Marriage as Communication: An Exploration of

Norms and Narratives in Early India

Kumkum Roy
University of Delhi

Comprehension of a particular act or form of communication requires an

understanding of both context and content, each of which is obviously
specific. Moreover, our task is rendered more difficult by the existence of a
plurality of understandings. As historians exploring communication, we face
the challenge of arriving at perceptions of the past, which we derive in large
measure (insofar as we rely on textual sources) from communicative devices

developed in contexts other than our own, and of communicating these in

our turn. Inherent in the enterprise are problems of methodology and, by

extension, validity. In what follows, I discuss prescriptions/perceptions

pertaining to marriage in order to raise questions and share uncertainties.
Marriage is an institution which receives considerable attention in early
Indian texts. This is implicitly, and occasionally explicitly, related to the
fact that the rites associated with marriage provide an occasion for giving
expression to relationships at several levels-between bride and groom,
within and amongst vamas, and amongst kinsfolk. Marriage is envisaged as
a focal point of a person’s social existence, a point in and around which are

woven concerns of power, status and well-being, both material and spiritual.
While these concerns are clearly long-term, intended to be worked out or
realised through the course and duration of married life, the rituals which
mark the commencement of married life bring them into sharper focus,
providing an occasion for foregrounding what are perceived to be the
essential features or ideal forms of the institution.
The communicative significance of rituals is widely recognised.’ The
enactment of social transactions within a context which is by definition
sacred and often public, is intended to confer a degree of both invulnerability
and sanctity on the desired outcome of the process. Hence it is not
surprising that defining the nature and scope of rituals acquires a certain
urgency, evident, for instance, in the Grhya Sutras.
The Grhya Sutras represent attempts to codify and brahmanise domestic
’
See M. Bloch, Ritual, History and Power, London, 1989, p. 122.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

184

rituals, including rites of passage. They are in prose, and are traditionally
ascribed to sütrakiiras (literally the makers of sütras). They were conceived
as related to but distinct from the srutis or revealed tradition. In terms of

communication, as texts, they are at once exclusive and inclusive. On the

one hand, the knowledge contained in the texts was meant to be transmitted

by and to brahmanas versed in Sanskrit. That is, their contents were

ideally the preserve of a restricted social category, ’a literate male elite.&dquo;
On the other hand, given that the sütras were envisaged as universal codes
governing social and ritual practice, they could acquire relevance only
when the prescriptions they incorporated and validated were communicated
to and enforced by (or for) non-literate, non-brahmanical groups.
Although the process of codification is an expression of the power of the
codifiers, yet its acceptance as such requires reinforcement, and is not
necessarily automatic. Besides, codes cannot afford to be entirely divorced
from practice-they may abstract from, modify, or privilege specific prac-
tices at the cost of others, but they are intrinsically related both to the
practices they recognise as well as to those they invalidate. Thus codes are
at once communicative devices, and such that are embedded in complex
networks of communication.
At another level, codified communication is complicated by the fact that
codifiers and those encompassed by the code belong to distinct social
categories. All were not expected to be familiar with the objectives and
contents of the codes. This meant that mechanisms had to be devised to
translate, enact and, if necessary, enforce. In the process, codes were
evidently subjected to a certain amount of transmutation.
Virtually all the principal Grhya Sutras begin with a discussion on
marriage. Marriage is one of the most significant rites of passage for men,
and possibly the most significant one for women, within brahmanical
prescriptive literature, and the codification of its ritual and its forms attains a
certain centrality in both the Grhya and the Dharma Sutras. Although this
is explicable in terms of the event marking the commencement of a new
phase of independent ritual activity for the male sacrificer/ one can also
argue that the focus on marriage is deliberate. What I am suggesting is that
while marriage was obviously constructed as a turning point of socio-ritual
existence, the fact that this particular rite of passage was invested with a
variety of meanings has to do with its (relatively) autonomous importance
for the codifiers. As we will see, although there were attempts to brahmanise
and by extension homogenise definitions of marriage, there are indications
that the codes, as communicative devices, had their limitations. It appears
that communicative processes were subject to contradictory pulls and
pressures. As a result, the texts, which represent in a sense the outcome of

2
Werner F. Menski, ’Marital expectations as dramatized in Hindu marriage rituals’, in J.
Leslie, ed., Roles and Rituals for Hindu Women, Delhi, 1992, p. 47.
3
For example, R.B. Pandey, Hindu Samsk
ras, Delhi, 1969, p. 153.
ā

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 185

such processes, are characterized by a certain amount of internal tension.

This is evident in spite of the fact that the texts intend to harmonise and
universalise the particular, locating it within a framework which aspires to
be holistic.
The potential for tension contained, if not suppressed, within the brah-
manical framework is often explored at length in non-prescriptive works,
including epic literature. The epics are narratives, but the care with which
they were compiled and transmitted points to their significance as vehicles for
the dissemination of a knowledge of norms. They are, moreover, less
systematically brahmanised.
On turning to the epics and their treatment of the ritual of marriage, we
encounter somewhat different problems, even if we ignore the vexed and
probably insoluble question of chronology. While both the Mahabhdrata
and the Ramdyana contain discussions on marriage, the extent to which
these passages are comparable remains an open question. Besides, the fact
that the depiction of the marriage of Rama and Sita constitutes a part of the
Balakanda, an admittedly late appendage to the text, complicates matters
further. If I nevertheless do venture to discuss the evidence from these
texts, it is because they provide us with alternative (though by no means
mutually consistent) definitions of marriage propagated through easily
accessible texts-texts which could be read, memorised, recited and heard.
In other words, unlike the sütras, the knowledge contained in the epics was
meant to be disseminated amongst a wide, not necessarily literate, populace.
However, and again unlike the sütras, the subject of the epics was not
universal, but was confined to the concerns of royalty. Thus in contrast
with the ritual codes the ideals of the epics were not necessarily meant for
explicit emulation or imitation by all and sundry.
Marriage functions as one of the nodal points with epic narrative. To cite
rather obvious examples, the marriage of Draupadi to the five Pandavas,
and of Sita to Rdma, is central to the unfolding of the plot, providing the
setting for complex problems and their resolutions. The epics describe
rather than define marriage. However, implicit in the descriptions are
definitions. Important in themselves, such definitions acquire an added
significance when compared with the prescriptions incorporated in the
scitras.
Both epics contain, moreover, didactic tracts on marriage. While the
significance of such sections cannot be ignored, as they constituted an
important means of simplifying and disseminating the values embedded in
brahmanical ritual, they are incidental within the framework of the epic
and could at best have served to qualify the more central depictions rather
than controvert them.
Given the deliberate yet different communicative nature of the prescrip-
tive texts and the epics, it is obvious that alternative understandings of
marriage were both created and disseminated. The contents of such alter-
natives, and the implications of their co-existence, demand attention.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

186

II

The brahmanisation of marriage, as retlected in the Grhya Sutras, was by

no means simple. What we find is an occasionally uneasy and not necessarily
uniform welding together of diverse beliefs and practices, and a tacit
recognition of the existence of alternatives which are not dignified through
incorporation within the code.’
The modification of marriage necessitated converting the occasion into a
public, visible enactment of a ritual which was ideally regulated. Hence the
insistence on performing the ritual in an outer hall or bahih gala,’ a space
accessible from both within and without the house.
The ritual itself can be viewed as an assemblage of analogies, based on
symbols whose meanings are often explicitly worked out. Thus, at one
level, marriage is constructed as a symbol of growth and increasing pros-
perity. This is underscored by the insistence on its being held at a specific
time-when the sun is in the northern hemisphere, during the fortnight
when the moon waxes, and under auspicious asterisms.~ó
Yet, intrinsic to this definition of prosperity is a notion of appropri-
ation-in other words, access to the well-being which is supposed to be
engendered through the ritual is not undifferentiated. What runs through
the ritual as a recurrent theme is the necessity of absorbing, incorporating
and transforming the woman, rendering her suitable for the task of ensuring
rather than enjoying prosperity. Consequently it is the attributes of the
bride rather than those of the bridegroom that receive attention. For
instance, while there are occasional references to choosing an intelligent
groom, the qualities expected of the bride were both more specific and
numerous; ideally she was to be intelligent, good looking, virtuous and free
from disease.’ Given that such attributes were difficult to discern, tests
x
were devised to simplify matters.
To an extent, the sequence of events which are worked into the ritualis-
ation of marriage develop these relations of prosperity. While there are
4
For instance, Ā
yana Grhya S
ā
śval tra (AGS) I.7.1, which enjoins the observance of jana-
u
padadharma and gr madharma at marriage and proceeds to delineate the common features of
ā
the ritual. The P
&
raskara
amacr; Grhya S tra (PGS) I.8.13 goes a step further in citing a ś
u ruti (Vedic
source untraceable), suggesting that on such occasions the gr
ma pram
ā nam, the authority of
ā
the village, should be followed, H. Oldenberg, trans., The Grhya S tras, Part I, Delhi, 1964;
u
N.N. Sharma, ed., The Āś
yana Grhya S
ā
val tra, Delhi, 1976; K.S. Ksemaraj, ed., P
u &
raskara
amacr;
Grhya S tra, Bombay, 1929.
u
5
yana
ā
ānkh
Ś Grhya S tra (SGS) I.5.2., PGS I.4.2, Oldenberg, The Grhya S
u tras, Part I;
u
S.R. Sehgal, The Śā
yana
ā
nkh tra, Delhi, 1960.
Grhya S
u
6
SGS I.5.5; AGS I.4.1. The latter text also recognizes that marriage may be performed at
any time according to some teachers (AGS 1.4.2).
7
For example, AGS I.5.2 and I.5.4. Occasionally the list of women to be avoided could be
rather long (e.g., Apastamba Grhya S tra [ApGS] I.3.11-14, Oldenberg, The Grhya S
u tras,
u
Part I; A.C. Sastri, ed., The Apastamba Grhya S tra, Banaras, 1928).
u
8 AGS 1.5.4-6; ApGS I.3.15-19.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 187

variations, the initiative is ideally to be taken by the bridegroom’s supporters

or kinsmen, the wooers or varakas.’ In one of the more elaborate accounts
of their roles,&dquo;’ the varakas were expected to strengthen their case by both
reciting the names of the gotra and bringing symbols of fecundity, including
flowers, fruits, barley and the water pot. These were to be placed on the
head of the bride by the preceptor as a means of endowing her with
offspring, cattle, energy and lustre. Thus the roles of kinsmen and priest-
hood were envisaged as complementary-while the former provided the
social and material prerequisites for the transaction, the latter staked
claims to sanctifying and in a sense completing the preliminaries. The
conjoined social and spiritual forces are, moreover, brought to bear on the
woman.
The distinction between bridegroom and bride is sharpened during the
subsequent phase of preparation. Here, the groom is expected to function
as an agent-setting up the sacrificial fire, invoking the gods, and worship-

ping his patrilineal ancestors.&dquo; He was also expected to pray to Agni as

janitr or progenitor, Soma as janimat (endowed with wives) and Pusan as
jnatimat (endowed with kinsfolk),’2 so as to attain identical attributes and
the specific woman in question. The bride, in contrast, is acted upon:
bathed, anointed, and dressed in new clothes.&dquo; In other words, the groom
is constructed as aligned with and capable of invoking ritual and social
authority, whereas attributes have to be conferred on the bride. This is
often conceived of in terms of a symbolism which is both literal and visible.
In one instance&dquo; the wearing of a new dress by the bride is symbolic of her
being endowed with wealth and sons.&dquo;
The crux of the ritual evidently centred on uniting two persona who were
construed as different and empowered, but not in identical ways. Hence
the bonds envisaged were not amongst equals, but had to be worked out in
terms of control, dominance and subordination. This is in a sense epitomised
in the grasping of the hand (panigrahana) which constituted the central
ritual episode. The groom was expected to hold the right hand of the bride,
different ways of doing this being advocated depending on the kind of
offspring desired.’&dquo; In other words, seizing the hand was viewed as symbolic
of an appropriation of the procreative powers of the woman, this being

9 For example, SGS I.6.1 ff; ApGS 1.2.16.

10 SGS I.6.1 ff.
11 SGS I.7.9-12.
12
Ibid.
13 SGS 1.11.1-2; PGS I.4.13; Khdira Grhya S
ā tra
u (KGS) I.3.6, Oldenberg, The Grhya
u
S
t ras, Part I.
14
PGS I.4.13.
15
As Apte observes, ’It appears that the bride is more or less a silent and passive person
during the marriage ritual.’ Usha M. Apte, The Sacrament of Marriage in Hindu Society ( from
the Vedic Period to the Dharmasastras
), Delhi, 1978, p. 126.
16
For example, ApGS II.4.11 ff; AGS I.7.3 ff.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

188

occasionally sealed by the use of a mantra equating the groom with heaven
and the bride with earth and, perhaps more telling, the former with the
composite säman or hymn and the latter with the component rcs or verse,
which is encompassed within the hymn and an integral part of it.&dquo; This was
reinforced by instructing the wife to be anuvrata or devoted
Occasionally the message of the panigrahana could be underscored by
formulae such as the jaya, implying an equation between the bride and
categories such as thought, intention and the new moon, while the groom
was equated with powers of thinking, intending and the full moon. This
was in turn mythicised, with the situation of the groom being conceived of
as parallel to that of Indra vis-A-vis the vis or common people.’9 Such

analogies located and envisaged the asymmetry between groom and bride
at a number of levels of abstraction, and their wide range, proclaimed in a
sacral context, may have made them appear inviolable. At another level,
the panigrahana or the recitation of the mantra could be accompanied by
the circumambulation of the sacred fire&dquo;’ which was probably conceived as
sanctifying, rendering visible and even setting in motion the marital bond.
Also integral to the ritual was the mounting of the stone. Once again, the
initiative in this situation rested with the groom who was expected to make
the bride tread on a stone in the hope that she would, like the stone,
become firm, and a destroyer of enemies.2I We find here an attempt to
both endow the bride with a specific kind of strength and, at the same time,
channelise its use. What is interesting is that the definition of strength in
this context appears to extend beyond the more common attribute of
procreative ability.
The one rite in which the bride was expected to participate more or less
actively was the lfjahoma or the offering of grain, an offering which
symbolised her transfer from one household to another. Here, the material
for the sacrifice was provided by her father or brother, and in the process
of making the offering, the bonds with her natal kin were acknowledged
but sundered. This was accompanied by a plea for the creation of new
bonds with her husband.&dquo; While at one level the ritual evidently conceded
a degree of recognition to the bride as an agent, her agency was perceived
in a specific and delimited context in which she was expected to actively
participate in the transfer of her self from the control of one set of kinsmen
to another.
The transfer of the bride was symbolised by the journey comprising

17
For example, SGS I.13.4.
18
In another instance the mantra concludes with a prayer to obtain offspring and a happy
conjugal life lasting a hundred autumns (AGS 1.7.6).
19 PGS 1.5.9.
20
AGS 1.7.6.
21
SGS I.13.12; PGS I.7.1; AGS I.7.7; ApGS II.5.2; KGS 1.3.19.
22
SGS I.13.15, I.14.1; AGS I.7.13; KGS I.3.20-23.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 189

the seven steps, the saptapadi, which marked the transition as particu-
larly auspicious and as a means of acquiring prosperity implicit in the
bride.23 Very often this preceded an actual journey to ward off various
hazards. 11
The rituals so far outlined thus envisaged the constitution of a particular
kind of bride, and her transfer from one domestic context to another.
However,, this was only deemed complete when the bride was incorporated
within her new, household, a process which was also substantially ritualised.
The couple was expected to observe silence till such time as they saw the
pole star or the asterism arundhatil5 in the hope that the values ascribed to
these stars, constancy and devotion, would be manifest in the new bride.
Finally, rites were recommended to either appease or destroy certain
attributes inherent in the bride which were thought to be threatening to the
model of domesticity which was advocated
If the ritual sequence established and marked out the nature of the
relationship between bride and groom, its enactment, involving the parti-
cipation of kinsfolk and priests, blurred the distinction between the domestic
and the public. In that sense, it provided a definition of marriage which was
located strategically, bridging the home and the ’public’ domain and sym-
bolising their unity. It also permitted the permeation of values from one
’

domain to the other.

The latter process is perhaps most obvious in the prescribed participation of
the priesthood,&dquo; representing an explicit understanding that the ideology
of varna provided the context for a ’true’ marriage. Insofar as marriages
corresponded to this model, they would suggest both an acceptance and a
reinforcement of the values implicit in varf)a. At the same time, the fact
that the participation of kinsfolk was viewed as equally essential> meant a
degree of public recognition on alternative, non-varf)a bonds, which, while
not necessarily egalitarian, would have rested on different principles of
hierarchy. Thus, while the involvement of the priesthood and of kinsfolk
was by no means equivalent, their co-existence and conjunction in the

ritual meant that alternative definitions of power relations would have

been rendered visible almost simultaneously. Although these could be
viewed as reinforcing one another when they were dramatised in such
contexts, they could also be viewed as competing. From this perspective,
23
SGS 1.14.6; AGS I.7.20; PGS I.8.1; KGS I.3.26. Incidentally, the mantra used in this
context was derived from the soma sacrifice, where it was chanted while leading the cow
(symbolic of prosperity) which was ritually exchanged for the ’purchase’ of
soma, Apte, The
Sacrament of Marriage, p. 104. The implicit equation of the bride with the cow would have
reinforced the understanding of the woman as an instrument of prosperity.
24
SGS I.15; ApGS II.5.21.
25 AGS I.7.22; PGS I.8.19; ApGS II.6.12; KGS I.4.4.
26 SGS I.16.3-4; PGS I.11.2; KGS I.4.12.
27 KGS 1.3.5.
28
For example, the brother of the bride had to be present at the jahoma—
ā I.3.20.
l
KGS

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

190

the recognition of the participation of kinsfolk as an integral part of

marriage in normative texts may represent an attempt to contain and
appropriate the power inherent in familial ties. Yet, the effort was vulner-
able and the very enactment of the normative pattern may have rendered it
open to question.
To some extent, such tensions surface within the normative literature
itself, as, for instance, in the treatment of the material transactions
which accompanied marriage. Some of these were ritualised. For instance,
the giving of new clothes to the bride was to be accompanied by the
chanting of mantras. 29 At the same time, gifts to priests are mentioned&dquo;
without being subject to elaborate ritualisation, presumably because there
were other exchanges which constituted a more significant element on such

occasions, and which did not involve the priesthood.&dquo;

In other cases, practices were tacitly accepted without being codified or
ritualised. Thus the practices devised by women could be adopted.&dquo; Avail-
able sources however do not indicate whether such norms coincided with,
or ran counter to, the rituals which were brahmanised. It is the fact that

they were incorporated but not sanctified that leads me to conjecture that
they were both too widely prevalent to be ignored and yet not entirely
amenable to the process of brahmanisation. This possibility is strengthened
if one examines occasional references to such alternative+-in one instance,
four or eight women who were not widows were to be fed with vegetables,
liquor and food, and were expected to dance four times.’3 While the
insistence on an even number of participants and their being in the ideal
married state harmonises with the analogical tenor of the brahmanise
ritual, their propitiation with liquor and the permission to dance are clearly
non-brahmanical. The simultaneous enactment of such alternatives could
pose a threat to the homogenising tendency implicit in codification.&dquo;
The problem of reconciling divergent possibilities is evident at another
level as well. While marriage is dealt with in all the Grhya Sutras, both the
order and the number of component rituals vary substantially (see Table
1). In other words, the normative tradition was itself contested between
authorities, implicitly if not explicitly.
Also interesting is the use of chants in the ritual. The mantras and their
possible origin have been examined at some length by Apte.35 Her detailed
29
KGS. I.3.6.
30 For example, PGS I.8.14 ff.
31 This possibility is explored below.
32
ApGS I.2.15.
33 SGS I.11.5.
34
In fact, as W.F. Menski observes, the grhya rituals reveal ’considerable tension between
priestly experts and their claims on the one hand and the expectations of the families
concerned and of the individuals within those families on the other.’ ’Marital expectations’,
p.50.
35
Apte, The Sacrament of Marriage.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 191

Table 1
Number and Sequence of Marriage Rituals

Note: The data presented are selective, based on five principal Grhya Sutras. Besides, they
exclude a range of minor ritual variations which would complicate the picture even
further. See for instance B. Sarasvati, Brahmanical Ritual Traditions in the Crucible of
Time, Simla, 1977, pp. 172-74. The number indicates the order in which rituals were to
be performed.

study indicates that chants were derived from disparate sources (both in
terms of texts and contexts, in the sense that verses not specific to marriage
were often incorporated in the ritual). The treatment of the famous vivaha
sükta (literally marriage hymn) is a case in point. This hymn of the Rg Veda
(X. 85) is generally recognised as composite and relatively late,36 although
parts of it may be early. Somewhat curiously, the composition or ’perception’
of the hymn is ascribed to Surya Savitn, a feminine mythical character,
commonly identified as the daughter of the sun. While the ascription of
hymns to mythical personages is not altogether uncommon in the Rg Veda,
what is unusual in this particular instance is the reference to Surya in the
third person whenever she is mentioned in the hymn. The ascribed author-
ship also presents a contrast with the ritualisation of the bride which
envisaged her as passive, as noted earlier. The uneasy co-existence between
such passivity and the dynamism implicit in ’seeing’ hymns, or even individual
chants, is obvious. That such incongruities were tacitly accepted, points to
the strengths of alternatives. In this case the overarching concern with
constructing the bride in terms of passivity was at least partly offset by the
necessity of paying lip-service to the possibility of the chants being ’dis-
covered’ by a mythical woman seer. 37
At the same time, the vivdha sükta was literally pulled apart in the ritual
context. Of the 47 verses which constituted the sükta, almost half (i.e., 23)
36
Ibid., p. 4.
37
Menski, ’Marital expectations’, suggests this may be an example of priestly appropriation
of a non-priestly tradition.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

192

do not seem to have been recommended for use in any of the Grhya
Sütras.3x What is more, the sequence in which the remaining verses were
used in the course of the ritual varies substantially from the order in which
they are arranged in the hymn. For instance, the use of verses 47, 45 and 46
(the closing verses) precedes verses 12, 16, 10, etc. In other words, the
logic which underlay the original compilation of the verses in the scikta was
questioned and substantially modified. Given the sanctity attached to or
claimed for the revealed tradition, such reworkings point to the existence
of a considerable pressure to adapt and reformulate the tradition, even if it
meant countenancing obvious discrepancies.
While the problem posed by such variants was probably resolved by and
for brähmwzas v~ho could claim affiliation with one or the other of the
numerous Vedic schools,39 such resolutions were possibly less easily imposed
on non-brihmavas. The latter may have exercised some option of selecting
from the code or even exerting pressure to modify it. The extent to which
such pressures posed limitations on the process of codification are unfortu-
nately difficult to assess. What is none-the-less evident is that communicative
processes were by no means one way-even so-called sacred ritual norms
could not be imposed from above by the priesthood, but were subject to
revisions by those who were technically considered to be actors rather than
directors in the ritual drama.

III

Marriage was conceptualised and dramatised in terms other than ritual or

sacral. This is evident in the depiction and definition of marriage in the epics.
We had seen how the ritual texts concentrate almost exclusively on devel-
oping the nature of the bond between the bride and the groom. While the
connections between marriage and vama and/or ties amongst kinsmen who
are brought together or separated through the ritual are occasionally
touched on, this was by no means a central concern. As opposed to this,
the epics shift the focus of discussion in a number of directions-although
the conjugal tie receives attention, it is embedded in a range of more or less
related themes-relationship amongst kinsfolk, amongst ksatriyas and
amongst varnas, especially between ksatriyas and brfhmapas. In other
words, the meanings ascribed to marriage in the epics are both broader and
different. There is, moreover, a marked reversal of roles, with the priests
being perceived/portrayed as participants rather than as in control of the
proceedings.
The focus on the participation of kinsfolk in the ritual and the negoti-
ations preceding and succeeding marriage both elaborate on and modify
38
This is based on Apte, The Sacrament of Marriage.
39
Sarasvati, Brahmanical Ritual Traditions, p. 177.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 193

the pattern which emerges from prescriptive literature. The role ascribed
to the mother figure Kunti in the marriage of the Pandavas is illustrative.
The prescriptive literature does not envisage a role for the mother of the
bridegroom. But Kunti is instrumental in suggesting the crucial visit to the
kingdom of Pancala and, certainly more important, in ensuring that her
sons enjoy the bride collectively.&dquo; Besides, she trains Draupadi in the

norms of domesticity4’ and after the conclusion of the ritual, blesses her

daughter-in-law in terms somewhat reminiscent of the mantras prescribed

in the brahmanised ritual. 42 In other words, Kunti’s role is envisaged as
decisive and almost priestly.
Vydsa, the paternal grandfather of the Pandavas through niyoga, is
assigned an almost identical role, buttressing the marriage at the point
where it most obviously transgresses prescription, in envisaging polyandry.
For this he invokes the virtually incontestable logic of past births and
actions.a; Given that Vydsa’s connection with the patrilineage of the Panda-
vas was ambiguous in both biological and social terms, his decisive inter-
vention represents an alternative to the more typical invocation of the
pitrs, commonly prescribed in rituals.
The marriage of the Pandavas was also construed as an occasion for
defeating, depriving and humiliating their rival kinsfolk, their Kaurava
cousins and their supporters. Implicit in this is an understanding of kins-
men as potential or actual competitors for brides-a theme undeveloped in
the normative tradition.
At the same time, the marriage was envisaged as a means of acquiring
kinsfolk. In fact, a recurrent refrain through the narrative is the empower-
ment of the Pandavas through their association with Drupada, their father-
in-Iaw.44 Once again, this presents a contrast with the brahmanical norm,
which focused more centrally on the transfer of the bride than on the
relations forged between affinal kin.
As an extension of this, marriage was conceived of as an occasion for
hierarchising relations amongst ksatriyas in general at two levels. On the
one hand, the Pandavas prove their superior prowess both through Arjuna’s
skill in archery&dquo; in the actual contest for winning Draupadi’s hand and in
the fight with rival ksatriyas, galya and KarI)a.46 At another level, Drupada

40
rata I. 156. 6;
bh
ā
Mah rata: The Adi
I.182.2-3, V.S. Sukthankar, ed., The Mah
bh
ā
Parvan, Poona, 1993.
41
Ibid., I.184.4-6.
42
Ibid., I.191.5-9. For instance, she hopes Draupadī will be like Indrānī, Rohini, Dama-
yanti, Bhadrā, Arundhati and Laksmī, give birth to sons, enjoy happiness, act as a yajñapatn
ī
(mistress of the sacrifice and, by extension, wife of the sacrificer or yajam ) and be devoted
na
ā
anuvrat
(
ā).
43
Mah 1. 157.6 ff, I.189.27.
ta
bhar
ā
44
Ibid., I.192.21.
45
Ibid., I.178.18.
46
Ibid., I.181. In the latter case the identities of ksgatriya and kinsmen coalesce, although
the protagonists are not aware of this.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

194

demonstrates his status through creating an impressive backdrop for the

contest for his daughter’s hand, with high walls, moats, archways, all richly
adorned.&dquo; This assertion of status is reinforced after the svayamvara, when
rulers who challenge the decision are defeated.
The relations envisaged between brahmanas and ksatriyas are slightly
more complicated. While the epic accepts a certain role for the priesthood
in the negotiation and completion of marriage, the priests are, in a sense,
marginal to the proceedings. Thus on the advice of a gandharva the
Pandavas appoint a priest, Dhaumya, for the occasion48 and there is mention
of a priest of the Somakas (the lunar line, incidentally not a gotra name)
officiating for Drupada,°9 but these are shadowy figures. Hence, although
their presence may mark a concession to priestly pretensions, they were
not depicted as directing or even guiding the proceedings.
As interesting a comment on the relative position of the brahmana and
the ksatriya is offered through the disguise adopted by the Pandavas. Their
being camouflaged as brihmavas affords them a certain degree of protec-
tion-they can pass for spectators rather than direct contestants in the
svayamvara.50 Arjuna’s participation and success at the critical juncture
then evokes discussion-is it justified for brahmanas to participate in such
contests? The issue is resolved partly through the use of arms and hence
through the logic of victory. It also turns out to be much ado about
nothing, for the Pandavas are in fact ksatriyas and hence legitimate contest-
ants. However, the means of arriving at the conclusion suggest that not
only are the Pandavas the best ksatriyas, but that k$atriyas are implicitly
superior to brahmanas as well.51
To what extent do the epics focus on the rituals of marriage? The answer
is, unfortunately very little. When we occasionally have details, they may
run counter to the prescribed pattern. For instance, there is no question of

testing or choosing the bride. Instead, it is the potential groom who is

subject to a test. Apart from the fact that this is a typical svayamvara, the
selection of five husbands is portrayed as partly Draupadi’s own doing-
her penance in a previous birth and her praying for a husband five times
bear fruit in this unique situation. 52 We also find that while the grooms
prepare and are prepared for the occasion, this involves little ritual activity,
instead is the acquisition of knowledge of their ancestors and of the Pancalas,
some kind of family or kin history, as well as a certain amount of instruc-
tion in the use of weapons and social norms.&dquo;
The contest itself is initiated by Dhr~tadyumna, who displays his sister,

47
I.176.16
Ibid., ff.
48
Ibid., I.174.2.
I.176.31.
Ibid.,
49
I.175.
Ibid.,
50
51
The point is probably reinforced by the narration of the long and complicated story of
Viśvāmitra and Vāsistha, which forms part of the prologue to marriage in both epics.
52
habh
&
M
rata,
ā
amacr; I.157.
53
Ibid. , I.153 ff.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 195

announces the terms and conditions as well as introduces the participants

in terms of name, gotra and achievements What is more, it is Arjuna’s
victory in the contest which transforms Draupadi into his wife or patnf, and
she consequently follows him out of the arena.&dquo;
It is only subsequently, almost as an afterthought, that Drupada sends
his priest to enquire about the Pandavas’ ancestry.56 But he is circum-
vented by a messenger who is better informed, thanks to Dhrstadyumna’s
spying, and announces arrangements for the wedding even before the
priestly negotiator has completed his task.5’ The participants at the
wedding feast include kings, ministers, princes, friends and attendants,&dquo;
amongst whom Yudhisthira introduces himself, declaring Drupada to be
his superior.59
The question of the panigrahana of Draupadi proves to be knotty. Can
five men take the hand of a single woman? As mentioned earlier, Vydsa’s
intervention at this point is crucial.’ The ritual itself is dismissed in a
couple of verses which refer to the offering of oblations with mantras, and
the circumambulation of the fire while holding hands,~’ the significant
point being that Draupadi was restored to virginity on each occasion.&dquo;
Rather more attention is lavished on the gifts Drupada gives, including
chariots with gold trappings, elephants, slave women, clothes, ornaments.&dquo;
Besides, Krsna sends ornaments, cloth, blankets, beds, seats, servants,
elephants, horse-drawn chariots, and gold,’ presumably in his capacity of
kinsman. Other kinsmen, the Kauravas, decide after much deliberation to
accept the situation, with the installation of the Pandavas in Khandava-
prasthaó5 marking, in a sense, the termination of the ritual.
The discussion on the wedding of Rama and Sitd in the Rfmfyava is
relatively superficial, but somewhat similar. The role of kinsmen is import-
ant, but not as central to the proceedings as in the case of the Mahabharata.
In fact, more in tune with ritual norms, it is Janaka, the bride’s father, who
takes the initiative, although the seer Vigvdmitra (an outsider from the
perspective of both kinship ties and ritual expertise) is certainly more
instrumental in bringing the eligible young men to Mithild, the bride’s natal
homes

54 1. 176.33 ff.
Ibid.,
I.179.23.
Ibid.,
55
Ibid., I.185.14.
56
Ibid., I.185.28, 186.1.
57
Ibid., I.186.11.
58
Ibid., I.187.11.
59
60 I.187.32, 188, 189.
Ibid.,
61
Ibid., I.190. 11-12.
62
Ibid., I.190.17.
63
Ibid., I.190.15-18.
64
Ibid., 1.191. 13-18.
Ibid., I.199.25.
65
66
&
m
R amacr; I.30. 13, G.K. Bhat, ed., The lak
yana ā Baroda, 1960.
B
nda,
ā

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

196

The themes of rivalry amongst ksatriyas and between ksatriyas and

brahmanas are present, but muted. Janaka himself claims to defeat dis-
gruntled kings who attack his city, frustrated in their attempts to string the
sacred bowó7 (the test he has devised for selecting his son-in-law). While
Rama succeeds, his triumph is not located in the centre of an immense
gathering of royal personages. It is more private. There is an implicit
contestation of the ritual claim of brahmanas in Janaka’s recitation of his
own genealogy’ as this contrasts with the priest Vdsistha’s recitation of the

lksvdku genealogyb9 on behalf of the groom’s lineage. However, this does

not develop into an open confrontation.
In terms of depicting the enactment of the ritual, the Ramayana is only
slightly better than the Mahäbhärata. Once again, priests of both royal
families are present. 70 Dasaratha offers worship to his ancestors&dquo; and
makes gifts of a hundred thousand cows, with gilded horns, for each of
his sons. We also have a reference to the päf)igrahaf)a72 and the circum-
ambulation, not only of the fire, but of the kings and sages as well.&dquo; In
other words, while the ritual process was described, the description was
modified to make space for alternatives.
While it is clear that marriage in the epics varies substantially from its
ritualised versions, the significance of the existence of such alternatives is
more difficult to assess. The concept of the svayamvara, for instance, is
rather intriguing. Although it represents a typical marital form for rulers in
literature, it does not seem to have been practised by rulers. Is it then a
literary device representing an idealisation of royal marital relations, an
expression of the conflicts inherent in forging marital alliances by and
between ruling families where questions of status and hierarchy could not
be resolved in terms of varna?
Related to this is the preoccupation with constructing a relatively different
dramatic pattern- the epics do not focus on sacred precept/practice (chants,
sequences of rites, the use of ritual objects). Instead there is far greater
concern with the participants in terms of support, power, even conflict.
The definition of marriage thus offered is in fact socialised rather than
sacralised.
What is as important is that the socialisation of the occasion did not
necessarily imply a reiteration of the social concerns voiced in and through
the brahmanical ritual texts. As we have seen, the claims to priestly control

67
Ibid., I.65, 21 ff.
68
Ibid., I.71. 1 ff.
69
Ibid., I.69. 17 ff.
70
Ibid., I.69.14.
71
Ibid., I.71.20.
72
Ibid., I.72.17.
73
Ibid., I.72.23.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

 197

on such occasions were contested by highlighting the roles of non-priestly

protagonists, including kinsmen and women. While the presence of such
personnel was recognised even within the brahmanised version of the
ritual, centralising their roles, as happens in the epic narratives, would
represent an implicit if not explicit questioning of attempts to convey

messages of varna relations through marriage rituals.

We have seen that prescriptive texts were primarily if not solely con-
cerned with defining marriage as an occasion for domesticating women.
Rejecting or even challenging the brahmanical model would then have
implied recognising a relatively autonomous position for women, or alter-
natively, would have required the construction of other means of appropri-
ation or control.
As noted earlier, the brahmanical model of the marriage ritual contains
internal indication of stress or tension. I had suggested that these were the
result of attempts to cope with pressures or threats posed by alternative
definitions, which may have been formulated in different terms, from
different perspectives. The epic model of marriage probably represented
one such alternative-an alternative which was widely disseminated through
the use of an accessible language, simple verses and a compelling narrative.
Although the epics and brahmanical prescriptive literature intended to
exclude and include audiences from different perspectives, their communi-
cative value and success in communication are obvious. Nevertheless, this
does not seem to have resulted in a simple acceptance of either model of
the marriage ritual. What is likely is that there would have been substantial
variations in ritual practices as people could have drawn on, modified,
adapted or even rejected elements from available alternatives. In other
words, there was possibly no absolute identity between communications
about marriage and communications through the enactment of marriage
rituals. What I am suggesting is that the latter could thus become an at least
partly open-ended locus of communication. For instance, the solemn ritual
elements could be and possibly were overlaid with other activities-gift-
exchanges, the participation of more or less people, belonging to identical
or different social strata, feasting, merry making and so on. The precise

importance of any or all of these activities could have varied substantially.

In such a situation, the codified ritual norm, while not explicitly violated,
could be marginalised. This in turn may have generated pressures to
circumscribe the occasion through further codification. Yet, as we have
seen, the process of codification evidently remained incomplete.
It is thus evident that e.ven forms of communication which were apparently
unidirectional, as in fact was the case with the normative texts, and, to a
lesser extent, with the epics, could be subjected to revisions in day-to-day
situations. Such revisions may not have necessarily involved a literal modi-
fication of the texts (although, as is well-known, this was by no means

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015

198

uncommon). What was in all likelihood more common were unwritten (but
not necessarily uncommunicated) challenges. What is more, such contest-
ations could be visible in a very literal sense, and it is likely that they
rendered the literate enterprise somewhat fragile. Freezing texts once and
for all in a written form would have probably made such documents even
more vulnerable, as they could no longer be adapted to changing or
different social milieus. In such a situation of fluidity and flexibility, the
development of both messages and means of communication would have
been a complex process. In coming to terms with the dynamics of such
situations we will hopefully move closer to an understanding of the richness
and diversity of communicative devices.

Downloaded from sih.sagepub.com at TU Muenchen on July 6, 2015