise,
but are also unrelated to the specific philosophical positions of these
different schools. Clearly therefore, these names have been thrown in,
not to suggest that they have literally described ~ in this way, but
to invoke figurative support for the recognition of ~a by the
brahmanical authority. Brahmanical authority accepts ~a; therefore
the entire tradition that it upholds must also accept him, no matter
whether or not they actually do so. This is the principal mechanism of
the Puran.ic process, and nowhere is it more visibly evident than in their
handling of the Vedas as symbolic authority.
Brian K. Smith has identified two criteria as constitutive of
Hinduism: (I) recognition of the authority of the briihmQ1Ja class, and
 Tire Bengal Pwil)8S I SS

(2) recognition of the authority of the Vedas.'3 He says that the briih-
ma~as are recogniud as religious authorities because of their intimate
and special relationship with the Vedas, the authoritative texts of
Hinduism. The briihm~ are the traditional bearers, purveyors, inter-
preters and protectors of the Vedas. The authority of the briih~ is
dependent on the authority of the Veda and the Yeda exists because of
the traditional function the briihmat,la bas assumed for its preservation.
Thus Smith arrives at the following definition of Hinduism: 'Hinduism
is the religion of those humans who create, perpetuate and transform
traditions with legitimizing reference to the authority of the Veda.,,.
The importance of the Vedas in Hinduism arises out of what Heester-
man calls the inescapable need for ultimate authority." The great para-
dox of Hinduism is however that, although the religion is inextricably
tied to the legitimizing authority of the Vedas, in post-Vedic times the
subject matter of the Vedas was and is largely unknown to those who
define themselves in relation to it, primarily because the lost their doc-
trinal and practical relevance to post-Vedic Hinduism. Thus the Vedas
became notional authority and came to represent not the Vedic texts, but
rather the totality of knowledge. And yet they are preserved in all their
pristine sanctity. This paradox is resolved if we look upon the Vedas as
the key to ultimate legitimation whereby even if they are unrelated to
current ways of human life and society, one still bas to come to terms
with them. Smith bas defined Hinduism as a process and not an essence.
Consequently, it is not the essential content of the Vedas that is of
defining import so much as the particular relationships Hindus establish
with them.56
As for Giorgio Bonazzoli here too canon is either a body of works
that are accepted as nonnative, or a complex of rules through which one
can establish whether a work should be accepted as authentic and nor-
mative or not.' 1 The Vedas provide the touchstone by which the briih-
m~ decide which texts should be considered normative and authentic.
In other words, the principal rule for determining canon in Hinduism is
whether or not it is Vedic. The method through which later texts such
as the Smrtis and the Puriil;ras are related to the symbolic authority of
the Vedas bas been characterized by Smith as 'strategies for ortho-
doxy' ." Smith bas observed a variety of such strategies for constituting
post-Vedic texts, doctrines and practices as Vedic: ' Reflection (this is
the Veda); restatement (this is based on the Veda); reduction (this is the
simplified Veda); reproduction (this enlarges the Veda); recapitulation
(this is the condensed essence of the Veda); and even reversal (the Veda
56 I /religious Process
is based on this).' 59 Smith points out that by representing through
deployment of these sttategies, new texts, doctrines and practices as
connected in some way or another to the Vedas, change is both legiti-
mized and denied, and continuity is both affirmed and stretched. 60 It is
immaterial whether the texts, doctrines and practices thus legitimized
have in reality anything to do with the Vedas. Only the assertion of the
briihmlll)as counts.
This is a universal process in the living tradition of Hinduism and
all post-Vedic texts raised to the status of nonnative and authentic have
claimed connection with the Vedas, in one form or another. This method
of seeking legitimation is particularly relevant for those texts which were
composed with a view to introducing brahmanical socio-religious prac-
tices in the peripheral areas where the Vedic tradition had not sufficiently
penetrated, such as the Upapuriil)as of Bengal, especially when, to begin
with, they were combining the functions of both the PuriiTJQS and the
Smrtis in these areas. Thus it can be observed that all the techniques
listed by Smith of establishing Puranic authority as derived from the
Vedas were operative in the Bengal Puriil)as.61 I will cite a selected few
instances from among innumerable such examples.
In the Devibhiigavata, Niiriylll}a, in response to Narada's question
on the sources of dhanna, says that Sruti and Smrti are the two eyes of
the God and Puriil)a is his heart. None other than that which has been
determined in these three is dharma.62 Similarly, in enumerating the
fruits of reading the Riidhiikavaca, typical of the Puriil)as, the
Brahmavaivarta declares that the result obtained from reading this
kavaca is the same as that of studying the four Vedas.63 Even if these
instances cannot be described as exactly what Smith calls 'reflection'
(this is the Veda), they certainly indicate an attempt to establish equiva-
lence between the Vedas and the PuriiTJQS.
But the major thrust of the Bengal Puriil)as was to try and establish
themselves as the 'restatement' (this is based on the Veda). Thus the
Brahmavaivarta says, blessed are the four Vedas because everything is
being done according to their prescription. All the siistras are derived
from the Vedas, and the Puriil)as also reside in them. 64 Instances of
' reduction' (this is the simplified Veda) are also numerous. For example,
the Devibhiigavata siates that in all the manvantaras, in every Dviipara
age, Vedavyisa composes the PuriiTJQS in order to protect dharma. It is
Vi~l]U, in the form of Vyisa, who divides the one Veda into four units
for the good of the universe. Knowing that in the Kali age the briih-
mal)as will be stupid and short-lived and thus unable to understand the
 The Bengal PuriQas I 51
Vedas, he composes the holy Purii1.1a Samhitiis in each manvantara. In
particular, women, Siidras and unworthy briihma1.1as are debarred from
listening to the Vedas. It is for their benefit that the Purii1.1as have been
written.65 The same idea has been indirectly expressed in the Brhan-
niiradiya Puriif.la.61> Here the siita says that there are two ways to salva-
tion, Jcriyiiyoga andjiiiinayoga. The long passage seems to indicate that
only yogis67 can aspire to salvation through the practice of jiiiinayoga
which involves, among other things, reading the Upan~ads and repeated
recitation of pra1.1ava and the twelve-, eight- or five-lettered mantras,61
presumably because this is difficult, while the common people (miinava)
should follow the simpler procedure of lcriyiiyoga which demands fast-
ing, listening to the Purii1.1as, and worshipping Vi~u with flowers. 69 The
implication is that Vedic scholarship, which is evidently superior, is for
the select few and the Puriif.las, which are in consonance with the Vedas
but much simpler, are for the masses.
The Bengal Purii1.1as also claim that they are 'reproduction' (which
enlarges the Vedas). Hence the Devibhiigavata declares that the half
verse signifying the central thesis of the Vedas that the Devi had told
the infant Vi~u lying on the banyan leaf during the great deluge, is the
key to the Devibhiigavata. As an explanation of that key, Brahma com-
posed the BhiigavataJiistra in myriads of verses. Vyasa summarized it
into the Devibhiigavata Puriif)a consisting of eighteen thousand verses
distributed over twelve sections, for the enlightenment of his son Suka.m
But the easiest way of linking the Puranic corpus with the Vedic was to
claim that the Puriif)as were the 'recapitulation' (the condensed essence)
of the Vedas. Thus the Brhanniiradiya begins with the statement that for
the good of the universe Vedavyasa has included the quintessence of the
entire VedavediingaJiistra in the Puriif.las. 71 The Devibhiigava1a ends
with the same assertion and adds that the result of reading and listening
to the Devibhiigavata is comparable to the fruits of reading the Vedas.n
'Reversal' (the Veda is based on this) is an attitude which was difficult
for the Bengal Puriif.las to adopt, because they were operative in an area
where the Vedic authority itself was not sufficiently well grounded and
it would have been presumptuous to challenge it from the point of view
of brahmanical interests which were uppermost in the Purii1.1as. Never-
theless there are certain indirect suggestions, such as in the Brahmavai-
varta, where Piirvati tells ~i:ia that she is known to him, but she does
not know him, for who can know him? Not even the Vedas themselves,
nor the Vedic scholar or interpreter,73 which implies certain superiority
 58 I Religious Process

over the Vedas. But the Bengal Puriil}as have permitted themselves to
go only that far.
On the whole the message of the Bengal Puriil}as is that they are the
confirming elaboration of the Vedas. They enshrine the essential wisdom
of the Vedas but simplify, condense or enlarge it, depending on the needs
of the common people, as perceived by the authors. The attempt was to
forge a legitimating relationship with the Vedas in every possible way.
The briihmalJas composed these Puriil}as, but, though necessary, this in
itself was not sufficient to establish their status as normative texts, unless
they also ensured the sanctifying support of the Vedas. The Bengal
Puriil}as therefore subtly insist on continuity with the Vedic tradition.
Innumerable statements such as 'it is a great secret even in the Vedas
and the Puriirµls' 14 or 'the four Vedas and the Puriil}as assert that
7
. Midhava is submissive to Ridhi', ' are meant to confinn this continuity
through endless repetition. Hazra cites an unidentified passage which
captures the essence of this continuity in its fullest elaboration,
That twice-born (Brihmal}8), who knows the four Vedas with the Aiigas (sup-
plementary sciences) and the Upan~ds, should not be (regarded as) proficient
unless he thoroughly knows the PuriQas. He should reinforce the Vedas with
the Itihisa and the Puril}a. The Vedas is [sic] afraid of him who is deficient in
traditional knowledge (thinking) ' He will hurt me'.76
V.S. Agrawala interprets the crucial line itihiisa-purii1Jiibhyiim vedam
samupabrmhayet as 'the metaphysical truth of the Vedas is intended to
be demonstrated in the ltibiisa-~a manner'. The Brahmavaivarta
77

closely follows this sentiment when it states, ' because of your ignorance
you could not know him who has been ascertained in all the Vedas,
Puriil}as and itihiisa as Sarvesa' .71 The Puriil}as generally, and the Ben-
gal Puriil}as in P.articular, have incorporated nearly everything that they
considered relevant in the name of the Vedas, irrespective of the fact that
often the matter had nothing to do with the 'Vedas or was even contra-
dictory to them. The Purii1Jas scrupulously maintained the cover of the
Vedas not so much out of deference to them (which of course was given)
as for their own survival. Thus when the Brahmavaivarta says that copu-
lation during the day is prohibited in the Vedas,19 it merely signifies that
for some reason the Brahmavaivarta or more precisely the briihmal}as
who were introducing brahmanism to Bengal through Brahmavaivarta
and such other Puriil}as, consider this act to be improper (or at least wish
this to be considered improper) and were merely calling on the Vedas
to witness that they endorse this stricture. We have noted that the Vedas
had lost practical and doctrinal relevance for post-Vedic Hinduism, and
 .~Bengal PwiQaa I 59
reformulations were necessary to match altered cin:umstances. This re-
formulation required the sanction of authority that could be passed off
as legitimate. As Kosambi points out, the Yedas did not help in this, since
'their text was fixed by immutable routine', and no word or syllable
could be changed. 'The Smrtis, once written, could not be tampered with
soon, or the conflict would mean loss of all authority. This meant raising
new texts [the Pura1J.as] to venerable antiquity in some plausible fashion
which would allow revision of their contents.'I()
This process has been characterized by Mackenzie Brown as shift
from sound to image of the holy word. ' Earlier, the "artha tradition"
was subservient to the ":Jabda tradition", the narrative history facilitating
and ensuring the success of the mantra.' In the Pura1J.as, narrative be-
came 'the primary holy word, reincorporating the old mantric tradition
under new terms. The saving story itself has taken on the character of
mantric efficacy though not the mantric immutability.' 11 Once the em-
phasis was on meaning, it was no longer immutable, for meaning is
always contextual and seldom, if ever, universal. Hence it can explain,
absorb, recreate itself, and construct its own meaning, depending on the
necessity. This flexibility of the Pura1J.as helped brahmanism to pene-
trate peripheral areas such as Bengal, and write its own texts to suit its
innovations. Thus R.C. Majumdar is mistaken in assuming a sharp dis-
tinction between the Vedic and Puranic traditions in Bengal when he
claims that the former was 'replaced' by the latter.12 Bengal was never
a stronghold of Vedic religion, and if the Vedas in Majumdar's under-
standing stand for brahmanisrn, this was carried forward in the Bengal
Purarjas in their notional acknowledgement and glorification of the
Vedic authority.

III

Invoking the authority of the Vedas for legitimation is common to all

the Pura1J.as, even if the tendency to eqiphasize the Vedic-Puranic con-
tinuum is somewhat stronger in the Bengal variants, because the need
for such authentication was greater in their c11S1e. The Bengal Pura1J.as
were, however, unique in that they were consistent in their effort to as-
similate local customs in the religious and cultural ethos and practices
they attempted to introduce. We have already noticed that even some of
the MahapurOIJ.as exhibit traces of regional interest,13 but none so com-
pletely and systematically as the Bengal PurlifJ.as which even evolved
general principles to justify such inclusions. It is the continuation of the
60 I Religious Process

same Puranic process of appropriating Vedic authority for Puranic in-

novations that they now extended to local customs, assimilating them
by according them the sanction of brahmanical authority. The ideal
gamut of Puranic Hinduism thus consists of the Vedas at one end of the
continuum and the regional/popular at the other, with the PurOl;las me-
diating between these two poles. This process reached its culmination
in the Bengal PurOl;las.
If the Puriit;1as represent an attempt to extend brahmaoical Hinduism
into areas hitherto beyond its pale and to broaden its base of mass
following, then assimilation of local customs in some form or other
would appear inevitable. Indeed, in the epics, which are justly consid-
ered proto-Puranic, this tendency is already discernible. For example,
the Mahiibhiirata says that the seven illustrious Citra.Sikh.al}<,tin ~is
unanimously proclaimed on Mount Meru an excellent siistra which was
made consistent with the four Vedas and was meant for the populace,
that it consisted of one lakh verses and dealt with the best /okodharma
(religious duties of the people).14 What is called lokadharma in the
Mahiibhiirata gets transfonned into iicara in the more explicit Bengal
Puriit;1as. Aciira is a difficult tenn to translate. According to H.H. Wilson
it means an established rule of conduct or custom or practice,15 but in
this context it also connotes specificity of region, a given time, and
consonance with one's stage and order of life. Thus when the Brhan-
niiradiya Puriil,la declares,
One who dedicates himself to devotion to Harl without transgressing one's own
iiciira, goes to the abode of ~r:iu which is witnessed by the gods. 0 powerful
sage, one who, while performing the duties declared by the Vedas and required
by one's own order of life, engages himself in meditation on Harl, attains final
beatitude. Dharma arises from iiciira and Acyuta is the lord of dharma. Being
worshipped by one who is engaged in [the practice of] iiciira enjoined by one's
stage of life, Harl gives everything. He who, though being a master of Vediinta
together with the airgas, falls from his own iiciira, is known as patita (degraded),
because he is outside (of srauta and smiirtaJ work. He who deviates from iiciira
enjoined by his own order of life, is said to be patita, no matter whether he is
given to devotion to Harl or engaged in meditation on him. 0 best of briihmalJOS,
neither the Vedas, nor devotion to Harl or Mahe5vara can sanctify that con-
founded fool (mu4ha) who has fallen from iiciira. Neither visit to holy places,
nor residence in sacred tirthas, nor performance of various sacrifices saves one
who has discarded iiciira. Heaven is attained by [the practice of] iiciira, happi-
ness is attained by [the practice of] iiciira, and final release is attained by [the
practice of] iiciira. What is not attained by [the practice of] iiciira?16
there is no uncertainty about its meaning. One must perform the duties
 11ie Bengal Purqu I 6 l

prescribed by the Vedas, one must demonstrate unswerving allegiance

to the Puranic high gods Hari or Mahe5vara--because these are
inescapable preconditions of dharma. But even these are ren~
ineffective if one deviates from the prevailing custom, for 'in all
sacred scriptures acara bas the first consideration. Dharma arises from
iiciira.' 17 Thus iiciira receives priority over all the other components
of brahmanism. In these instances of course the term iiciira bas been
used in its widest possible sense which includes the duties of both
iiJrama and v~a.
Yet iiciira also bas specific implications. This is borne out by the use
of such compounds as vrddhiiciira" [practice of old men] in the same
text and the meaning gets even more specific when the text says, 'com-
plete iiciira [is to be practised] in [one's] own village, half [of the same]
on the way (during ajomney), 0 best of sages, there is no rule in limes
of illness as well as in great distress'.'9 Finally, the Purii1.1a sheds all
ambiguity and asserts, 'all the va~as should perform the practices of
the village (griimiiciira) prevailing at the lime, in consonance with the
ways of the Smrtis',90 and again 'the practices of different countries/re-
gions (deiiiciira) should be followed by the people born in those par-
ticular countries. Otherwise one is known as patita and is excluded from
all dharma. ' 91 The IJrhanniiradiya further proclaims, 'correct statement
is truth. Those who are committed to dharma will tell the truth which
is not contrary to dharma. Therefore the statements wise men make after
proper consideration of lime and space (deiakiiliidivijfiiiniit), which are
in conformity with their own dharma, are truth. ' 92 Thus it does not
hesitate to designate even truth context specific and hence conditional,
such is the magnitude of importance placed on the local customs. The
Devibhiigavata virtually equates dharma with iicara. It says, 'the
dharma of the village, of the caste, of the country and of the lineage
bas to be performed by all men. Under no circumstances can these be
disregarded. Men of improper conduct (duriiciiraM are condemned by
the people. ' 91
Apart from iiciira, the other expression the authors of the Bengal
Purii1.1as frequently use is laulcika (literally, popular). The Brhaddharma
apportions dharma into two fundamental categories, vaidika and
laukika.94 At limes they seem to be independent of and yet consistent
with one another: 'adultery in women is prohibited in the Vedas and in
popular custom',9' and at other times they form a continuum: 'as milk
cannot be separated from its whiteness ... similarly there is no distinc-
tion between Ridhi and Midhava in popular perception, in the Vedas
62 I Religious Process

and in the Purii!Jas' .'>6 Significantly, however, the Brahmavaivarta

asserts that in case of a conflict between these two sources of dharma,
popular custom must prevail. When the gods told Pirvatl, 'whatever is
approved in the Vedas is dhanna and anything to the contrary is the
absence of dhanna', Pirvati replied, 'who can decide on the basis of
the Vedas alone? Popular custom is more powerful than the Vedas.
Hence who can renounce popular custom?' 97
Indeed, these Purii!Jas were written and revised over a long period
of time and the contributions of multiple authors (presumably without
absolute unanimity of opinion in every matter of detail), inttoduced
confusing inconsistencies in them. Therefore, a stray statement should
not be taken to represent the central argument of a Purii!Ja. At the same
time, upholding popular custom appears a dominant attitude in these
texts, and since it runs as a continuous thread throughout the corpus of
the Bengal Purii!Jas, its importance in the Puranic scheme of things is
inescapable. Besides, the Vedas constitute such an infallible symbol of
sacred authority for the Puriil)as that if defied (by none other than
Pirvati) even once in favour of popular custom, this significant deviation
should be given due weightage.
An example of how inconspicuously popular custans were incorpo-
rated into brahmanical proceedings will make our meaning clear. In
connection with the preparation of the planting of the lndradhvaja, the
Devi Puriil)Q instructs that amidst the beating of drums, the music of
prostitutes, the blowing of conch-shells and recitation of the Vedas by
briihmQl)as the pole should be taken to its hoisting place. 91 The songs
of prostitutes must have been a popular custom that was retained. So
long as the primary prerequisite of brahmanism-thc recitation of the
Vedas by briihmal)as-was fulfilled, it did not object to entertaining a
prevailing social custom. On the contrary, the practice was conferred
with the seal of brahmanical authority by declaring it a part of the ritual
scheme. It did not upset the Puranic arrangement and at the same time
it assured the local participants of continuity with existing usage, thus
paving the way for an easier acceptance of the innovations the Purii!Jas
were attempting to introduce. This has indeed been the basic technique
of the Purii!Jas in their assimilative process. To select an instance at
random, the Devibhiigavata decrees that after defecation a briihmQl)Q
should cleanse himself with white coloured clay and the other ~as
with red, yellow and black coloured clay respectively, or whatever the
practice in whichever country.99 In the same context it later adds that in
all matters one should act according to considerations of place, time,
 The Bengal PW'il)as I 63

material and capability.100 Purification of the body is an essential re-

quirement Provided that is performed, the Purib)as are willing, even
eager, to accommodate local constraints.

IV

In view of this explicitly stated Puranic need to acknowledge, and if pos-

sible, integrate regional specificities, we are confronted with an impor-
tant question-the status of a normative text, such as the Purib)as, in a
given region. The question assumes greater significance if this happens
to be a region where brahmanism came comparatively late. The question
has been partly addressed, although inadvertently, in a major contem-
porary debate over the interpretation of the Puriil)as, in the journal
Purib)a. The crux of the discussion has been the possibility and signifi-
cance of critical editions of Smrti literature. Since a consensus exists on
later human elaborations of the 'original' Purii1,1asamhitii by Vyisa, the
issue was: if a Puranic text is edited along the lines laid down by
V.S. Sulcthankar for the critical edition of the Mahiibhiirata, will it vio-
late the Puranic spirit? Madeleine Biardeau identified the problem as that
of the application of methods of textual criticism to what is basically an
oral tradition. only subsequently written down. She has pointed out that
in India the spoken word is more revered than the written and more
importantly, a further distinction is drawn between Sruti (revealed lruth
which is immutable) and Smrti (remembered truth which is of human
composition). She believes that the latter acquires nonnative value only
through its acceptance by the briihmaf)as of a particular region and it is
this locus of authority at the local level which, Biardeau feels, Suk-
thankar has betrayed by trying to establish the earliest text and then treat
it as authoritative by virtue of a supposed greater chronological prox-
imity to Vyisa. She claims that this method introduced the historical di-
mension into the realm of myth, where it cannot exist. 101 V.M. Bedekar
defended Sukthankar's method against Biardeau's criticism on the
grounds that once a tradition is committed to writing, it becomes liable
to textual criticism, regardless of its origin and the text so constituted
is simply a text which may or may not have been at sometime some
place accepted as authoritative. 1112 A .S. Gupta joined issue by attempting
to find a compromise between the rigour of Sukthankar and the scepti-
cism of Biardeau. He views the entire Puranic corpus as a whole, the
individual Puriil)as representing its diffm:nl parts. He believes lhal il is
possible to characterize each of these parts according to its sectarian
64 I Religious Process

preference, and that it is then open to an editor to treat additions that

do not accord with that bias as spurious and relegate them to the notes
and appendices. 103
The details of this debate are irrelevant here. Our concern is only an
interesting by-product of this debate, namely th~ mechanism of the
acceptability of such a text in a given region, particularly when such a
region is not exposed to brahmanical influence for a long time. Biardeau
has expressed her opinion unequivocally, ' any epic or puriz}ic story is
true if the local brahmins recognize it as part of their beliefs. These
brahmins .. . warrant the authority of the local tradition because they
are well-versed in the iruJi. And if such a story is recognized by the
brahmins, it is attributed to Vyisa ... regardless of whether it is men-
tioned in any of the classical text[s).' 104 Biardeau illustrates this with an
'
experience she had with the Srivai~va pundits of the Simhacalam
pii{hsiilii in Andhra. These pundits admitted to her that the local Puriilµl,
relating the story of Narasiriiha and Prahlada, was quite different from
the Skanda Purii1;1a version of the same story, though this was its avowed
source. In spite of this difference the local Purii1JQ was authoritative for
them, since it expressed their belief and was therefore considered SUJ»
rior to any other local version. Biardeau believes that there is no con-
tradiction here since the Srivai~ava briihmQ1Jas saw themselves as the
source of the authoritativeness of the story without caring too much for
the origin of the tradition in time or place. ios
For my purpose this is an important point, the significance of which
became lost in the debate. It should be noted, however, that both the
critics of Biardeau concede her this point in passing. Bedekar, for ex-
ample, does not deny that the traditional locus of authority was the local
briihma1;1as who expressed their 'subjective preference' 106 and Gupta
does not dismiss the later revisions as mere corruptions, for he recog-
nizes in them 'the desire on the part of the redactors to revise the text
of the Purai)as from time to time and keep them in line with the current
religious and social ideas of their times' . 107 Thus, an inevitable condition
of the Purii1;1as seems to be to remain 'in line'. But in line with what?
And who was to determine this?
Coburn rightly points out that the intention of the Purii1;1as is to
'make the original Vedic truth available in a contemporarily relevant
way'. 108 Whatever be the connotation of the term 'Vedic truth', I
suggest that the contemporaneity relates both to time and place be-
cause it is absurd to assume homogeneity of tradition for so large
and diverse a country as India, and that too in the early medieval
 The Bengal Puri.Qas I 65
period. How is this conformity with time and place ensured? For an
answer to this it will be useful to recall Mackenzie Brown's discus-
sion on the general role of the Puriil)as in Hindu culture. 109 He argues
that for the Hindus, truth is something to be discovered by each
generation. But it is merely a rediscovery, because truth was fully
revealed in the past by the Vedas. The Puriil)as are primarily a means
of rediscovering and reinterpreting truth. They are also an easier form
of truth adapted to contemporary conditions: 'Implicit in this is the
notion that mankind can accept only a modified fonn of truth. It is
assumed that the revisions are made in complete harmony of the truth
contained in the sruti. The Puri.Qas represent then an interpretation
or clarification of the sruti, revealing the eternal, immutable truth in
a comprehensible fonn to all mankind in his changing, historical
situation.' 110
Since Gupta, Coburn, and Brown are concerned with the 'revisions'
of the Mahiipuriil)as, their explanations naturally rest with changing
needs of the 'time', although Brown is more inclusive in the sense.that
'historical situation' takes into cognizance the specificities of both time
and space. They however establish the principle behind the process
through which new materials were incorporated in the Purii1.1as. How-
ever, in the case of the Upapurii1.1as, which were evidently composed
with a view to transmitting brahmanical religion in the peripheral areas,
the demands of place must have received precedence over the ever-'
changing needs of time. Otherwise it is difficult to explain the repeated·
insistence of the Bengal Puriifµa on the necessity to honour deiiiciira-
lokiicara. Most of these local customs must have been inconsistent with
current brahmanical tradition. And yet the essential features of both of
these had to be retained in some fonn for them..to appear acceptable to
both. Moreover, these customs, even within a region, are hardly ex-
pected to be uniform. Thus, which of the available choices is to be
accepted and incorporated in the local Purii1.1as, with suitable modifica-
tions, if need be?
In matters of such adjustments the judgement of the local briihma1;1as
was crucial. For instance, in the example cited by Biardeau, there was
the brahmanical version of the Narasiriiha-Prahlada story contained in
the Skanda Purii1.1a and many local variants. For the Srivai~r:iava
briihmQf.las, however, the version of the local Purii1;1a was considered
superior to all others. Biardeau explains this in terms of the 'authorita-
tiveness· of the local briihmQ1Jas. 11 ' This is indeed true, but how in the
first place did the version of the local Purii1.1a come about, if it was
66 I Religious Process

neither an exact reproduction of the Skanda Puriil;la, nor in complete

agreement with the many local variants? It is reasonable to speculate
that the particular version adopted or constructed for the local Stha/a-
puriiTµJ was one which carried the fundamentals of brahmanism and at
the same time appeared least disruptive to local traditions, in the assess-
ment of the contemporary local briihmaTµJS. Brown suggests that the
Puranic redactions reflect the 'reality as perceived and interpreted' by
the redactors. 112 This reality of the redactors, or in the case of the Bengal
PuriiTJas the authors and redactors, consisted of mediating and striking
a delicate balance between brahmanical authority and the local beliefs,
customs and traditions. Thus a text could attain normative status in a
given region only if it was approved, and in many instances written and
rewritten, by the local briihmaTµJS and not by an impersonal brahmanical
authority. The Bengal Purii1Jas often repeat epic and Puranic stories, but
this in itself was considered insufficient These had to be retold with a
minor or pronounced revision, depending on the necessity as perceived
by the local briihma1Jas. This is an integral part of the Puranic process
and a particularly important one for the Bengal Puriil;las.
An example will serve to illustrate the foregoing discussion on the
general aspects of this process. The following is an account of the
circumstances leading to the introduction of Puranic religion in
Kimarlipa, as described in the Kiilikii Pura1Ja:
In the olden days, everyone started going to heaven after taking bath in the
water of the river, drinking its water and worshipping the deities in the
mahapi(ha ofKimarilpa. Yama, out of deference to PirvatJ, could not stop them,
nor could he take them to his abode. Yama went to Vidhiti and complained
that people in Klmariipa, after pcrforrni11g the ritual worship, are becoming
attendants of Siva and Kimilchyi; he has no jurisdiction there. Brahmi, Yama
and Vifr;iu went to Siva and requested him to malce an arrangement so that Yama
retains his power over the people of Kimariipa. Siva then asked Ugratiri and
his Gaoas to drive away the people from Kimariipa. When everybody, including
the members of the four va~as and even the briihma1Jas started being thrown
out, the sage Vasi$tha, residing at Sandhyicala, became extremely angry. When
Ugratiri and the Gar;ias came to him he said, ' I am an ascetic and you want to
evict me too? Therefore you, along with the MiilfS, will be worshipped in the
viima way. Since your Pnuluithas, are roaming around in a state of drunkenness
like the Mlecchas, this Kiimarllpa qetra will be treated as Mleccha country. I
am a sage versed in the Vedas. Mahidcva, like an inconsiderate Mleccha, tried
to evict me. He should therefore stay here bedecked with ash and bones, loved
by the Mlecchas. Let the Klmariipa ~etra be infested with the Mlecchas. Till
such time as Vi$1}U himself comes to this place, it·w.ill ren;iain like this. The
•1
 11te Bengal Puril)as I 61
Tantras which express the glory of Klmarilpa will be rare.' Thus in the
Kimariipa pifha-the abode of the gods-die gods became Mlccchas, Ugl'lliri
became vama and Mahidcva came to be devoted to the Mlecchas. Within a
moment Kimariipa became devoid of the Vedas and the four~- Although
after Vifl)U came here, Kimarilpa was released of its curse and began to deliver
the fruits of worship again, gods and men remained oblivious of its former
glory. m

The story admits two significant facts about K.imarilpa: that it was
devoid of the Vedas eod the brahmatJ.as, and here the gods and the
goddesses were worshipped in the left (vama) way, prior to the pene-
tration of brahmanical ideology. The story also reluctantly concedes that
the process ofbrabmanintion was not completely successful in the leod
of the Mleccbas, despite the cleansing presence of V~u. Indeed, the
story is remarkably soft on the prevailing form of religion in Kimariipa,
as it bas no word of censure for a non-bnhmanical mode of worship
which made one defy death, eod it is only through the obviously artifi-
cial device of a curse that the text creates an excuse to condemn the
anti-Vedic religious practice of outcastes and finds a way of introducing
V1'1:1u. This story is symptomatic of the message of the Puriil)as, par-
ticularly for Bengal.
Van Kooij suggests that the long exposition in the Ka/ika Puriil)a
on ceremonies and rites for the worship of the goddess in her various
fonns bas no intrinsic unity and appears to describe multiple cults
and practices current at the time of the final redaction of the text
However, it is possible to draw a distinction between two categories.
One category closely corresponds to the general kind of worship,
called common worship (samiinyapiija) in the text, as practised in
large parts of India not only among the Saktas but also among the
Vai~vas and the Saivas. The second category consists of more
special rituals in which orgiastic rites play a major role and which
were kept outside the bnhmanical sphere for a long time. 114 Since
this first category of common ritual covers by far the greater part of
the fragment on Devi worship, Kooij takes this to be a clear indication
of the author's concern to have the deities of Kimarilpa propitiated
by a cult form concordant with the contemporary brabmanical practice
commonly acceptedm in large parts of India and thus to draw the
frontierland of Kimarilpa into the Hindu fold.
For the second category, dealing with the religious forms of the
heterodox communities, the author/redactor of the Kalika Puriil)a is
willing to allow much less i:oom. This tradition is called the left method,
68 I Religious Process

or viimabhava, that is, unorthodox behaviour characterized by the use

of intoxicating liquors and meat, and the stress laid on sexual inter-
course.' 16 It is centred on the ambivalent goddess Tripura-bhairavi, but
significantly the Kiilika Purii')a recommends this form of worship for
brahmanical deities as well, when they assume their heterodox left
shapes. 117 Although the Kiilika Purii')a sanctions these forms of worship,
it warns that a person practising them does not absolve himself from his
debts to seers, gods, ancestors, men and demons. In other words, it does
not release him from the obligation to carry out the five great sacrifices,
which is the right method or da/qi')abhiiva, i.e., the orthodox beha-
viour.' 11 At the same time the text recognizes that orthodox behaviour
may be an obstacle for an individual who wants to attain final emanci-
pation through the viima method. 119 Moreover, the results that can be
obtained by the left worshipper, such as happiness, political power,
radiating body, ability to seduce women, the freedom to move about
like wind, and the power to control ghosts and wild anirnals 120 are so
alluring that compared to these, absolving one's debts pales into insig-
nificance. Thus it seems that the Kiilika Purii')a does accept the viima-
bhiiva and actually instructs that some goddesses should be honoured
with both forms of worship, 121 although, as Kooij rightly points out, it
remains true to the general tendency of the Purii')as to stimulate the
performance of the brahmanical rituals and sacrifices, and attendance to
the established fonns of social behaviour. 122 Indeed, the text adds that
those who worship through the left method do not fulfil the expectations
of the other gods, while the right worshipper satisfies evecybody. That
is why the right method is the better of the two. 123 •
Juxtaposed in this way the text of the Kiilika Purii')a, as indeed the
entire corpus of the Bengal Purii')as, would appear hopelessly ambiva-
lent. But seen in the context of the strength of local traditions, on the
one hand, and the intentions of the purveyors of brahmanical religion,
on the other, much of these inconsistencies are resolved. The brahmani-
cal religion had to be introduced in eastern India, where strong local
traditions had developed, as unobtrusively as possible and as much in
keeping with the local customs as practicable. In other words, the strat-
egy was necessarily assimilation rather than imposition. Thus while not
a single Mahiipurii')a ever dealt exhaustively or even principally with
Sakti worship, although chapters on the praise and worship of the dif-
ferent forms of Devi are found in some of them, the foremost preoccu-
pation of the Bengal Purii')as is the goddess, primarily because the
 The Bengal PUJiJ)as I 69

tradition of goddes~ worship and its associated customs were so finnly

entrenched in that region.
There was no problem in principle with the goddess cult as such,
because she was already integrated with the brahmanical religion. It was
the near-indispensable corollary of goddess worship, the Tantras, which
was problematic since it disregarded the infallibility of the Vedas and
the authority of the briihma1.1as. Hazra shows that until about the begin-
ning of the ninth century Tantric elements were systematically purged
from Puranic literature and it was only later that the Purii1.1as started
recognizing the Tantras as a result of 'the great spread of Tantricism
among the people', 124 when it was no longer possible to ignore them.
Even here the Bengal Puriilµls had to declare that only those Tantras
could be accepted as authoritative that were not in conflict with the
Vedas. •2s This obviously is a face-saving device, for the Tantras are, by
their very nature, external to the Vedic tradition and all that it signifies.
However, the authors of the Bengal Purii1Jas had to arrive at an inter-
mediate position where accommodation of local customs would not
render the features of brahmanical religion unrecognizable, nor the
assertion of brahmanical religion constitute such a threat to the local
people that they would be persuaded to reject the entire package.
All theories of social change through cultural contact accept this proc-
ess as the primary prerequisite of transmission of new ideas and order.
Strasser and Randall have made a summary of the findings of those who
have studied similar assimilative processes in a number of contexts:
In the receiving system we will always find willingness to accept and to integrate
innovations in varying degrees .... The acceptance or rejection of an innovation
may, first, depend upon congruence conditions. An innovation is congruent if
it shows some similarity to the existing cultural elements with which the public
is confident. Secondly, the acceptance may depend upon the condition of com-
patibility that is met if the innovation fits into the new SO<:ial context without
incurring destructive conscquences. 124
It is precisely to satisfy the conditions of congruence and compati-
bility that the Bengal Purii1Jas laid down separate provisions for separate
groups of people. For instance, the Kiililai Purii1Ja says that the briih-
ma1Ja should offer intoxicating liquor to the goddess only once, and that
too through others, ' while it imposes no such restriction on sUdras. 121
The use of liquor was indispensable in viima worship and had to be
complied with by every member of society. But once this insuperable
requirement is fulfilled, the text is willing to make exceptions for dif-
ferent social groups. Liquor plays no part in the brahmanical da/cyi1.1a
70 I Religious Process

method of worship, recommended by the PurO!µls to be superior to the

vama, and therefore, for the briihmal)ll outsider, the offering of liquor
is merely the ritual acknowledgement of a local practice. But local
siidras were presumably used to this form of worship and hence allowed
to continue with their custom without hasty attempts at pe1emptory
reforms. 12ll The consideration of minimwn compatibility must have
weighed heavily with these PurOl;las, for the Kalika Puriil)a, despite its
obvious preference for the brahmanical way of worship, suggests that
when one goes to a pi(ha outside of one's region one should worship
according to the advice of the local people; that one should follow the
prevailing form of worship in a particular region, such as in Q4ra or in
Piiicila. 129 In other words, the text concedes that so long as the locally
in:evailing form of worship is not brought in line with the brahmal)ical
way, it is safer not to disturb the existing equilibrium.

v
I may add that this process of tacit adjustment with the local traditions
through the mediation of some legitimating authority is in no way
unique to the Bengal Puriil)as. This technique is indeed central to the
complex process of reformulation and transmission of tradition
throughout India in all periods. It is only when the Puriil)as them-
selves became firmly established as the source of brahmanical author-
ity at a later date, that the Vedas were replaced by them as the
effective symbol of mobilization, particularly at the village level. Two
examples illustrate this point
D.F. Pocock, in his study of beliefs and practices in a village in
central Gujarat, has discussed a popular local text in vernacular (as
contrasted with the high-profile texts sucli as the Bengal PurO!µls, which
were written in Sanskrit and contributed to the creation of a regional
great tradition) on ' The Most Glorious History of the Goddess'.130 A
typical book of this genre-thick, cheaply bound, printed in large char-
acters and containing garish illustrations- it lays down the rules for
animal sacrifice by the briihmQl)QS. In Pocock's translation the passage
reads:
Although many Brahmans arc meat eaters they should not openly make flesh
offerings. This prohibition derives from the first part of Kalika-purana. That we
arc to understand that this custom is directed to Kshatriya appears from that
section called Sharadatilaka which says (here follow two lines of Sanskrit which
the editor glosses in Gujarati). Pure Brahman who follow the ordinances should
 The Bt:ngal Purqas I 71
make pure (siilvik) oft'erinp for, when they have ablndoned offerings usocWed
with violence:, these then become for other castes the highat 131
There is no such section as 'Sharadatilaka' in the Kalika PurQi;ia, but
chapter 65 of the PurQi;ia is entitled Saradotantra which of course does
not even' mention the topic of sacrifice. The fact of the matter is that
Guj~ is a predominantly vegetarian area and the anxiety of jhe norm-
setters to en.sure that at least the brahtnal)as do not violate this custom
is understandable. What is interesting is that it invokes a spurious Pu-
ranic authority to enforce this prohibition.
The other example is directly related to Bengal and is a Bengali genre
of local-level texts known as the pancali, composed in simple verse
usually in praise of a local deity. Cheaply produced thin books, these
are read aloud before an audience on an appointed day of the week or
month as a part of a ritual practised by women (vrata-kathii). The book
that I discuss here is an intriguing variant of this genre. 132 Impressively
printed and bol·n d, it spreads over 285 pages which suggests that it was
not meant to be recited in one sitting. Entitled Sri Sri Padma Puriil}.a
vii V11aharir Piil~coli, it begins in typically Puranic Sanskrit with the
dhyona and eulogy of Manasi, GqeS&, the DaSivatira and a number of
goddesses. Then follows the central narrative in Bengali, a retelling of
the Cind Sadigar- Lakhindar story of the popular Manasiimailgala of
medieval Bengal, without innovation but suitably prefaced with stereo-
typed descriptions of the beginning of creation, the origin of the Vedas
and time, the churning of the ocean, the sacrifice of oaqa, the burning
of Madana and so on, episodes utterly irrelevant to the content of the
story and not mentioned at all in the original Manasomailgala. From
the style of the composition and choice of topics and their arrangement,
it is clear that the author of this text was claiming for his work the status
of a Purar;ia. It indeed says as much: ' Says dvija Vamsidlsa that one
who listens to this [text] will be absolved of his sins, as it is stated in
the Y-almiki Puro1,1a .' m
In both these instances the point to note is that either a Puriil}.a is
quoted (whether or not the citation is authentic), or the style of the
Puriif,IOS is simulated (however transparent the simulation may be) to
acquire authority for a local custom or to elevate the status of a primarily
regional deity in the divine hierarchy-a process already initiated by the
Puriil}.as. Besides, in both these instances the Puriil}.as perform the same
fimction for these local-level vernacular texts as the Vedas did for the
Purii!J.as. In orthodox Hinduism ultimate authority does stem from the
Vedas, but in course of transmission, as the Vedas get further removed
72 I Religious Process

frOm the actual context, the symbol of authority gets transformed and,
in the case of Bengal. it is largely appropriated by the epi~Pmaoic
tradition. particularly in what may be termed the secondary elaboratio.n
of bnhmanical authority at the grassroots level, for in the Bengal
Purii1Jas and in the vernacular rendering of the epics, the conditions of
congruence and compatibility with regional requirements were already
partially fulfilled, and their general acceptability paved the way for
greater absorption in the future.

NOTES

I. MA. Mehendale, 'The Puri.Qas', in R.C. Majwndar (ed.), The Classical

Age, Bharatiya Vidya Bhavan, Bombay, 1970, pp. 291- 2.
2. Ibid., pp. 296-7.
3. Ibid., p. 294.
4 . Hazra, Puriif)ic Records, Chapters II to IV are devoted to working out the
chronology of the MahiipurOJ)as.
5. For an example of how Hazra arrived at his dating of the Puranic corpus
as a whole, see ibid., pp. 5-6.
6. Ibid., pp. 6-7.
7. In Appendix I Hazra has drawn up an exhaustive list of the nwnerous
verses quoted by the Nibandha writers from the various Puraf)Q3. Ibid.,
pp. 265-335.
8. kaleniigrahaf)am dr~{Vii purOJ)asya tato nrpa I
vyiisa-riipam aham lq-tvo samhiiriimi yuge yuge II
This is what the fish says to Manu in Matsya Puriif)a, 53. 8-9, cited in
ibid., p. 6, note 38.
9. Hazra provides a glimpse of the assortment of topics dealt with in the
Puriif)as:
glorification of one or more of the sectarian deities like Brahml, Vi$Qu and Siva,
... numerous chapten on new myths and legends, and multifarious topics con-
cerning religion and society, for instance, duties of the different castes and orders
oflifc, sacraments, customs in general, eatables and non-eatables, duties of women,
funeral rites and ceremonies, impurity on birth and death. sins, penances and
expiations, purification of things, names and descriptions of hells, results of good
and bad deeds, .. . pacification of unfavourable planets, donations of various types,
dedication of wells, tanks and gardens, worship, devotional vows .. . places of
pilgrimage, c:onscc:ration of temples and images of gods, initiation, and various
mystic rites and practices.
R.C. Hazra, 'The PuriQa', in S.K. De, U.N. Ghosal, A. D. Pusalkar and
R.C. Hazra (eds), The Cultural Heritage of India, Vol. II, The Rama-
krishna Mission Institute ofCultwre, Calcutta, 1969, pp. 246-7.