The R̥gveda: Introduction

I. Who, What, Where, When?  3

A. Veda and R̥ gveda 3
B. Date and cultural context of the R̥ gveda 5
C. What do the hymns do?  7
D. The poet  9
E. Structure of the R̥ gveda 10
II. History of the R̥ gvedic Text 13
A. Language and compositional technique  13
B. Preservation and transmission  14
C. Indigenous commentarial tradition  18
D. Western scholarship on the R̥ gveda and major resources  19
III. Power of the Word  22
A. Words, truth, and knowledge  22
B. Homology  23
C. The wordsmiths  24
IV. Ritual 25
A. Model of the ritual  25
B. Basic elements of the ritual  27
C. Soma ritual  30
D. Other rituals  32
V. The Gods 35
A. Nature of the gods  35
B. Devas and Asuras  37
C. Viśve Devāḥ 38
D. Indra  38
E. Agni “Fire”  40
F. Soma  42
G. Aditi and the Ādityas 43
H. Savitar  44
I. Sūrya 45
J. Uṣas “Dawn”  45
K. Vāyu / Vāta “Wind” 46
L. Aśvins 47
2 The Rigveda

M. Maruts  49
N. Heaven and Earth  50
O. Tvaṣṭar 51
P. R̥ bhus 51
Q. Pūṣan 52
R. Viṣṇu 52
S. Rudra  53
T. Sarasvatī and the Rivers  53
U. Vāc “Speech”  53
VI. R̥ gvedic People and Society  53
A. R̥ gveda as history  53
B. Āryas, Dāsas, and Dasyus  54
C. Social and political organization  57
VII. Language and Poetics  59
A. Grammatical structure and language use  59
B. Hymn types and structuring devices  62
C. Imagery, metaphors, and similes  67
D. Riddles  70
E. Metrics  71
VIII. Translation Principles  75
A. Sanskrit into English: Problems and solutions  75
B. Format of the translations  81
C. Hymn distribution by translator  83

India has a magnificent tradition of religious literature stretching over three and a
half millennia, with a vast range of styles and subjects—from almost impersonal
reflections on the mysteries of the cosmos, the divine, and humankind’s relation
to them to deeply intimate expressions of worship. This literature is justly cele-
brated, not only within the religious traditions that gave rise to the various works
but around the world among people with no ties to those religious traditions. The
R̥ gveda is the first of these monuments, and it can stand with any of the subse-
quent ones. Its range is very large—encompassing profound and uncompromising
meditations on cosmic enigmas, joyful and exuberant tributes to the wonders of the
world, ardent praise of the gods and their works, moving and sometimes painful
expressions of personal devotion, and penetrating reflections on the ability of mor-
tals to make contact with and affect the divine and cosmic realms through sacrifice
and praise. Thus, much of what will distinguish later Indian religious literature is
already present in the R̥ gveda. Yet, though its name is known, the celebration of the
R̥ gveda is muted at best, even within its own tradition, and, save for a few famous
hymns, its contents go unnoticed outside of that tradition.
India also has a magnificent literary tradition, characterized in great part by
sophisticated poetic techniques and devices and a poetic self-consciousness that
glories in the transformative work that words can effect on their subjects. Again,
 Introduction 3

the R̥ gveda is the first monument of this literary tradition and at least the equal
of the later literature. The exuberance with which the poets press the boundaries
of language in order to create their own reflection of the complex and ultimately
impenetrable mysteries of the cosmos and the verbal devices they developed to mir-
ror these cosmic intricacies resonate through the rest of the literary tradition. Yet,
again, the R̥ gveda figures very little in standard accounts of Indian literature and is
little read or appreciated as literature.
Thus the R̥ gveda is not only the beginning but also one of the paramount expressions
of both the religious tradition and the literary tradition, combining these two roles in
a text that displays great variety, skill, and beauty. Surely it deserves a modern English
translation that makes these riches available to a wider audience. Yet it does not have
one; the only readily available complete English translation, the nineteenth-century
product of R. T. H. Griffith, conceals rather than reveals the wonders of the R̥ gveda
and would (properly) discourage any sensitive reader from further pursuit of the text.
Why this lacuna? The answer is quite simple: the R̥ gveda is very long and very hard.
Neither of these factors alone would necessarily hinder translation—both very long
texts, like the Sanskrit epics, and very hard texts, like the Avestan Gāthās, are receiving
their due—but the combination of the two has proved very daunting. We two transla-
tors, after some fifteen years of concentrated effort on the translation and more than
forty years of living with and working with the text, can attest to the rigors of the
task—but even more to its joys. And we feel privileged to have spent so much time in
intimate contact with the poets who shaped such an extraordinary religious and liter-
ary achievement at the very dawn of the Indian tradition.
In the introduction that follows we try to give readers some grounding in the
world and worldview of the R̥ gveda and to provide enough information to approach
the translation without undue bafflement. It is not meant as a comprehensive treat-
ment of the many subjects touched on, but only a stepping stone to the text itself
and the readers’ direct experience of the hymns.

I.  Who, What, Where, When?

A.  Veda and R̥ g veda

The R̥ gveda is the oldest Sanskrit text, composed in an archaic form of the lan-
guage, known as Vedic or Vedic Sanskrit. It is a collection of over a thousand
poems, composed by a number of different poets over the course of some consid-
erable period of time. The poems are primarily hymns praising various gods and
ritual elements and procedures, designed to be recited during ritual performance;
that is, they are liturgical compositions. However, they are also finely crafted and
self-conscious literary productions of the highest quality.
As the first text in Sanskrit, the R̥ gveda is somewhat isolated, and many of
the difficulties of its interpretation stem from the fact that there are no parallel
or closely contemporary texts. Yet, it is poised between two bodies of textual
4 The Rigveda

material that can contribute to its interpretation, and the characteristic features
from these two types of texts, mingled uniquely in the R̥ gveda, help account for
its distinctive quality. On the one hand, it stands at the end of a long tradition of
Indo-European and Indo-Iranian praise poetry, most nearly mirrored in the Old
Avestan Gāthās attributed to Zarathustra. On the other, it stands as the earliest
of the ritual texts collectively known as the Vedas and forms a part of the inter-
locking ritual system set forth in the Vedas.
There are four Vedas: the R̥ gveda, Sāmaveda, Yajurveda, and Atharvaveda. The
first three are the provinces of individual priests, who function together to per-
form the solemn rituals of the Vedic liturgical system, later, in the middle Vedic
period, known as śrauta rituals. Each of those three Vedas also represents a differ-
ent type of ritual speech. Thus, the R̥ gveda belongs to the Hotar priest, who recites
or chants the poetry; the Sāmaveda to the Udgātar priest, who sings the poetry to
set tunes called sāmans. The vast majority of the verbal material in the Sāmaveda
is borrowed from the R̥ gveda. The Yajurveda is the realm of the Adhvaryu priest;
his verbal product is the yajus, a short verbal formula that generally accompanies
the physical actions that are the main task of the Adhvaryu. Each of these three
priests is accompanied by other priests who share their principal functions. So in
the later soma ritual, for example, the number of priests can be sixteen or seven-
teen. The Atharvaveda stands outside of this ritual system and consists primarily
of hymns and spells of a more “popular” nature, often magical or healing. Despite
its lack of connection to the solemn ritual, the Atharvaveda is especially important
for R̥ gvedic studies because it is linguistically the closest text to the R̥ gveda and is
thus the second oldest text in Sanskrit. The two texts also share a number of pas-
sages and hymns, although the Atharvaveda often varies the wording or order of
verses. The R̥ gvedic hymns found also in the Atharvaveda are often drawn from the
younger layers of the R̥ gveda.
We will treat the structure of the text in more detail below; here we will pro-
vide only the most general outline. The text consists of 1028 hymns divided into
ten books or maṇḍalas (lit. “circles”), of varying lengths. The arrangement of the
hymns within each maṇḍala and the arrangement of the maṇḍalas themselves attest
strongly to the deliberate quality of the collection and organization of the hymns,
as we will demonstrate below. Maṇḍalas II–VII are known as the “Family Books,”
each attributed to a different bardic family. Maṇḍala VIII contains smaller collec-
tions attributed to particular poets or poetic families, and has a somewhat aber-
rant character. Maṇḍala IX contains all and only the hymns dedicated to Soma
Pavamāna, “self-purifying soma,” the deified ritual drink at a particular moment in
its ritual preparation. Maṇḍalas I and X were added to the collection later, though
they both contain much that is contemporaneous with the linguistic and religious
level of the core parts of the R̥ gveda, as well as some more recent and “popular”
material. Both I and X contain exactly 191 hymns, a synchronicity that was clearly
not by chance.
 Introduction 5

B. Date and Cultural Context of the R̥ g veda

As was mentioned above, the R̥ gveda is part of the long tradition of Indo-European
praise poetry, composed and performed orally and deploying inherited set verbal
formulae, on which the poets also ring changes. Thus, whatever date(s) we assign
to the actual composition of the particular hymns found in the text, the temporal
horizon of the R̥ gveda stretches a good deal further back, in that the poetic tech-
niques and even some of its precise verbal realizations go back many centuries, even
millennia.
The dating of the R̥ gveda has been and is likely to remain a matter of contention
and reconsideration because as yet little has been uncovered in the material record
or in the hymns themselves that allows us to date the period of the R̥ gvedic hymns.
One attempt at dating begins with an absence. Since the R̥ gveda does not mention
iron but does mention other kinds of metal, it is likely a pre–Iron Age, Bronze Age
text. The dates at which iron appears in the archaeological record in South Asia
differ in different parts of the subcontinent. For the northwest, which comprises
the geographic horizon of the R̥ gveda, iron began to be manufactured around
1200–1000 bce. The R̥ gvedic hymns, therefore, would have to have been composed
no later than this period. However, iron is attested in the Atharvaveda. While the
R̥ gveda is older than the Atharvaveda, there is no basis for assuming a substantial
gap in time between the end of the R̥ gvedic period and the Atharvaveda. Therefore
the date of the latest portions of the R̥ gveda is not likely to be very much earlier that
1200–1000 bce. It is also likely that the period of the composition of R̥ gvedic hymns
did not extend more than several centuries before this terminus ad quem. Witzel (in
Jamison and Witzel 1992: 2 n. 2) has noted that the poets and kings mentioned in
the anukramaṇīs (indices) and in the hymns themselves comprise perhaps five or six
generations. Generously rounding these numbers, we can then place the period of
the composition of the R̥ gvedic hymns sometime within the period 1400–1000 bce
or, even more approximately, within the second half of the second millennium bce.
At best these dates encompass only the hymns of the R̥ gveda as we have them. The
poetic conventions on which the R̥ gveda was built are very much older, extending
back to the Indo-Iranian period with roots into the Indo-European period. The
R̥ gveda is only the surface of a very deep tradition.
While the date of the R̥ gveda remains problematic, the hymns provide informa-
tion that helps identify the geographic area in which the hymns were composed.
Above all, the rivers mentioned in the text help establish the place of the R̥ gveda.
These rivers range from the Kabul and Kurram rivers in present-day Afghanistan
to the Ganges in the east. Its center is the greater Punjab, the region of the Indus
and its major tributaries. Following the likely internal chronology of the R̥ gveda,
geographic references in the text suggest a movement from the northwest toward
the east. Thus while the earliest parts of the R̥ gveda were likely composed in the
northwest, in the latest parts of the text the area has extended further into the
subcontinent, and its center has shifted toward Kurukṣetra, roughly the area of the
modern state of Haryana.
6 The Rigveda

One of the reasons that the R̥ gveda is difficult to date is that there is no material
evidence that we can clearly associate with the people who composed the R̥ gveda,
the people who called themselves Āryas. Nor would we expect very much material
evidence, since the hymns make no mention of any permanent religious structures
or enduring settlements. The Āryas formed instead a semi-nomadic pastoralist
society, in which seasons of settlement alternated with seasons of migration. This
migration likely contributed to the extension of the culture into new areas. The
period of movement was also the season of conflict in the competition for land and
the season of cattle-raiding, especially for younger males eager to acquire assets on
which to establish their own livelihood. Cattle were the primary source of wealth,
although the hymns also mention sheep, buffaloes, goats, and camels. Horses too
were essential and prized, since they enabled the Āryas’ mobility and contributed
to their success in battle. Although the economy was fundamentally pastoral, the
Āryas practiced some agriculture during the times of settlement; one hymn (IV.57)
specifically celebrates agricultural divinities, and the plow is occasionally men-
tioned. The hymns refer to yáva “barley” or “field grain,” which was used both for
food and in the rituals. The R̥ gveda does not attest rice cultivation.
In addition to the absence of material remains, another difficulty in describ-
ing the cultural context of the R̥ gveda is that its hymns depict only a part of the
religion and society at the period. First, the R̥ gveda represents the continuation of
an elite tradition also attested in the Avesta and therefore quite ancient. As such,
it reflects the religious practice only of the upper strata of Ārya society. Second, it
is primarily a collection of liturgical hymns for use in the soma sacrifice, surely the
most prestigious ritual of the period but still only one kind of ritual, representing
a particular and limited set of religious concerns. Finally, the soma sacrifices were
sponsored and performed by socially elite men, and they reflected the religious con-
cerns of these men. The text did not directly address the religious lives of women
or of other social classes nor indeed even other aspects of the religious lives of elite
males. Thus, while the R̥ gveda is a sizable text and from it we can derive a great
deal of information about the soma rite and about those who participated in it,
we are still dealing only with a segment of Ārya religion and society. However, we
can gather information on non-elite concerns and on the daily life and pursuits of
the elite incidentally, often through similes or imagery modeling ritual elements
and procedures or through the crediting of gods with activities also appropriate to
humans, such as warfare.
Indirectly, we can also get some information about other aspects of religion.
First, although the soma rite was primarily focused on the god Indra, already in
the R̥ gvedic period it had begun to incorporate the worship of gods around whom
independent ritual traditions existed. So, for example, the Aśvins were worshiped
already during the Indo-Iranian period and in the Pravargya rite, which is not a
soma ritual. But already in the R̥ gvedic period the Aśvins were recipients of soma,
and by the time of the later Veda the Pravargya rite had been incorporated into
the soma tradition. Moreover, especially in book X, there are hymns that address a
 Introduction 7

variety of religious interests separate from those of the soma rite. There are funeral
(X.14–16) and wedding (X.85) hymns. There are hymns against cowives (X.145),
against rivals (X.166), against witchcraft (X.155), against miscarriage (X.162), and
against disease (X.161, 163). There are hymns for the safety of cattle (X.169), for
conception (X.183), and for successful birth (X.184). In short the R̥ gveda already
attests rites that address domestic and individual issues principally associated with
the Atharvaveda. These hymns point to substantial ritual activity outside of the
soma rituals.

C.  What Do the Hymns Do?

The overwhelming majority of R̥ gvedic hymns have as their major aim to praise the
god(s) to whom the hymn is dedicated and to induce said god(s) to repay the praise
with requested favors. To a certain extent different gods receive different types of
praise, but the praise generally focuses on the appearance, qualities, and power of
the gods and on their remarkable deeds. Some divinities attract particular atten-
tion to their appearance: for instance, the seductive beauty of Dawn, the glittering
ostentation of the Maruts, the endlessly fascinating transformations of physical fire
and its divine embodiment Agni. Others, like the Ādityas, have few if any physical
characteristics, but are more celebrated for their mental and moral qualities. The
supernatural powers of almost all the gods receive abundant praise, though again
the types of power lauded differ from god to god.
Their powers are actualized in their deeds, the recounting of which occupies a
large portion of many R̥ gvedic hymns. Some gods have a robust narrative mythol-
ogy, and episodes from this mythology are constantly related or alluded to; the
most prominent example is Indra with his catalogue of great victories over both
divine and mortal enemies. Those without much narrative mythology tend to be
credited with general cosmogonic deeds or with the regular maintaining and order-
ing of the world and its inhabitants.
This praise of divine powers and deeds is not a disinterested act, for the aim
is to persuade or constrain the gods to mobilize these same powers on behalf of
their worshipers and to replicate their great deeds in the present for the benefit of
these same worshipers. In the all-pervasive system of reciprocity and exchange that
might be termed the dominant social ideology underlying the R̥ gveda, praise of the
gods requires requital:  they must provide recompense for what they receive from
those praising them. Worshipers are not shy about specifying what they want in
exchange: the good things of this world—wealth, especially in livestock and gold,
sons, and a long lifespan—and divine aid in defeating opponents, be they enemies
in battle or rival sacrificers. The sign that the praise has been successful is the epiph-
any of the god(s) addressed, so that many hymns urgently invite the dedicand(s)
to journey to the particular sacrifice in which the poet is participating and then
jubilantly proclaim the arrival of the god(s) at that particular sacrificial ground as
the ritual is taking place.
8 The Rigveda

This epiphany at the sacrifice brings us back to the liturgical role of the hymns,
for it should never be forgotten that almost all the hymns in the R̥ gveda were com-
posed to accompany the physical acts of the ritual, which are happening simultane-
ously with the recitation. But the word “accompany” here is too weak. The hymns
are not merely verbal background music, as it were. As will be discussed further
below, another important aspect of Vedic ideology is the belief in the power of the
word:  words make things happen. The physical actions of ritual alone would be
insufficient; it is the skillfully crafted, properly formulated hymn, the verbal portion
of the ritual, that makes the liturgical acts effective.
While the great majority of R̥ gvedic hymns have a liturgical form that obviously
reflects the soma rite, there are examples in which this model is not evident. Among
them are the ākhyāna or “narrative” hymns, as Oldenberg (1883, 1885) called them,
which take the form of a dialogue between two or more figures. These hymns occur
in the later portions of the R̥ gveda, especially book X. They include, for example,
dialogues between the sage Agastya and his wife Lopāmudrā (I.179); between Yama
and his twin sister Yamī (X.10); between the celestial Apsaras Urvaśī and her aban-
doned mortal husband Purūravas (X.95); among the monkey Vr̥ ṣākapi, the god
Indra, and Indra’s wife Indrāṇī (X.86); and between Indra’s dog, Saramā, and the
tribe of Paṇis (X.108). It is possible that some of these hymns comment on the
soma rite, but others were composed for different ritual purposes, which have to be
surmised, if they can be surmised at all, from the contents of the hymns. Our intro-
ductions to individual hymns discuss possible applications. A  few may represent
individual or domestic concerns, such as the recovery of lost cattle (X.108); oth-
ers may embody the differing viewpoints of ongoing ritual controversies (I.179) or
provide a dramatic modeling of a particularly important ritual (the Horse Sacrifice
in X.86) or the mythological underpinning for a series of hymns (X.10 for the fol-
lowing funeral hymns). An evergreen controversy concerns the form of the ākhyāna
hymns. Oldenberg (1885) argued that the oldest type of epic composition mixed
poetry and prose. The poetry principally consisted of the words spoken by char-
acters in the narrative, and the prose provided the narrative context for the verses.
This form is found in the Pāli Jātakas, the stories of the Buddha’s former births, for
example, in which the verses are considered canonical but the connecting prose is
not. Oldenberg suggested that the ākhyāna hymns conformed to this type and that
what we have preserved is a skeleton of canonical dialogue that originally had prose
narrative attached to it. Oldenberg’s theory has the advantage of explaining why
these hymns are difficult to interpret and why even the speakers of particular verses
are not readily identifiable. While we find the theory attractive, many scholars have
found it unnecessary and understand these hymns to have been recited as they are
transmitted to us.
There are also hymns that, though they may be ritually employed in the later Veda,
were perhaps not composed for ritual use. Gonda (1978: 25–38) compares some of
these hymns to medieval stotras: expressions of emotion, praise, and devotion to the
gods. However, such functions do not preclude their application in rites, even if their
 Introduction 9

original ritual context is not clear to us. A better possibility for non-liturgical hymns
is the type that comments on the ritual and its meaning. Generally occurring in the
latest strata of the R̥ gveda, these include X.129, the Nāsadīyasūkta (“that not exist-
ing did not exist”); X.121, the Hiraṇyagarbhasūkta (“Golden Embryo”); X.90, the
Puruṣasūkta (“Hymn of the Man”); and I.164, the “Riddle Hymn” of Dīrghatamas.
These are sometimes called the “philosophical” or “speculative” hymns of the
R̥ gveda, but this is a misleading description, since they are not primarily abstract
philosophic reflections on the nature of things. Rather, they are better viewed as
forerunners of the Brāhmaṇa and Āraṇyaka texts that interpret the ritual in general,
particular rituals, or aspects of the ritual. So, for example, X.90 comments on the
sacrifice through the symbol of the “Man,” which represents both the world and the
sacrifice. X.121 concerns the royal consecration rite, and X.129 sets forth the creative
power of knowledge and therefore the power of the poets and priests who possess
it. In many of these hymns the meaning of the ritual is expressed in terms of a cos-
mogony or cosmology. While such hymns share many themes and draw upon com-
mon stores of symbols, there is not a single R̥ gvedic cosmogony or a single R̥ gvedic
cosmology to which they refer. Rather, they represent imaginable worlds that explain
why things are as they are. To force the hymns into the straitjacket of a unitary view
of the world underestimates the power and originality of the poets who produced
these cosmogonic and cosmological models.

D. The Poet

Who is the poet, and why is he composing poetry? The poets participate in an
elaborate patronage system. They are hirelings, but of a very superior sort. As
craftsmen of the word, their contribution to the success of the sacrifice that estab-
lishes and maintains the mutually beneficial relationship between men and gods is
critical, and they serve the patrons, often royal patrons (whatever “royal” meant at
this period), who arrange for and underwrite the sacrifice. The poet provides the
praise poetry that the patron needs to put the gods in his debt, and he speaks on
behalf of his patron, in making specific requests of the gods for goods and services.
The poet’s reward comes as a second-hand or indirect benefit of the success of his
verbal labors: the patron should receive from the gods what he asked for, and he
provides some portion of that bounty to the poet in recompense. This payment
from his patron is sometimes celebrated by the poet at the end of his hymn, in a
genre known as the dānastuti, literally “praise of the gift,” in which the largesse of
the patron—cows, horses, gold, women—is catalogued and glorified. Or, if it is less
than expected or desired, scorned. The tone of the dānastuti is often teasing and
jokey, and the language colloquial.
But the making of poetry is not simply a business proposition. Poets take great
pride in their work and often reflect on their part in the poetic tradition and also on
their ability to use the tools of the tradition in innovative and creative ways. They
are self-conscious, naming themselves and addressing themselves, calling attention
10 The Rigveda

to their verbal tricks and achievements and their ability to bring fame and material
success to their patrons and glory to their gods. Some poets have very distinctive
poetic personalities, as we will have occasion to remark throughout the translation.
The R̥ gvedic poet’s social position and his role in the patronage economy was
clearly inherited from Indo-Iranian and Indo-European times, and one of the
closest parallels is found in ancient Greek, in the poetry of Pindar (5th century
bce), who was hired to celebrate the victors in the various Greek games and did so
in verse as elaborate, finely crafted, and deliberately obscure as that found in the
R̥ gveda. For further on the Indo-European poet, see Watkins (1995).

E. Structure of the R̥ g veda

The R̥ gveda comprises 1028 sūktas or hymns, which contain a total of slightly more
than 10,500 verses and which are divided into ten maṇḍalas, or books, of uneven
size. Within each maṇḍala there is a further division of the hymns into anuvākas
or “recitations” consisting of several hymns. The number of hymns in an anuvāka
varies within a maṇḍala, and the number of anuvākas in each maṇḍala varies from
maṇḍala to maṇḍala. As its name suggests, the anuvāka division was created prin-
cipally to provide convenient units for memorization and recitation. Although this
division is occasionally indicated in editions of the R̥ gveda, we have not included
it in the translation. There is a second division of the R̥ gvedic corpus into eight
aṣṭakas, but this is a purely mechanical arrangement also created to facilitate mem-
orization. In this latter division each of the eight aṣṭakas has eight adhyāyas, each
adhyāya has thirty-three vargas, and each varga has five verses. Since unlike the divi-
sion of the text into maṇḍalas, neither of these divisions reflects the contents of the
R̥ gveda, therefore we will use the division into maṇḍalas exclusively.
Invaluable work on the organization and history of the R̥ gveda was done by
Bergaigne (1886, 1887)  and Oldenberg (1888:  191–270), ably summarized and
amplified by Witzel (1995a, 1997). Following their work, the structure of the R̥ gveda
and the broad outlines of its compositional history are as follows. The core of the
R̥ gveda and its oldest part are the “Family Books,” so called because the hymns in
each maṇḍala are attributed to poets belonging to the same family lineage. These
comprise Maṇḍalas II–VII. The family lineages are the following:

II Gr̥ tsamada
III Viśvāmitra
IV Vāmadeva
V Atri
VI Bharadvāja
VII Vasiṣṭha

Within the R̥ gvedic corpus, the six Family Books are generally ordered according
to the increasing number of hymns in each successive maṇḍala. So Maṇḍala II con-
tains the fewest number of hymns and VII the greatest. Within each Family Book
 Introduction 11

the hymns are ordered first by deity. Thus the hymns to Agni come first, followed
by those to Indra. After these collections are the hymns to other deities, generally
arranged by the decreasing number of hymns to each deity within the maṇḍala.
Within each deity collection the hymns are arranged by their length, beginning with
the longest hymns. If two hymns are of equal length, they are ordered according to
meter, with the hymns in longer meters placed before those in shorter meters.
The arrangement of the Family Books and their hymns, therefore, functions
like an index. If you know the poet (and therefore the family of the poet), the deity
to whom the hymn is addressed, the number of verses in the hymn, and the meter,
then, in principle at least, you can locate the hymn within the collection. Perhaps
for that reason, the oral recitation of a R̥ gvedic hymn is traditionally preceded by
the identification of the poet, deity, and meter. It is this kind of information that is
provided by the anukramaṇīs or indices to the R̥ gveda. Because the anukramaṇīs
come from a later period, there has been some question about the value of their
information for the R̥ gvedic period. Some of the identifications of poets, in book X
in particular, are derived from the content of the hymns and can be rather wonder-
fully fantastic. In that book, for example, there are hymns attributed to serpents
(X.76 and 94), to the “Golden Embryo,” Hiraṇyagarbha (X.121), to the god Indra
(X.48–50), and to Yama (X.14) and Yamī (X.154), the first humans. Such identifi-
cations are not exclusive to book X. Among the possible composers of VIII.67 are
listed Matsya Sāmmada, king of the sea creatures, and fish that have been caught
in a net. Also, in the ākhyāna hymns and any other hymn in which the verses are
supposed to be spoken by a god or a legendary being, the anukramaṇīs ascribe
authorship to that god or being. Thus, the composition of the dialogue among the
monkey Vr̥ ṣākapi, the god Indra, and his wife Indrāṇī (X.86) is attributed to the
three of them.
However, these creative identifications are much more the exception than the
rule. Even though the anukramaṇīs were composed and redacted long after the
R̥ gvedic period, they are an invaluable resource, for, by and large, their identifica-
tions of the poets of hymns are plausible. The collections they mark by assigning
groups of hymns to certain poets or poetic circles correspond to the organization of
the R̥ gveda and to verbal, metrical, and thematic connections among these hymns.
The great majority of the roughly five hundred poets named in the anukramaṇīs
also appear in Pravara lists of brahmin ancestors (Mahadevan forthcoming), which
supports the plausibility of the anukramaṇī identifications. Therefore, the tradi-
tions transmitted in the anukramaṇīs can be a helpful guide in understanding rela-
tionships among hymns, in identifying collections of hymns, and in determining the
relative ages of hymns.
Generally younger than the Family Books, Maṇḍala VIII largely comprises the
hymns of two poetic traditions: that of the Kāṇvas (in 1–48 and 60–66) and that
of the Āṅgirases (in 67–103). The hymns of the Āṅgirasa group are probably some-
what younger than those of the Kāṇva group. However, not all the hymns in either
of these two groups are from Kāṇva or Āṅgirasa poets. Rather, both collections
12 The Rigveda

include hymns by poets who belong to other families and are known in other
maṇḍalas. These hymns were probably added to book VIII because they were com-
posed in forms and meters characteristic of the hymns of VIII. Both the Kāṇva
and the Āṅgirasa collections in VIII consist of a large number of hymns whose
verses are arranged in pragāthas or tr̥ cas, that is, in sequences of units (strophes)
consisting of two or three verses. There are pragāthas and tr̥ cas outside of VIII and
non-strophic hymns in VIII, but because of the predominance of these structures
in that book, pragātha and tr̥ ca poetry of various poets was relocated into VIII.
The significance of this collection is not entirely clear, although the marked forms
of the hymns suggest that they or the priests who produced them may have had a
distinct ritual function, and it is noteworthy that a large proportion of the R̥ gvedic
material borrowed into the Sāmaveda comes from VIII. It may be that the priests
who created Maṇḍala VIII were, like the Sāmavedic priests, those who chanted
R̥ gvedic verses. The arrangement of hymns in VIII generally follows that of the
Family Books: they are first organized by poet or poetic circle and then by deity.
But the organization of the book is less transparent than that of the Family Books.
Maṇḍala I also consists of two collections. One, I.51–191, probably dates from
around the time of the Kāṇva hymns of VIII, and the other, I.1–50, is slightly later
than the Āṅgirasa hymns of VIII. The collection of I.51–191 consists of the hymns
of nine groups of poets, organized according to the same principles as the Family
Books. The collection of I.1–50 consists of hymns in six groups, each attributed to
a single poet. Within each of these six groups, the hymns are collected by deity, but
the six differ in their arrangement of the hymns within the deity collections. The
groups of I.1–50 are also distinguished by the prevalence of hymns in gāyatrī meter
and in pragāthas, like the hymns of VIII, while the hymns of I.51–191 are primarily
in triṣṭubh and jagatī meter.
Maṇḍala IX is unusual, because it is a liturgical collection of hymns to Soma
Pavamāna, the soma “purifying itself ” as it runs across or through the sheep’s wool
filter. It includes hymns by poets already known from the Family Books as well as
by later poets. The collection is dominated especially by poets from books I, V, and
VIII. It was therefore created after the Family Books and contains hymns from
various periods. Like the Family Books, it is arranged in groups according to meter
and then within each metrical grouping, according to decreasing number of verses.
Maṇḍala X is a collection of hymns that belong to the youngest strata of the
R̥ gveda and forms a kind of appendix to the text. However, it shows organizational
principles comparable to those we have seen in the other books. It consists of collec-
tions of hymns by individual poets, which are ordered according to the decreasing
number of hymns in each collection or, when collections contain an equal number
of hymns, according to the number of verses in the first hymn of the collection. By
roughly the second half of X, the collections are reduced to single hymns by indi-
vidual poets. Finally, there is a short and late supplement to the R̥ gvedic collection,
the Vālakhilya hymns, which are collected in VIII.49–59.
 Introduction 13

While this is the general organization of the text, there are many exceptions
to these ordering principles. So, for example, we remarked above that the Family
Books, II–VII, are ordered from the shortest to the longest. While this is generally
true, it is not completely the case. Consider the list of the Family Books and the
total number of hymns in each:

II 43 hymns
III 62 hymns
IV 58 hymns
V 87 hymns
VI 75 hymns
VII 104 hymns

Although we would have expected the sequence of maṇḍalas to show a steadily

increasing number of hymns, instead book IV has fewer hymns than III and book VI
fewer than V. In a similar fashion, hymns can appear out of order within the various
groupings that make up both the Family Books and the other books of the R̥ gveda.
Such discrepancies have arisen through insertions of hymns and redactional com-
binations and divisions of hymns. These alterations occurred after the initial col-
lection of the R̥ gveda, when the order of books and hymns was established, and by
the time or at the time of Śākalya’s final redaction of the text around the middle
of first millennium bce. Attention to these discrepancies can be an effective tool
in reconstructing the compositional history of individual hymns and of groupings
of hymns, since they allow us to see where an alteration has occurred. Oldenberg
(1888:  193–94) provides the following example. In a series of single hymns with
decreasing numbers of verses, V.83 is a hymn to Parjanya of ten verses and V.85 is a
hymn to Varuṇa of eight verses. Between the two is V.84, a hymn to Earth of three
verses. Clearly the hymn to Earth is out of sequence and was likely inserted between
the hymns to Parjanya and to Varuṇa. While the fact of its insertion is obvious,
there are several possible explanations for how and why this occurred. For example,
V.84 could be a later composition that was later added to the R̥ gveda, or it could
have been composed earlier but have been moved to its current place within the
collection. In either case, this little hymn is actually a riddle depicting Earth during
a violent storm and must have been felt as an appropriate pendant to the Parjanya
hymn (see Jamison 2013).

II.  History of the R̥gvedic Text

A. Language and Compositional Technique

The R̥ gveda was composed in an archaic form of Sanskrit that is richer in forms
and less grammatically fixed than Classical Sanskrit, but essentially identical in
structure. For further discussion of the language, see section VII below. The text
14 The Rigveda

was composed entirely orally and transmitted entirely orally for a very long time,
probably several millennia. But it was a type of oral composition very different
from what that designation now generally brings to mind in scholarly, especially
Homeric, circles. It was not an anonymous floating body of infinitely variable ver-
bal material (re-)composed anew at every performance, generated in great part from
fixed formulae that formed the poet’s repertoire. In contrast to the vast sprawl of
epic, on which the usual model of oral-formulaic composition was formed and
tested, R̥ gvedic oral composition was small-scale and verbally complex. Though
orally composed and making use of traditional verbal material, each hymn was
composed by a particular poet, who fixed the hymn at the time of composition and
who “owned” it, and it was transmitted in this fixed form thereafter.
R̥ gvedic verbal formulae work very differently from those in epic compo-
sition. Rather than deploying fairly sizable, metrically defined, and invariant
pieces—ready-made surface structures, in the felicitous phrase of Paul Kiparsky
(1976: 83)—our poets seem to operate with deep-structure formulae. Invariant rep-
etition is fairly rare, and when it occurs, the repeated formulae tend to be short, gen-
erally shorter than the pāda (= verse line) and not necessarily metrically fixed. But
the poets often assume knowledge of an underlying formula, which seldom or never
surfaces as such, but which they ring changes on—by lexical or grammatical substi-
tution, scrambling, semantic reversal, and the like, confounding the expectations of
their audience while drawing upon their shared knowledge of the underlying verbal
expression. These deep-structure formulae tend to be shared across bardic families,
and we can in fact sometimes identify cognate formulae in other Indo-European
poetic traditions, especially in the Old Avestan Gāthās.

B.  Preservation and Transmission

The structure of the R̥ gveda points to several stages in the creation of the R̥ gvedic
text as we now have it. Collections of hymns were first made by the families of poets
who produced them, and these early collections defined the various poetic tradi-
tions and helped train new poets within those traditions. At some point a unified
consolidation was made of six family traditions, which formed the original collec-
tion of the Family Books, II–VII. As discussed above, the books were arranged
from shortest to longest and the hymns of each book were organized according to
the same principles. Then, probably at several intervals, the hymns of books I and
VIII (except for the Vālakhilya hymns) were added, and book IX was assembled
from hymns composed by poets of the other books of the R̥ gveda and from hymns
of younger poets. The last major additions to the collection were the hymns of
book X.
We do not know the precise mechanism for the formation of the R̥ gvedic collec-
tion or the circumstances that brought it about. There must have been some cen-
tralized authority or agency that could consolidate the different family traditions
and impose a single set of organizational principles on their collections. Michael
 Introduction 15

Witzel (cf. 1995a, 1995b, 1997, 2003)  has suggested that this authority was first
the Bharata tribe, as it attained hegemony over the other Vedic tribes during the
R̥ gvedic period, and then later the Kuru state, which arose around 1000 bce. In
his view, the initial collection and organization of the Family Books, the Kāṇva
hymns of VIII, and the nine collections of I.51–191 occurred under the Bharatas,
and the complete collection of the R̥ gveda under the Kurus. These consolidations
of the religious traditions supported the political consolidations of the Bharatas
and of the Kurus and reinforced their rule by means of a unified religious practice
approximating a state religion. The Kuru period saw the creation not only of the
complete R̥ gveda but also of the other saṃhitās, and the fixation and canonization
of Vedic sacrifices. The Vedic rites created at this time were composites, fashioned
from different family traditions. They included extended recitations constituted of
verses extracted from various parts of the R̥ gveda and thus from various family
traditions. The purpose of such composite rites was to create a ritual system that
represented the unity of the Vedic tradition. This process is already apparent in late
hymns of the R̥ gveda itself (cf. Proferes 2003a). For example, R̥ V IX.67 is a hymn
to “self-purifying” soma. Rather than being the product of a particular poet or even
a particular family of poets, it includes verses from poets representing the princi-
pal brahmin lineages. It reflects an attempt to create an “ecumenical” liturgy, as
Proferes (2003a: 8) calls it, one in which all the major poetic traditions had a place.
The creation of the R̥ gvedic Saṃhitā reflected a significant ritual change, since it
marked an emphasis on liturgical appropriation and repetition of earlier material
rather than, as in the R̥ gvedic period itself, on the creation of new hymns. However,
the tradition of R̥ gvedic composition did not simply come to a halt at the close of
the R̥ gvedic period. The R̥ gveda Khila (Scheftelowitz 1906) is a collection of hymns
that do not form part of the Śākalya recension. Some of these hymns may go back
to the R̥ gvedic period, but most were likely composed in the following period, dur-
ing which the hymns, chants, and recitations of the Atharvaveda, the Sāmaveda,
and the Yajurveda were composed or assembled. The Atharvaveda itself also rep-
resents the extension of hymnic composition into a wider variety of ritual contexts,
a process already visible in R̥ gveda book X. Beyond the Veda, elements and tech-
niques typical of R̥ gvedic composition appear in later praśastis, epic poetry, and
even in kāvya (see Jamison 2007: chap. 4).
The R̥ gveda did not remain unchanged after its collection. As described above,
the collection of hymns was arranged according to definable principles, but the
text of the R̥ gveda we have does not always follow these principles. Most of the
changes were made at an early period since they are reflected in all the versions
of the R̥ gveda that we have or that are described in later literature. These versions
were the product of Vedic schools or śākhās, which became the institutions through
which the R̥ gveda collection was preserved and transmitted.
The R̥ gveda translated here is the R̥ gveda of the Śākala school, established by
Śākalya, a teacher of the late Vedic period. There were other schools that produced
other recensions of the R̥ gveda, although most of these other recensions are now
16 The Rigveda

lost. The Caraṇavyūha, a Yajurvedic Pariśiṣṭa, lists five R̥ gvedic schools:  Śākala,
Bāṣkala, Āśvalāyana, Śāṅkhāyana, and Māṇḍūkāyana. Other sources give larger
numbers of R̥ gvedic schools. There are seven according to the Atharvaveda
Pariśiṣṭa, and twenty-one according to Patañjali (ca. 150 bce), although the last
number reflects not the number of versions of the R̥ gveda, but rather of schools
that studied the R̥ gveda. Of the five recensions mentioned in the Caraṇavyūha, the
oldest may be the Māṇḍūkāyana, although little is known about it. The Bāṣkala
school may have survived into the sixteenth century (Chaubey 2009: vii), and per-
haps the Bāṣkala R̥ gveda still exists somewhere in manuscript. But even without a
manuscript, much is known about it from other texts. It probably dates to around
the time of the Śākala recension and was close to the Śākala recension in sub-
stance. According to the Anuvākānukramaṇī, the Bāṣkala R̥ gveda included the first
seven hymns of the Vālakhilya, but rejected the other four, and after R̥ V X.191,
the last hymn in the Śākala recension, it had a second saṃjñāna hymn, or hymn of
“agreement,” consisting of fifteen verses. It also rearranged Maṇḍala I, so that the
Kutsa collection (I.94–115) followed the Parucchepa collection (I.127–139). This
rearrangement conforms better to the expected order of the collections that consti-
tute Maṇḍala I and therefore may represent either an older tradition than that of
the Śākala recension or a later correction made according to perceived principles.
The Āśvalāyana R̥ gveda has recently been published (Chaubey 2009). It was based
on the Śākala recension, but includes an additional 212 verses, all of which are
later than the rest of the R̥ gveda. The Śāṅkhāyana R̥ gveda was very similar to the
Āśvalāyana R̥ gveda. A sixth R̥ gvedic school was the Śaiśirīya school, mentioned in
the R̥ gveda Prātiśākhya. Its recension again closely resembled the Śākala recension
and indeed the Śaiśirīya school might have derived from the Śākala school or have
been merged with it (cf. Bronkhorst 1982/83). It again contained a few more verses
than does the Śākala recension. In short, the differences among the reported and
attested recensions of the R̥ gveda are very minor, consisting of variant ordering of
some existing materials and the inclusion or not of a relatively few late verses. There
seems no need to mourn the loss of these recensions.
These schools produced a saṃhitā text, that is, a continuous text of the R̥ gveda
that includes the phonological alterations that occur between words—a phenom-
enon characteristic of the Sanskrit language in general known as sandhi or “putting
together.” It is this basic form of the hymns that would have been recited in their
ritual contexts. But in order to secure the text, these schools also produced other
forms of the R̥ gveda that supported its memorization. According to Patañjali,
Śākalya not only created a recension of the saṃhitā text, but also a padapāṭha text.
This latter text provides a grammatical analysis of the words of the R̥ gveda by
restoring the forms of the words before the application of the sandhi rules when
the words are strung together. It shows the schools’ interest not only in preserving
and transmitting the R̥ gveda, but also in understanding the text they transmitted.
This history gives us reason to be confident that the Śākala R̥ gveda is close to
the R̥ gveda that was created at the beginning of the first millennium, even though
 Introduction 17

the Śākala recension probably dates to some five hundred years later. We also
have evidence for minor changes in the Śākala text itself. In the Śākala Padapāṭha,
there is no analysis for six verses in the Saṃhitā: VII.59.12, X.20.1, 121.10, 190.1–
3. They are probably missing from Padapāṭha analysis because they were not
part of the text of the R̥ gveda at the time of the creation of the Padapāṭha, but
were added to the Śākala text at a later period. Note again that these adjust-
ments primarily occur in book X, the latest part of the R̥ gveda and apparently
its most fluid.
When we say that the Śākala R̥ gveda is substantially the text created at the
beginning of the first millennium bce, we need to acknowledge one significant
area in which the R̥ gveda recensions show demonstrable change since the col-
lection of the R̥ gveda. This is in the phonetics of the text. The recitation of the
R̥ gveda in different regions and times apparently reflected the different contem-
porary dialects and conventions of recitation in those regions and times. Such
change is apparent in the Śākala recension in its handling of the phonological
alterations that take place between words. The Śākala school imposed a further
set of euphonic or sandhi rules on the text that developed during the centuries
between the composition of the text and the Śākala recension. The result is that
the saṃhitā text does not always reflect the metrical structure of the verses. In
most cases, the changes are sufficiently regular that it is not difficult to restore the
text to its metrical shape. For example, in the saṃhitā text the last verse of the
first hymn of the R̥ gveda reads: I.1.9 sá naḥ pitéva sūnávé, ’gne sūpāyanó bhava /
sácasvā naḥ svastáye. This hymn is composed in gāyatrī meter, so it ought to have
eight syllables in each pāda. But the elision at the beginning of pāda b gives a line
of seven syllables, and pāda c also apparently has seven syllables. Originally, the
verse must have been recited without the elision in b: sá naḥ pitéva sūnáve, ágne
sūpāyanó bhava. And in pāda c svastáye must have been recited quadrasyllabi-
cally su(v)astáye. While it is usually not difficult to restore the meter, that work
has been done for us in the edition of the R̥ gveda by Barend A. van Nooten and
Gary B.  Holland (1994), which gives the metrically restored text of the Śākala
recension.
These kinds of phonetic and euphonic changes were natural in the oral trans-
mission of the text, more natural than the rigid oral preservation of the text after
the Vedic period. Because such changes are natural, they were likely not deliber-
ate alterations. More importantly, the reciters of the R̥ gveda did not deliberately
change and, for the most part, did not change at all the order of the books of the
R̥ gveda, the order of verses within hymns, the words of the hymns, or their grammar.
There were a few—but relatively few—changes to the order of hymns, such as that
reflected in the difference between the Bāṣkala and Śākala recensions in the order
of Maṇḍala I. This early “freezing” of the text is very important and one of the
characteristics that makes the R̥ gveda so valuable for understanding the linguistic,
religious, and literary history of South Asia. The R̥ gvedic tradition has preserved
a very ancient literature with extraordinary fidelity, with no grammatical or lexical
18 The Rigveda

modernization or adjustment of contents to later conceptual conditions. It could

have been otherwise. In R̥ gvedic hymns that also appear in the Atharvaveda, the
latter text not uncommonly shows a different verse order, and in both Atharvavedic
and Sāmavedic versions of R̥ gvedic hymns there can be differences in wording and
in grammatical forms. In these cases, with few if any exceptions, the R̥ gvedic ver-
sion of the hymn is the older, and the versions of the other Vedas are modifications.
Up to the creation of the recensions of the R̥ gveda and long afterward, the
transmission of the R̥ gveda was oral. At some point, however, the R̥ gvedic schools
did produce manuscripts of the text. It is difficult to say when this occurred, but the
transmission of the text likely remained exclusively oral at least until around 1000
ce. The oldest manuscript in the collection of the Bhandarkar Oriental Research
Institute collection dates only to 1464; the Sampurnanand Sanskrit University in
Varanasi has an older R̥ gveda manuscript from the fourteenth century—thus a gap
of considerably over two millennia between the fixation of the text and our earliest
written evidence for it. Even when these activities did begin to occur, copying and
preserving manuscripts never displaced memorization of the text as the primary
means of transmission of the R̥ gveda until quite modern times.

C. Indigenous Commentarial Tradition

With regard to indigenous commentary, the situation of the R̥ gveda differs mark-
edly from that of standard Classical Sanskrit texts, in that there is no unbroken
commentarial tradition that might preserve the understanding of the text by the
authors and audience at the time of composition. Although we find implicit com-
mentary on some parts of the R̥ gveda already in later Vedic texts, it is clear that in
many case this “commentary” is based more on adaptation, speculation, or fancy
than on a direct transmission of the purport of the text, as when the Śatapatha
Brāhmaṇa (XI.5.1) sketches a narrative background for the Purūravas and Urvaśī
dialogue (R̥ V X.95) that distorts or misunderstands crucial portions of the hymn.
The first complete de facto commentary on the R̥ gveda, dating probably from
the late Vedic period, is Śākalya’s Padapāṭha mentioned above, which simply con-
sists of an effectively linguistic analysis of the continuous text (the Saṃhitāpāṭha)
of the R̥ gveda into individual words (padas). Because of sandhi, an important and
pervasive feature of the Sanskrit language whereby all words undergo significant
phonological adjustment to adjacent words in context, the phonological restora-
tion of the underlying pausal forms of words from the continuous reading is no
mere mechanical operation, but presupposes a grammatical and semantic analysis
of the text.
The Nighaṇṭu is a collection of difficult Vedic words probably made likewise
in the late Vedic period. The Nighaṇṭu and the commentary upon this collection
by Yāska in his Nirukta provide early lexical and etymological approaches to the
R̥ gveda, though understanding “etymology” in a synchronic, rather than our cur-
rent diachronic sense.
 Introduction 19

Various indexes or anukramaṇīs to the R̥ gveda, also mentioned above, were com-
piled probably around the middle of the first millennium bce, attributed to Śaunaka;
these were not fully preserved. A comprehensive index, the Sarvānukramaṇī, attrib-
uted to Kātyāyana and dating perhaps to the mid-fourth century bce but drawing
on the earlier indexes, provides, for each hymn, the poet, the god(s) to which it is
addressed, and the meter(s) in which it is composed. The Br̥ haddevatā, also attrib-
uted to Śaunaka, is in essence also an anukramaṇī, specifically an index to the dei-
ties of each hymn, but in expanded form, with a number of interesting narratives
and legends interspersed in the dry sequential listing of hymns and their divinities.
None of these ancient tools and treatments remotely approaches the standard
type of commentary familiar for later Sanskrit texts. For this the R̥ gveda had to
wait until the medieval period. The most influential and lasting commentary on the
text was made by Sāyaṇa in the fourteenth century ce in South India, although there
were a number of pre-Sāyaṇa commentators, some of whose work survives in part.
Sāyaṇa’s work essentially superseded these earlier works, and remains enormously
important in both indigenous and Western interpretations of the text: Max Müller’s
edition of the R̥ gveda includes Sāyaṇa’s commentary, and Geldner’s translation, for
example, owes much to Sāyaṇa. It should be remembered, however, that Sāyaṇa is
temporally closer to our own age than to that of the R̥ gveda, and he was writing
in a very different geographical, political, and religious landscape from that of the
R̥ gveda. It is therefore more useful to read Sāyaṇa not as a direct conduit of the
“true meaning” of the R̥ gveda but as a scholar grappling with the same problems
as modern interpreters, and bringing to bear all the intelligence and knowledge he
can muster, just as we do.

D.  Western Scholarship on the R̥ g veda and Major

Resources
Because the transmission of the R̥ gveda has preserved the text remarkably well, we
have a trustworthy and uniform text of the R̥ gveda. The published editions of the
R̥ gveda differ in their presentation of the text, not in the text itself. The landmark
edition of the R̥ gveda was that of Max Müller (1849–74), who published both the
Saṃhitā and Padapāṭha text in devanāgarī script, together with Sāyaṇa’s commen-
tary. Before the completion of Müller’s R̥ gveda, in 1861–63 Theodor Aufrecht pub-
lished the text in Roman transliteration, together with selected Padapāṭha analyses.
A  second edition, the standard edition of Aufrecht’s R̥ gveda, appeared in 1877.
The R̥ gveda and Sāyaṇa’s commentary were published again under the editorship
of N. S. Sontakke (1933–51), together with the Padapāṭha. Also noteworthy is the
aforementioned publication of the metrically restored version of the R̥ gveda text by
Barend A. van Nooten and Gary B. Holland (1994).
The uniformity and reliability of the text of the R̥ gveda cannot be said of its
translations, which vary considerably. The standard scholarly translation remains
that of Karl Friedrich Geldner into German. This translation was published in full
20 The Rigveda

in 1951, but was complete in the 1920s and partially published in limited fashion
then. Although it remains a remarkable philological accomplishment, whose worth
we two translators have appreciated more and more over the years, it of course
could not take account of the advances in Vedic scholarship over the last eighty
years or so. Louis Renou (1955–69) was able to finish most of a French translation
of the R̥ gveda, in a series of thematic publications, under the general title Études
védiques et pāṇinéennes, organized by the divinity addressed rather than the order
of the R̥ gvedic text. But he left undone substantial parts, notably the Indra and
Aśvin hymns, and the later publications are rather sketchy. T.  Ya. Elizarenkova
(1989–99) completed a Russian translation of the text. Currently in preparation is a
new German translation of the R̥ gveda under the direction of Michael Witzel and
Toshifumi Gotō. The first volume of the Witzel-Gotō translation (2007), which cov-
ers Maṇḍalas I and II, has appeared, with the second volume scheduled for 2013.
Unfortunately, English has not been as well served as these other languages.
Aside from anthologies, the English version that is in general use is R. T. H. Griffith’s
translation, which was first published in four volumes between 1889 and 1892, then
in a revised edition in 1896, and then yet again in another revised edition, this time
by J. L. Shastri, in 1973. Griffith’s translation has been reprinted several times since
1973 and is available online (http://www.sacred- texts.com/hin/rigveda/index.htm).
Sadly, this translation really does not deserve as many rebirths as it has had. Its phi-
lology was already dated when it was published, and the English style of the transla-
tion is cloying and almost unreadable. Now, well over a century later, it should have
long since been superseded. There was at least one serious effort to do so. From the
late 1940s until the early 1960s, H. D. Velankar steadily published English transla-
tions of the R̥ gveda, which were a decided improvement over Griffith’s work (for a
list of his translations, see the Bibliography). These were published as independent
volumes dedicated to books II, V, VII, and VIII of the R̥ gveda and as collections
of hymns to different deities published in the Journal of the University of Bombay.
Partly because these translations are scattered, incomplete, and difficult of access,
they have received less attention than they might otherwise have done. In addition
to these complete or extensive translations of the R̥ gveda, several anthologies of
R̥ gvedic hymns have appeared. In English, the most notable are those of Wendy
Doniger O’Flaherty (1981) and Walter H. Maurer (1986). The principal problem
with any such anthology is that translators understandably choose hymns that they
think will be especially interesting for their readers and accessible to them. As a
result, they tend to create a distorted view of the R̥ gveda that does not reflect the
liturgical functions and scope of the text.
Both the partial and the complete translations just mentioned were fundamental
for the present translation. In our introductions to individual hymns, we will often
make reference to them, especially when they suggest interpretations of the text we
deem particularly worthy of note. Rather than giving a fuller bibliographic refer-
ence, we will refer to them only by the last names of the translators. The exceptions
are Renou’s translations, for which we will give the relevant volume and page of his
 Introduction 21

Études védiques et pāṇinéenes, and Velankar’s translations, where again we will cite
the specific source.
In addition to these translations, there are a number of other essential resources
for translating and interpreting the R̥ gveda. Among the older works of particular
significance are the Wörterbuch of Hermann Grassmann (1872–75), whose pre-
sentation of the lexicon of the R̥ gveda is somewhat antique but still very useful.
Grassmann’s work has now been complemented and in some respects superseded
by Alexander Lubotsky’s concordance to the R̥ gveda (1997), which provides the
verse-line context for R̥ gvedic words. Among older works of significance, of par-
ticular note is Hermann Oldenberg’s Noten (1909, 1912) on the complete text of
the R̥ gveda, whose insights remain remarkable even a century after they were pub-
lished. Maurice Bloomfield’s Rig-veda Repetitions (1916) is an invaluable resource
for the study of R̥ gvedic formulaic language and its variations, though it predates
the discovery of oral-formulaic composition.
In the last fifty years or so, R̥ gvedic scholarship has blossomed with studies that
have significantly advanced our understanding of the language of the R̥ gveda—its
lexicon, morphology, and syntax—of its compositional techniques, and of its con-
ceptual universe and ritual procedures and context. We cannot offer even a partial
list of such books, let alone articles, since we would omit too many. We will mention
only a few, those which were especially close at hand as we worked through the text
and whose insights we have often adopted. Particularly important was Manfred
Mayrhofer’s etymological dictionary of Old Indo-Aryan (EWA 1986–2001) and his
previous version (KEWA 1951–76), not completely superseded by the newer one.
Other works include Salvatore Scarlata’s study of nominal compounds ending in
roots (1999) and Jared S. Klein’s studies of Vedic particles and discourse structure
(1985). One of the subjects in which there has been substantial progress just in the
last decades has been the Vedic verbal system. The flood of monographs on the ver-
bal system probably began with Johanna Narten’s work on the s-aorist (1964), fol-
lowed closely by Karl Hoffmann’s influential study of the injunctive (1967). Other
verbal subsystems treated more recently include the -áya-stems (Jamison 1983), the
first-class presents (Gotō 1987), the intensive (Schaefer 1994), the perfect (Kümmel
2000), the desiderative (Heenen 2006), and the ya-presents (Kulikov 2012). All of
these works, as well as many other and many shorter studies, have been of critical
importance as we have worked our way through the text.
Finally, we wish to note one last development not just in R̥ gvedic but in Sanskrit
scholarship more generally. Among older scholars there was a tendency to deal
with difficulties in a text by emending it. By the beginning of the twentieth century,
however, scholars increasingly recognized that they could gain a much better under-
standing of the text by accepting the text as transmitted. This process is visible in
Oldenberg’s scholarship. Before the Noten, Oldenberg published translations of the
Agni hymns of the first five books of the R̥ gveda (1897). There he was willing to
suggest text emendations to smooth rough spots in the hymns. By the time of the
Noten, however, he had become much more apt to accept the text as it stands and
22 The Rigveda

to explore ways of accounting for that transmitted text. We too are committed to
accepting the traditional text and more importantly to allowing the poetry of the
R̥ gveda to remain complex, elusive, jagged, unsettled, and even unsettling.

III.  Power of the Word

A.  Words, Truth, and Knowledge

One reason for the intricacy of R̥ gvedic poetry and the careful thought that the
R̥ gvedic poets put into it is the importance that Vedic culture attached to the spo-
ken word and to the truth that it embodied. The elegantly formulated truth, spoken
in a ritual context, was powerful. The word for “truth” is r̥ tá, a crucially resonant
word that, with some reason, some other translators have rendered “order” or “cos-
mic order.” The term r̥ tá essentially defines what a being or object is and what it
does, and it structures the relationships of beings and objects with other beings and
objects. By speaking these truths of essence and relationship, the poets could make
the truths real and actual in the present. So, for example, the great Indra hymn,
I.32, begins, “Now I shall proclaim the heroic deeds of Indra, those foremost deeds,
which the mace-wielder performed. . . . ” Why does the poet proclaim these deeds?
It is not simply to honor the god, although his proclamation surely does do that. It
is also to state the truth of these heroic deeds, so that these deeds will become real
once again. As Indra once before smashed Vr̥ tra, who was the symbol and epitome
of all obstacles, so once again he will smash obstacles. Formulating the “truth” of
Indra is part of what makes Indra real and present. Similarly, the story of Indra
and the Vala cave is essentially a story of the power of the truth. According to
this myth, Indra and the Aṅgirases opened the Vala cave and released the cattle
and the dawns by the songs they recited. These songs were powerful because they
contained the truth that the cattle were the dawns, and therefore, by singing this
truth Indra and the Aṅgirases obtained both cattle and dawns. In X.108 the poet
narrates part of the story of Indra and Vala and then states his expectation that
the truth of his song, which is the truth of the Vala story, will bring cattle back to
him: “Exchanging with the truth, let the cows come up, which Br̥ haspati [=Indra]
found hidden. . . . ” Or again, in IX.113.2 the poet declares the soma is pressed “with
real words of truth, with trust, and with fervor.” That is to say, it is not just the
physical pressing of the soma plant that produces soma juice. Soma is also cre-
ated by the intensity of the priests, by their confidence in the effectiveness of their
actions, and by the truth they speak about the soma and about the power of the
soma to strengthen the gods and to give life to mortals. Words, commitment, and
ritual all combine to make the soma real.
The product of the formulating of a truth, the verbal formulation itself, is the
bráhman, and the poet who formulates truth is the brahmán. While the later ritual
tradition will rely on ancient formulations of the truth passed down from the early
 Introduction 23

Vedic period, the contemporary religious system of the R̥ gveda required ever-new
formulations of the truth. These new formulations would attract the gods and make
them present at the ritual, and these new formulations would make the truth they
expressed newly real. For this reason the work of the poet was essential, for only
poets could continue to produce these new formulations. The poet needed to be a
master of many skills, but of all the skills he needed, the one that was most essen-
tial was knowledge. The truths that the poets formulated were often hidden truths,
founded on enigma and paradox. One such kind of truth was homological truth
that connected objects across spheres, which will be discussed in the next section.
The importance of knowledge for the Vedic poets is underscored by the variety
of words for the poets’ thinking and its articulation in the hymns. The poets speak
of their dhī́ or dhītí “insight” or “vision,” matí or mánman “thought,” and manīṣā́
“inspired thinking.” The hymn and the understanding that gives rise to it are so
closely related that the boundary between them becomes permeable, and the poets
use words for thinking and knowing for the hymns themselves. In I.2.7, for example,
the poet calls on Mitra and Varuṇa “who bring success to our ghee-covered insight
(dhíyaṃ ghr̥ tā́ cīm),” that is, to the hymn that is accompanied by offerings of ghee, or
again, in VI.8.1 the poet offers Agni Vaiśvānara a “newer thought” (matír návyasī),
which purifies itself (pavate) in the poet’s mind and emerges in his speech.

B. Homology

The Vedic mental universe is structured, in great part, by a web of identifications

or equivalences among elements in the ritual realm, in the cosmic realm, and in the
realm of the everyday. These homologies play an enormous role in the thought of
the Brāhmaṇas and Upaniṣads, where knowing the hidden connections between
apparently disparate elements, particularly the equivalences between cosmic and
ritual elements, gives the knower some power to control the cosmic by manipulation
of the ritual. In these later Vedic texts, such types of equivalences have a technical
name—bandhu “bond, connection” (a word of course cognate with English “bind,”
etc.)—and the difference between the cosmic and the mundane system of equations
is pervasive enough to be distinguished by technical terms: the cosmic level is called
adhidaivam or adhidevatam, roughly “concerned with the divine,” while the mun-
dane is called adhyātmam, roughly “concerned with the person(al).” These two sys-
tems meet in and focus on the sacrifice, a controlled and orderly sphere of human
activity, as a way of modeling the complex web of relationships that obtain in the
two other realms, which are not under human control. The level of the sacrifice is
called adhiyajñam “relating to the sacrifice.”
Although the correspondences are not as systematically worked out and pre-
sented in the R̥ gveda as in those later texts, this way of thinking is just as pervasive
in that text, though backgrounded:  the poets assume that they share with their
audience a web of associations, and this shared knowledge allows the poet to sub-
stitute one element for another without overt signaling. Thus, for example, the first
24 The Rigveda

rays of light at dawn are homologized to cows, as in the Vala myth discussed above,
and therefore the goddess Dawn is called “the mother of cows” and images of
ruddy cows overrun the hymns to Dawn. Cows also have a ritual association: milk
is mixed with the soma juice after it is pressed, and so the soma hymns are filled
with depictions of the soma (as bull) running toward or mixing with “the cows,”
that is, the milk. And cows are also homologized to waters, particularly the waters
confined by Vr̥ tra and released by Indra’s killing of that demon.
Modern readers of the R̥ gveda must try to internalize the many associations and
identifications that formed the unconscious mental universe of the contemporary
audience; otherwise almost every hymn in many of its parts will seem unintelligible
or nonsensical. Other important homologies include that between the sun and the
fire, especially the ritual fire, and, in general, the many associations between the
various parts and procedures of the ritual and both cosmic and everyday elements.
For example, the ritual itself, or the praise hymn specifically, is often identified with
a chariot, and the crafting of poetry is homologized to chariot-making. Soma is
often identified with a bull, but often otherwise with a horse, and the ritual prepara-
tion of the soma is equated with the grooming of this horse. Or he is a bird settling
into his nest or a king embarking on conquest. The ritual ground itself is often iden-
tified with the cosmos, with the ritual fire a pillar connecting heaven and earth and
any movement on the ritual ground implicitly compared to a vast journey across
or around the cosmos. It is not possible here to list anything approaching all the
important connections that underlie R̥ gvedic poetry, though we will try to sketch
many such associations in particular circumstances. But the reader must be alert
to, and open to, this overall substitution principle, and also recognize that these
homologies are not mere poetic embellishments, imagery for its own sake, but an
implicit statement about the way things really are, the pervasive underlying connec-
tions unifying apparently disparate elements.
It is one of the poet’s main jobs to find and articulate these hidden connections,
and the theme of “secret names” and their discovery is an important one in the
R̥ gveda. Indeed, one of the great deeds attributed to Indra (as Br̥ haspati) and to the
poet-singers, the Aṅgirases, divine counterparts of the human poet—the splitting
of the Vala cave and the release of the cows imprisoned therein—turns in some
versions on discovering the cows’ hidden names (e.g., IV.1.16, X.68.7). The training
of a poet clearly involved both producing and solving quite recherché associations,
and the genre of riddle displays these skills to great advantage.

C. The Wordsmiths

A number of different words are used to identify the workers in words, for which
we use the cover term “poet.” These different terms often in fact have the same
referent, sometimes in the very same passage, but emphasize different aspects of
verbal craftsmanship. The most important terms are brahmán, kaví, ŕ̥ṣi, and vípra.
Although it is not always possible to draw fine distinctions among these words, and
 Introduction 25

there remains considerable disagreement about their application (we two transla-
tors, for example, are not entirely agreed on the function of the kaví ), the words do
point to different types of relationships between the poet and his product and its
source. As noted above, the brahmán is the “formulator,” who discerns and puts into
words the hidden truths that undergird reality. His product is a bráhman, a “sacred
formulation.” The term ŕ̥ṣi, generally rendered here as “seer,” is especially applied
to the famous seers of the past, single or as a group, and to current poets who con-
sciously aspire to the status of their predecessors, and like other words for verbal
craftsmen, it is also regularly used of gods, especially Agni and Soma. The vípra, a
derivative of the root √vip “tremble,” appears to be an “inspired poet” and contrasts
with the kaví, the poet par excellence, but also the “sage poet,” endowed with the
knowledge embodied in poetic skill (kā́ vya). The original sense of the word kaví
and the function of the person so named is disputed; it has well-attested Iranian
cognates, starting with Avestan kauui, whose sense diverges from its Vedic corre-
spondent. We two translators have discussed these issues in separate treatments
(Brereton 2004; Jamison 2007: chap. 4).
There are a few less well-attested terms, such as kārú here rendered “bard” or
“praise-poet,” the transparent agent noun stotár “praiser,” as well as designations
that seem more related to performance, such as jaritár “singer,” vāghát “cantor” or
“chanter,” rebhá “husky-/raspy-voiced (singer).”

IV. Ritual

There have been a variety of approaches to the R̥ gveda, by both ancient and mod-
ern scholars. Following the lead of the Br̥ haddevatā, some interpreters have been
concerned primarily with the myths and legends of the R̥ gveda. Others have sought
to understand the deities of the R̥ gveda by exploring the meaning and derivation
of their names. And as early as the later Vedic period, still others understood the
R̥ gvedic deities as the powers of the natural world or as symbols of microcosmic
processes. But the R̥ gveda is first of all a liturgical text. The great majority of its
hymns were composed for rituals and, more specifically, for the soma rituals of
their period. After the collection of the R̥ gveda, its verses were adapted to the reci-
tations and chants of the classical soma ritual and employed in a variety of other
ritual contexts. Understanding the R̥ gveda, therefore, first requires understanding
the Vedic rituals and, above all, the soma ritual.

A. Model of the Ritual
The ritual as depicted both in the R̥ gveda and in the later Vedic texts treating the
classical śrauta system (and indeed the much later pūjā of classical Hinduism) is
modeled as a hospitality ceremony and festive meal, offered to the visiting gods. The
poets eagerly invite the gods to journey to attend the sacrifice. When they arrive,
26 The Rigveda

they are greeted and provided with seats near the center of the action, on a special
grass strewn on the ritual ground to make the ground more comfortable. This grass
strew is called barhís and it has an exact cognate in Avestan barziš “cushion, pillow.”
In some hymns the horses pulling the gods’ chariots, particularly Indra’s two fallow
bays, are unhitched and offered refreshment of their own. The gods are then offered
a meal, that is, the oblations, generally consisting of ghee, little grain cakes, and,
in the most important sacrifices, soma, and they are entertained while they eat, by
the hymns of praise that constitute the R̥ gvedic corpus. The end of this entertain-
ment is often announced in the final verse of the hymn, and sometimes the gods are
explicitly sent on their way back to heaven. Particularly nice examples of this envoi
are found in I.61–63 and I.82, hymns designed to accompany the Hāriyojana or
“Fallow-bay-yoking” oblation, the ritualized hitching up of Indra’s horses for the
return journey; in I.82.5–6 Indra is charmingly urged to go home to his “dear wife.”
The “ritual ground” alluded to above is the sacralized space within which the rit-
ual proceeds. It is not a permanent space dedicated to ritual performance, much less
a building or temple; rather, judging from the evidence of the later śrauta texts, it is
demarcated and sacralized for each performance. The space is defined especially by
the presence of three ritual fires, the focal point of the sacrifice and the recipient(s)
of the oblations. In the later texts these are the Āhavanīya (the fire “to be offered
into”) set to the east, the Gārhapatya (“householders’ fire”) to the west, from which
the Āhavanīya is taken out and carried to the east, and the Dakṣiṇāgni (“southern
fire”), and their locations and functions are precisely fixed. The R̥ gveda does not
attest these names for the fires, but it does often refer, though often obliquely, to
the three fires of the ritual. Moreover, the god Agni and his physical representative,
the offering fire, regularly receive the epithet puróhita “placed to the east/in front,”
appropriate to the removal of the Āhavanīya fire from the western Gārhapatya
and its movement to the east. (The third word in the R̥ gveda is in fact puróhitam
I.1.1.) The poets sometimes seem to treat this movement as a sacred mystery (e.g.,
III.31.1–3), a fact that may point to the three-fire model as a ritual innovation of
the pre-R̥ gvedic period. There is no parallel to it in Avestan ritual, which uses only
a single fire.
One aspect of the “entertainment of guests” model of Vedic sacrifice is the fear,
constantly expressed, that the gods will choose to go to another party. The R̥ gvedic
sacrificers know that everyone in the larger sacrificial community follows much the
same ritual calendar and therefore many competing sacrifices are taking place at
the same time. The invitation to the gods often explicitly urges them to pass over
other sacrifices and come to ours, and in order to make ours more attractive, we
must provide the best hospitality, in particular the best entertainment in the form
of exquisitely crafted and novel hymns.
An important feature of the “guest” model is the fact that it envisions and
requires direct interaction between gods and mortals, on the mortals’ turf—or
rather on a space that has been rendered temporarily neutral through its sacral-
ization—though it is on earth, the human realm, rather than in the gods’ world,
 Introduction 27

heaven. The interaction in turn requires epiphany in the technical sense: the gods
must appear in person, as it were. The poets crave this epiphany and fear its failure
to materialize. It is especially Indra whose epiphany they anticipate, and the fear
that it may not happen also leads to the more general fear that Indra does not exist,
a possibility that is often put in the mouth of doubtful “others.”
There is another, more recessive model of sacrifice perceptible in the R̥ gveda,
especially in Agni hymns, whereby the offerings go to the gods in heaven, rather
than requiring the gods to come to earth to receive them. The libations ascend on
the smoke of the offering fire, an image conceptually close to the Homeric sacrifice.
There seems to be no conflict in the minds of the R̥ gvedic ritualists between these
two models, and they are not exclusive to different groups of poets but can in fact
occur in the same hymn; see, for example, the Agni hymn I.1, where the “guest”
model dominates, with the gods’ travel to the sacrifice explicit in verses 2 and 5, but
the oblation’s movement to the gods is alluded to in verse 4.

B.  Basic Elements of the Ritual

Later Vedic texts give detailed descriptions of the soma rituals, which varied in
length and complexity. The problem for the interpretation of the R̥ gveda is that
these classical rites were clearly not the same as the rites of the R̥ gvedic period
itself. While the classical rites drew on R̥ gvedic traditions, they did not simply con-
tinue them. In the R̥ gvedic period these rituals were probably much more variable,
for even though their rituals were based on a shared inherited tradition, different
priestly families performed the rites differently. For example, in the classical one-day
soma rite, soma juice is pressed out of the soma plant and offered to various deities
in the morning, at midday, and in the evening. But the third or evening pressing
stands apart from the other two. While soma plants are freshly pressed in the morn-
ing and at midday, in the evening the soma is extracted from the already-mangled
soma stems left over from the earlier pressings. Not only the soma preparation but
also the overall character of the Third Pressing is different. It has a greater focus on
the sacrificer and his wife, the gods of the Third Pressing form a less coherent group
than those of the other pressings, and in general, the rites of the Third Pressing
have the appearance of an appendix. It is possible that some family traditions did
not have a Third Pressing and that, among those that did, the form of the Third
Pressing was more fluid. Or again, in the Midday Pressing of the classical soma
ritual, there are offerings to Indra alone and to Indra along with the Maruts. This
form may represent a compromise between family traditions that offered soma only
to Indra at this second pressing and those that also included the Maruts.
Variation in the rite also occurred over time. The R̥ gvedic period itself extended
through several generations of poets and priests, and the rites evolved during this
period. In the classical soma ritual the sacrificer must be accompanied by his wife,
who has her own ritual responsibilities (see Jamison 1996a). But in the R̥ gveda,
the wife of the sacrificer is very rarely alluded to and then almost exclusively in
28 The Rigveda

the younger parts of the text. Therefore the wife of the sacrificer may have been a
late addition to the soma ritual, and indeed the R̥ gveda even offers evidence that
her introduction may have been controversial, with some ritualists defending and
others rejecting her presence in the rite (see Jamison 2011, forthcoming a, and forth-
coming b). Ultimately, those who wished to include the sacrificer’s wife won the day,
for in the classical soma ritual her presence is required even if her ritual functions
are limited.
The consolidation of the ritual tradition and the creation of the classical forms
of the soma rite probably occurred after 1000 bce, during the period in which the
other saṃhitās were compiled. The innovations effected by this consolidation not
only changed the form of the ritual, but also altered its very nature. Fundamental
to the R̥ gvedic rite was the need for poets to create new formulations of the truth
in order to attract the attention of the gods, to make them present at the rite, and
to define their action. But in the later Vedic period, not only did the ritual offer-
ings become increasingly standardized, so also did the ritual recitations and chants.
Novelty and innovation were no longer primary values, but instead were supplanted
by the ability to remember the old compositions. The compilation of the R̥ gveda
reflects the need to preserve the poetry of the past in order to employ it within
the ritual, for in the later Veda power resided not in the new formulation but in
the ancient one. Or rather and more generally, power was understood as residing
in the ritual performance as a whole and with the priests who carried out the rite.
In later Vedic interpretation the ritual even came to exercise power over the gods
themselves. The ritual became less an invitation to the gods and more a manipula-
tion of them. One marker of this shift was the rise in the later Veda of Prajāpati, the
Lord of Creatures. Prajāpati was the sacrifice itself, and his superiority to the other
gods therefore symbolized the dominance of the sacrifice over the gods. Knowledge
remained important: priests should understand the secrets of the rite that they per-
formed. But the knowledge and skills of a poet did not have a place in the evolving
ritual order.
Given the fluidity of the rites during the R̥ gvedic era, we cannot really speak of
the R̥ gvedic soma rite, even though there were certainly common characteristics of
the soma rituals performed throughout the period. Moreover, given the changes
that occurred after the R̥ gvedic era, we cannot simply project the classical soma
rites back into the R̥ gveda. While many technical ritual terms and elements were
passed down from the R̥ gvedic period to the classical rites, these terms may not
have had the same significance for the R̥ gveda that they do for the classical tradi-
tion. This is perhaps most obviously the case for the brahmán. In the later ritual the
brahmán became the formal overseer of the ritual. In the R̥ gveda he was rather the
poet who formulates the truth (further see Brereton 2004).
We have already noted that fire is at the center of the Vedic rites and that the
three fires of the classical śrauta ritual model were likely present in the R̥ gvedic
ritual. It is clear that already in the R̥ gveda the principal offerings were made into
the fire, and the gods received them through this mediating force. It is no wonder,
 Introduction 29

then, that the first hymn of the R̥ gveda is a hymn to Agni, the god of fire, since
his appearance ritually marked the beginning of the sacrifice, nor that Agni is the
recipient of more hymns than any god but Indra.
Sacrifices were carried out by men performing a variety of priestly functions.
In II.1.2 and II.5.1–8 Gr̥ tsamada poets provide a list of seven priests, to which the
Gr̥ hapati “houselord,” who is the primary beneficiary of the rite, is added as an
eighth. Although the correspondence is not precise, Oldenberg (1917:  385)  com-
pares this list to a list of eight priests in the Avesta and argues that it reflects an
old tradition that was still alive in the R̥ gveda. In II.1.2 these seven priests are the
Hotar, Potar, Neṣṭar, Agnīdh, Adhvaryu, Praśāstar, and Brahman. As Minkowski
(1991:  113)  points out, these seven essentially correspond to the classical list of
priests in the R̥ tuyājas, the “sacrifices in turn” (which adds the Acchāvāka priest)
and in the R̥ tvigvaraṇa, the “choosing of priests,” and to the priests who recite at
the Prasthitahomas, the “presented oblations.” It is likely, therefore, that the list of
seven does rest on a well-established tradition.
The Hotar was the principal priest, and the R̥ gveda repeatedly mentions the
“seven Hotars” (VIII.60.16; IX.10.7, 114.3; X.35.10, 61.1, 63.7) or the seven hótrāḥ
“priestly functions” (X.17.11), by which it probably refers to the group of seven
priests led by the Hotar. Although in the R̥ gveda and in the later tradition, the
Hotar is the priest who recites the hymns and is secondarily associated with the
root √hū “call,” his name originally meant the “pourer” (derived from the root √hu
“pour”), which indicates that his ancient function was both to pour the offerings
and to recite. In the R̥ gveda he was likely often the composer of the hymns he
recited as well. He is the one priest who has an exact correspondent in the Avestan
tradition, the Zaotar, who already in the Gāthās of Zarathustra also composed and
recited hymns.
On the basis of the likely analysis of his name, the Potar was the “purifier,” likely
a purifier of the soma (cf. TB I.4.8), but his R̥ gvedic function is unclear. In the clas-
sical soma rite, the Neṣṭar, the “leader,” is associated with the wives of the gods and
the wife of the sacrificer. Judging from passages like IV.9.4, in the R̥ gveda he had a
connection with the former, which may have expanded to the latter with the intro-
duction of the sacrificer’s wife into the ritual. In the later tradition the Agnīdh or
“fire-kindler” laid and tended the sacrificial fire, in addition to kindling it. He also
acted as the primary assistant to the Adhvaryu, the “ceremoniant,” the priest prin-
cipally in charge of the ritual acts of the sacrifice. According to II.5.4 the Praśāstar,
“director,” is the priest who knows the “enduring commandments” (dhruvā́ vratā́ ni)
that govern the rite. He was likely the forerunner of the Maitrāvaruṇa priest of the
later tradition, who gives the praiṣa or “command” for a particular ritual act to
take place (Minkowski 1991:  118). And finally, there is the Brahman. According
to Oldenberg (1917:  396), the Brahman in this list corresponded to the later
Brāhmaṇācchaṃsin, an assistant to the Hotar who recited hymns in honor of Indra.
Oldenberg may be right, but the Brahman in II.1.2 may have been the brahmán
elsewhere in the R̥ gveda, a composer and reciter of hymns. His liturgical functions
30 The Rigveda

were probably not so restricted as those of the Brāhmaṇācchaṃsin. Other passages

contain partial lists of priests and mention priests who are not included among the
seven of II.1.2, although for the most part their functions overlap those already
described. One remarkable omission from the above lists of priests is a priestly title
corresponding to the singers of the Sāmaveda. Nonetheless, there were singers at
the R̥ gvedic period, since especially those hymns composed in pragātha meters were
likely chanted rather than recited. The Udgātar, later the principal priest of the
Sāmavedic tradition, is mentioned as a singer of sāmans in II.34.2, and among those
priests already mentioned the Potar may have chanted the Soma Pavamāna hymns
of book IX. Also, it is possible that unlike the later tradition, in which singers were
distinguished from reciters, R̥ gvedic priests both chanted and recited according the
requirements of the ritual.
A number of other terms describe the ritual participants and priests and their
functions in the rites. While the exact role of the R̥ gvedic kaví continues to be
debated, he was linked to the production of poetry and to the mastery of esoteric
knowledge expressed in his words. Our usual translation of kaví as “sage poet”
attempts to capture these two sides of him. Since it has an Avestan cognate kauui,
the term is an old one, and both the roles of the kaví and the kauui undoubtedly went
through many changes in their long histories. Originally, however, the kaví/kauui
may have had specific ritual functions within the forerunners of the Avestan and
Vedic rites. Those functions may have complemented those of another figure, the
uśíj, a term that also has an Avestan cognate, usig. Although anciently the uśíj/usig
may have been a specific priestly office, the R̥ gveda uses the term uśíj to describe
various priests who tend the ritual fire. To give the sense of uśíj as a technical term
for such a category of priests, we have translated it as “fire-priest.” So then, likely
reflecting a very ancient distribution of roles, the uśíj priests were masters of ritual
action, while the kaví was the master of words and knowledge.
Thus, like other aspects of the R̥ gvedic rites, the R̥ gvedic priesthood was not
fixed. For the R̥ gveda we can better understand the names of priests as describing
priestly functions rather than naming professional ritual specialists. Thinking of
them in this way helps us understand why hymns use different names or vary the
names. In a period before the rise of the priestly śākhās, there was less need for
priests to have a determined or single identity. The same people who acted as priests
in the ritual could be warriors or perform other social roles in other contexts.

C. Soma Ritual

As we have frequently had occasion to remark, most of the hymns of the R̥ gveda
were composed for the soma ritual. The central rite of this sacrifice was the prepa-
ration of the soma juice, which was then offered to the gods and shared among
male participants in the rite. In the R̥ gvedic period the stalks of the soma plant
were probably placed on a stone and crushed using another stone or stones. The
extracted juice was either transferred to a vessel that contained water and then
 Introduction 31

poured through or onto a woolen filter to purify it. In either case the soma then ran
into another vessel, in which it was mixed with milk. In the Agniṣṭoma of the classi-
cal tradition, there are three soma-pressings in a single day, although, as remarked
above, this may not have been true for the whole R̥ gvedic period or for all ritual
traditions during the R̥ gvedic period. The R̥ gveda also knows the Atirātra or “over-
night” form of the soma ritual, in which there are still three pressings on one day,
but in which the rite continues across the night. The final offerings of the Atirātra
are then made on the morning of the second day.
One of the perennial problems in R̥ gvedic and Avestan studies has been the iden-
tity of the soma plant or its Iranian equivalent, the haoma plant. In the R̥ gveda
the effect of soma juice on both humans and gods is described by the verbal root
√mad, roughly “exhilarate” or “elate.” By these translations we mean that the soma
juice invigorated those who drank it and heightened their senses in some fashion.
We could be more precise about the effect of soma if we knew from what plant
it was extracted. Early speculation that the soma juice was an alcoholic drink of
some sort clearly missed the mark, since the preparation of soma does not allow
for fermentation and √mad does not mean “intoxicate,” if that implies drunken-
ness and not transport. Of the substantial number of possibilities proposed in
more recent times, two have dominated the discussion. (For a review of the various
theories, at least of the time of its writing, see Houben 2003.) The first is that the
soma plant was a stimulant, and the most frequent candidate for that stimulant
is one or another kind of ephedra. Although not original to him and defended
by other scholars, the interpretation of soma as ephedra was argued with particu-
lar plausibility by Harry Falk (1989), largely on the basis of internal evidence in
the R̥ gveda. Also in favor of this hypothesis are the use of ephedra in Zoroastrian
ritual even in modern times and the discovery of traces of ephedra at various sites
of the ancient Bactrian-Margiana Archaeological Complex, a culture with appar-
ent connections with Indo-Iranian culture. Neither of these discoveries confirms
the ephedra hypothesis, and there have been and continue to be many critics of it.
Another set of proposals envisions the soma as a hallucinogen. This argument was
most famously put forth by Wasson (1968), who identified the soma plant as the
Amanita muscaria, the fly agaric mushroom. Similarly Flattery and Schwartz (1989)
argued that previous attempts to identify the soma/haoma plant had overvalued
the Vedic evidence and undervalued the Iranian. On the basis of the latter evidence,
their candidate for the soma plant was Peganum harmala, mountain rue, which also
has psychoactive properties. Despite Flattery and Schwartz’s admonitions, recent
defenders of the view that soma was a hallucinogen have continued to focus on the
internal evidence of the R̥ gveda. For them √mad implies not so much stimulation
as ecstasy or visionary experience. Stuhrman (2006), for example, cites the hymns’
light imagery and the unexpected associations made by the poets to argue that these
are best explained as reflexes of hallucinogenic experience.
This is not an issue that we can resolve, and we would leave it aside if we could.
But the identification of soma affects the interpretation of some hymns and
32 The Rigveda

particularly the translation of the various forms of the root √mad. In general, we
find more textual evidence to support the interpretation of the soma juice as a
stimulant than as a hallucinogen. Neither the imagery of the poems nor the vision
of the poets requires a hallucinogen to explain them. Our view of the hymns is that
they are careful, often intricate compositions that attest to the skill and imagination
of the poets. There is no need to assume the poets experienced the effects of a hal-
lucinogen, and some reason not to do so. To explain what is bizarre and obscure in
these hymns by pharmacology can inhibit the effort to see the underlying logic and
intention of the hymns. While there is much that remains obscure in the R̥ gveda,
interpreters of the text have been able to make progress by the simple assump-
tion that the hymns do make sense and that the poets did know exactly what they
were doing.

D. Other Rituals

In describing Vedic ritual we have thus far been discussing the soma sacrifice. But
while the soma sacrifice dominates the R̥ gveda, the collection includes hymns com-
posed for other rites as well. One of the sub-rites of the classical soma ritual is the
Pravargya, which according to most ritual sūtras is performed twice daily on the
three days leading up to the soma-pressing day. At an earlier stage, likely repre-
sented by the R̥ gveda, the Pravargya was an independent rite, which was only later
incorporated into the soma tradition. Although the Pravargya is not mentioned by
name in R̥ gveda, the text refers to the rite at the center of the Pravargya, an offering
to the Aśvins of gharmá, a mixture of hot milk and ghee, and the underlying verbal
lexeme prá √vr̥ j is used in ritual context.
A number of other hymns, especially in book X, also refer to a Rājasūya, a royal
consecration rite, and other rites defining and affirming royal sovereignty. The cen-
tral rite in the classical Rājasūya was the royal unction, in which water was poured
on the king. This water conferred on the king royal power and authority. Again, the
unction rite itself is not mentioned, but the symbolism of water as a substance that
confers power on the king and over which a successful king must exercise power is
very much part of the R̥ gvedic tradition. While R̥ gvedic kingship may have differed
significantly from later kingship, other classical symbols of kingship also appear
in the R̥ gveda, such as the association of the king and the sun (cf. X.121) and the
identification of king with Varuṇa and Indra, who represent two kinds of sover-
eignty (cf. IV.42).
The animal sacrifice, either as an independent rite or as a part of the soma sacri-
fice, is not very prominent in the R̥ gveda and is generally alluded to by its parapher-
nalia, actions, and verbal accompaniment rather than treated directly. The most
salient piece of equipment is the yūpa, the post to which the sacrificial animal is
tied. The yūpa is celebrated in a single hymn, III.8, which is actually a composite of
two parallel hymns, one appropriate to a single post for a single sacrificial animal,
 Introduction 33

one to multiple posts for several victims. The hymn concerns only the preparation
of the posts; there is no mention of the sacrificial victim(s).
One of the most dramatic actions in the later animal sacrifice is the threefold
circumambulation of the victim by a priest, the Āgnīdhra, carrying a firebrand.
When in some R̥ gvedic hymns the ritual Agni is said to “go around,” the refer-
ence seems to be to this circumambulation—though again the victim is mentioned
barely or not at all (see esp. IX.92.6, 97.1, where soma’s circling of the filter is com-
pared to the Hotar’s circling of “the fixed seats provided with [sacrificial] beasts”;
see also I.173.3, VII.18.22, X.22.14). Elsewhere (VI.1.3, V.43.7) the first and most
highly prized part of the sacrificial beast to be ritually offered in later śrauta ritual,
the omentum, is alluded to, again without direct mention of the sacrificing of the
animal.
The most significant representation of the animal sacrifice in the R̥ gveda, albeit
again indirect, is found in the litanies known as the Āprī hymns. In the later ritual
the Āprī litany accompanied the fore-offerings of the animal sacrifice. The R̥ gveda
contains ten such hymns; in them a set series of subjects or key words—ritual per-
sonnel (e.g., Tvaṣṭar), qualities (e.g., “well kindled”), or equipment (e.g., ritual
grass)—are treated in a fixed order, though with variable wording, generally in
eleven or twelve verses. One of the pieces of equipment is the yūpa, the wooden post
mentioned above, though under the epithet “Lord of the Forest” (=tree). But the
actual sacrifice of an animal is never mentioned in these hymns; at best the victim is
delicately referred to as an oblation.
There is also indirect reference to the animal sacrifice in mythological allusion to
the Śunaḥśepa story, in which the young brahmin boy so named is almost offered
up as a sacrifice to the gods, though he is rescued at the last minute. Although the
story is best known from the very full narrative in the Aitareya Brāhmaṇa (VII.13–
18), his dramatic release is alluded to in the R̥ gveda quite explicitly (see esp. V.2.7,
also I.24.12–13).
Perhaps the most extravagant and dramatic ritual in the classical śrauta system
is the great royal sacrifice, the Aśvamedha or Horse Sacrifice, to be performed by
a king to consolidate or display his power. It involves letting a stallion roam at will
for a year, accompanied by armed troops who fight the sovereigns of any territory
into which the horse strays. When the horse returns at the end of the year, it is
sacrificed, along with numerous other victims, with due pomp but also with almost
unimaginably outlandish accompanying actions. At the climax of the ritual the
chief wife of the king has sex (or simulated sex) with the just-slaughtered horse on
the ritual ground, while the other queens and their female attendants circle around,
singing and dancing and trading obscene jokes with the officiating priests. Two late
R̥ gvedic hymns (I.162–163) directly treat the Horse Sacrifice, although the later
sexual extravaganza is either unknown to them or, more likely, delicately omitted
from discussion. The first (I.162) describes the (literally) gory details of the sacrifice
itself, while commending the sacrificed horse and all its gear to the gods, while the
second (I.163) lavishly lauds the horse and identifies it with the sun on its journey
34 The Rigveda

to the gods. Moreover, if we are correct in our interpretation of X.86, the famous
salacious three-way conversation among Indra, his wife Indrāṇī, and Indra’s pal,
the monkey Vr̥ ṣākapi, this hymn is a burlesque or parody (though a serious one) of
the Horse Sacrifice, with the monkey standing in for the horse-victim, Indrāṇī for
the wife who must mate with the animal victim, and Indra for the king and sacri-
ficer who stands by and watches this mating. This interpretation presupposes that
the dramatic sexual aspect of the Aśvamedha was already present in the R̥ gvedic
version of the rite and is simply not mentioned in I.162–163.
In addition to the rituals that will be codified in the later classical śrauta system,
the R̥ gveda marginally treats rituals that will form part of the gr̥ hya, or “domestic,”
ritual system, primarily life-cycle rites. The R̥ gvedic treatments are almost exclu-
sively found in late portions of the text, in Maṇḍala X, and often have parallels
in the Atharvaveda. The funeral is treated in a series of hymns (X.14–18) in the
Yama cycle. Of particular interest are X.16, which concerns the cremation fire and
the actual burning of the dead man’s body, and X.18, which describes the funeral
service, the burial, and the return of those still living to their lives. Verses 7–9 of
X.18 have attracted special attention because they appear to depict the widow of
the dead man, first lying beside the dead man and then being recalled to life and to
remarriage—thus suggesting that while the later institution of satī or widow burn-
ing is not attested in the R̥ gveda, the ritual representation of the widow’s ceremo-
nial death (though followed by ceremonial rebirth) could have provided a model for
a more literal enactment. Many of the verses in this group of hymns are found also
in the Atharvaveda funeral hymns, XVIII.1–4. Another hymn, X.56, describes the
ascent of the body of the deceased by means of the cremation fire and its transfor-
mation into an immortal body in heaven (see Brereton forthcoming a).
A long and episodic hymn, X.85, is devoted to the wedding. Many of its verses
are found also in the Atharvaveda wedding hymns and are utilized in the gr̥ hya sūtra
protocols for the wedding ceremony. After a long mythological prologue, the hymn
proceeds (sometimes in jumbled fashion) from the wooing of the bride-to-be to the
journey of the newly married couple to the new home, with a very interesting (and
barely comprehensible) treatment of the deflowering of the bride.
The R̥ gveda also provides a certain amount of evidence for the institution of the
Svayaṃvara or bridal “self-choice” marriage, familiar to anyone who has ever stud-
ied first-year Sanskrit and read the Mahābhārata story of Nala and Damayantī,
whose engagement took that form. Most of the R̥ gvedic evidence is mythological,
as the archetypal divine bride in the text is Sūryā, daughter of the sun, who exer-
cises her choice (see especially Jamison 2001), but there is also lexical evidence (see
Jamison 2003) as well as incidental imagery, especially in Dawn hymns, of young
girls in such a situation.
Pregnancy and birth, a major preoccupation in the later gr̥ hya material, is barely
represented in the R̥ gveda. There is a charm for safe childbirth (X.184), which fol-
lows immediately on a brief dialogue between husband and wife attempting to con-
ceive (X.183), and in a short series of hymns against disease we find one against the
 Introduction 35

dangers of miscarriage (X.162). Maṇḍala V also contains a charm for safe child-
birth embedded in an Aśvin hymn (V.78.7–10) and motivated by the mythological
case of (the male seers) Atri and Saptavadhri, trapped in tight quarters like a child
in the womb.

V. The Gods

A. Nature of the Gods

The great majority of R̥ gvedic hymns are dedicated to individual gods or to groups
of gods. Following a classification articulated already in I.139.11, Yāska (Nir.
VII.5) divided the gods into three categories: gods of the earth, gods of the mid-
space, and gods of heaven. This simple scheme has the advantage of being clear
and the disadvantage of being misleading. For while the R̥ gvedic pantheon includes
deities who represent the visible realities and powers of the natural world, it is not
fully composed of such gods. A better starting place for most of the gods is their
names. As we shall see, the R̥ gveda attaches great importance to the names of gods.
By invoking, varying, and meaningfully placing gods’ names in their verses and by
echoing the sounds of those names, the poets bring about the presence of the gods,
their epiphany. More than in any other single feature, the essential nature of a deity
is expressed in the god’s name. The god is who the god is because the god obeys
the truth embedded in that name. Thus the goddess Dawn is the Dawn because
she adheres to the truth that she appears in the morning before the sun. The god
Parjanya, Thunderstorm, is the Thunderstorm because he adheres to the truth that
he sends the rain. To be sure, the meanings of the names of some gods, such as
Viṣṇu, the Nāsatyas, or the Maruts, are unclear or controversial, and the names of
other gods, such as the Aśvins (“Horsemen”), imperfectly represent their charac-
ters. By and large, however, the names of gods define their natures and actions. This
is one reason that the gods of the R̥ gveda can appear to be flat, since they generally
lack the complex personalities that the gods of classical India possess. But there is
considerable complexity in the relations among R̥ gvedic deities, in the fusion and
fission of deities, and in the dimensions of gods indicated by different names and
epithets.
Starting with their names, we can classify gods according to their different
spheres of action. This categorization will be porous, since some gods cross bound-
aries of domains and functions, and will be more orderly than R̥ gvedic reality, but
it gives an approximate shape to the R̥ gvedic pantheon. There are at least five cat-
egories of divinities. First there are gods of nature, the powers that represent and
govern natural phenomena and entities, such as Sūrya, the “Sun,” Vāyu “Wind,”
Parjanya “Thunderstorm,” Uṣas “Dawn,” and Dyaus and Pr̥ thivī “Heaven and
Earth.” With some exceptions, the names of these deities are also words for the
phenomena they represent. So, for example, the word sū́ rya can signify either the
36 The Rigveda

sun god or the sun itself, or, rather, it signifies both the sun god and the sun, since
the two are not fully distinguished. However, these gods of nature do not act only
within the natural sphere defined by their names, but enlarge their sphere of action
on the basis of their natural characteristics. In I.115.1, for example, the Sun is called
“the eye of Mitra, Varuṇa, and Agni,” because the Sun, as he transits the sky, looks
down upon the actions of human beings and observes whether they conform to
the ritual and social principles governed by these other deities. In V.80.5 the light
brought by the goddess Dawn disperses not only the physical darkness of night but
also the “powers of darkness,” the dangerous forces at work within the world. And
according to V.83.9 the god Parjanya, Thunderstorm, not only shakes the world,
but also smashes evildoers. Thus the principle of hidden connections and corre-
spondences allows the gods of nature operate in other spheres in manners analo-
gous to their natural functions.
A second category of divinity includes those defined by the social sphere in which
they operate. The most prominent deities in this category are the three principal
Ādityas: Mitra, Varuṇa, and Aryaman. As detailed below, these gods represent the
different principles that define social relations, and they ensure that human beings
act according to these principles. As the gods of nature have functions within the
social world, so these gods of the social world also have functions within the natu-
ral world. They are associated with the sun, as illustrated by I.115.1, and Varuṇa
especially governs the waters, granting them to those who uphold the principles
he represents, withholding them from those who do not. In this way, the processes
of the visible world become the assurance of the reality of the principles of the
social world.
Still other gods are defined by an action or function that their names embody.
Perhaps the most obvious example is Savitar, the “Impeller,” who compels humans
and other living beings to action or sends them to rest. Similarly, there is Tvaṣṭar,
the “Fashioner,” Viśvakarman, the “All-Maker,” and various other minor “agent
gods” as Macdonell (1897:  115)  called them. Perhaps fitting into this category,
albeit awkwardly, are the Aśvins, the “Horsemen.” As the name of these two gods
suggests, they characteristically drive their chariot, and their mobility is a signifi-
cant part of what defines them. They ride to accomplish many purposes: to heal, to
rescue, or even to facilitate marriage. As charioteers and riders, therefore, they move
within a variety of places and spheres.
Fourth are gods who embody aspects of the ritual, a category dominated by
the two gods who, except for Indra, are the most frequently invoked deities of the
R̥ gveda, Agni and Soma. Agni is the sacrificial fire and Soma the central offering
at the principal sacrifice in the R̥ gvedic tradition. The particular significance of
the ritual gods is their accessibility to humans. Various gods can be present at
the sacrifice, but Indra and other gods like Mitra and Varuṇa remain invisible.
Gods like the Wind, the Dawn, or the Sun are perceptible, but they are distant
or amorphous. Agni and Soma, however, are visibly, tangibly present, right in
front of the priests and sacrificers, and their presence can be reliably brought
 Introduction 37

about by human action. They are the representives of the divine within reach of
humans, and therefore they can create the link between gods and humans upon
which the life of both gods and humans depends. So, Agni is the messenger who
conveys the gods to mortals (e.g., I.14.12, III.6.9, IV.8.2, VII.11.5), and he is
the Hotar priest who brings humans’ offerings to the gods (e.g., I.1.1, VIII.60.1,
X.7.5). And likewise, Soma descends from heaven to the human world (IX.61.10,
63.27, 66.30) and, when offered by humans, goes from these humans to the gods
(IX.25.4, 39.1).
The last category belongs to Indra, who stands apart from all the other gods.
Although it might once have had other resonance, the word índra means only Indra,
which makes it not quite unique but a still rarity among the names of gods. The
greatest number of hymns, nearly a quarter of all the hymns of the R̥ gveda, are
dedicated to him. This preeminence in the R̥ gveda is not surprising since the soma
sacrifice is primarily a sacrifice to Indra. Indra and Vāyu are the first of the paired
divinities who receive soma in the morning; Indra alone or with the Maruts receives
soma at midday; and, at least according to some R̥ gvedic traditions, Indra and the
R̥ bhus receive soma in the evening (cf. IV.35.7). Thus, even though the soma sacri-
fice gradually incorporated other rites and other gods, Indra and the offerings to
Indra remained central to it.

B. Devas and Asuras

Beginning with the Vedic prose texts, one of the most enduring mythological struc-
tures is the perpetual conflict between Devas (devá being the Sanskrit word for
“god”) and Asuras, with the two (almost) balanced groups contending with each
other in numerous myths and myth fragments in all sorts of situations. The Asuras
are, as it were, the anti-Devas, with negative traits exactly corresponding to the posi-
tive ones possessed by the Devas. In the various combats depicted, the Devas always
prevail, but only barely. This conflict continues to be prominent in the post-Vedic
religious landscape, as in the well-known story of the churning of the ocean of milk
in which the two moieties fight over the treasures churned up.
An apparent mirror image of this paired opposition is found in Old Iranian in
the Avestan texts, where ahura, the direct cognate of Sanskrit ásura, is the title of
the head of the pantheon, Ahura Mazdā “Lord Wisdom,” and the daēuuas (exact
cognate of Sanskrit devá) are the enemies of all that is good. Although it has always
been tempting to superimpose the Avestan and middle Vedic situations upon each
other, the R̥ gveda makes serious difficulties. There the term ásura is generally in the
singular, used as a title (“lord”) in a positive sense, and is often applied to divini-
ties who are otherwise identified as Devas. A particularly striking example is found
in VIII.25.4 where Mitra and Varuṇa are called Devas and Asuras simultaneously
(devā́ v ásurā “[the two] Devas, [the two] Asuras”). The Asuras as a defined group
only begin to appear in the late R̥ gveda. For further discussion see Hale (1986). The
38 The Rigveda

history and significance of the Avestan/post-R̥ gvedic mirror-image pairing of the

two terms remain unclear.

́ e DevĀ Ḥ
C. Visv

The term víśve devā́ ḥ “all the gods” or “the All Gods” is common in the R̥ gveda,
and the Anukramaṇī identifies the divinity of a large number of hymns as víśve
devā́ ḥ. The term, both in its usage in the text and in its application to a hymn type,
is employed in a number of different senses. On the one hand, it is a handy way to
refer to the whole divine community, to ensure that no god has been left out of a
generic eulogy or request for aid. In this usage the gods are not treated as individu-
als but as an undifferentiated group, opposed to mortals or, later, the Asuras (see
just above). This group ultimately becomes conceived of as a sort of corporate
entity, the All Gods. On the other hand, many Viśve Devāḥ hymns do not encom-
pass the whole group, but name a series of individual gods, each one often allotted
a single verse in a list hymn (e.g., VI.49). Here the phrase “all the gods” is a way
of indicating that the hymn is not targeting a single god, as in the majority or
R̥ gvedic hymns, but selecting from the group. And finally a number of hymns with
the Anukramaṇī designation “all gods” actually have very little to do with the gods
at all, but contain meditations on the mysteries of the cosmos, of the sacrifice, or of
the powers of poetry and ritual speech (see, e.g., I.105).

D. Indra

As the preeminent god of the R̥ gveda, Indra has a variety of roles. But first of
all Indra is a warrior, upon whom depend the protection and prosperity of his
worshipers. His weapon is the vájra, the mace. In later tradition, when Indra was
reduced to a storm god, the vajra became a thunderbolt. But in the R̥ gveda it was a
weapon, which could be thrown at an enemy or smashed down upon him, and the
principal means by which Indra asserted his power.
The foremost story of Indra in the R̥ gveda is the narrative of the battle between
Indra and Vr̥ tra. Vr̥ tra was a gigantic cobra, who was twisted around a mountain
that enclosed the waters. In order for life to exist Vr̥ tra had to be destroyed. Indra
battled the serpent, alone, according to some hymns, or with the help of the Maruts
or other gods, according to others. After a furious battle Indra killed Vr̥ tra with his
mace and smashed open the mountain, releasing the waters. This myth is occasion-
ally merged with others, so that not only the waters but also the cattle and the sun
emerge from the mountain. The name Vr̥ tra means “obstacle,” and one of the char-
acteristic epithets of Indra is vr̥ trahán, which can mean either “smasher of Vr̥ tra”
or “smasher of obstacles.” There is little difference between these two interpreta-
tions, however, since Vr̥ tra is the paradigm of all obstacles. To evoke Indra as the
smasher of Vr̥ tra, therefore, is to evoke him as the god who smashes all obstacles.
 Introduction 39

The narrative of the destruction of Vr̥ tra is associated particularly with the midday
soma-pressing, which is dedicated to Indra alone or to Indra and the Maruts.
The Vala myth is the second great narrative of Indra and a complement to the
Vr̥ tra story. According to this myth a group called the Paṇis captured the cattle
and kept them trapped in the Vala cave. Indra opened the Vala cave and released
the cattle and the dawns. Remarkable in this story is that Indra does not release
the cattle using his mace as his weapon, but rather using the power of the truth in
the songs he chants. Accompanying him and joining him in his chant are groups
of priests, the Aṅgirases, sometimes along with the Navagvas or the Daśagvas. In
this narrative, therefore, Indra is a priest-king rather than a warrior-king as he is in
the Vr̥ tra myth. In his role as priest-king Indra is also called bŕ̥haspáti, the “lord of
the sacred formulation.” Br̥ haspati appears not only as Indra, but also as a separate
divinity alongside Indra. Gradually, as Indra and the Vedic king, who personifies
Indra, progressively lose their priestly functions in the late R̥ gveda and in the later
Vedic tradition, Br̥ haspati increasingly stands apart from Indra. As the Vr̥ tra story
is connected with the Midday Pressing, so the Vala story was associated with the
Morning Pressing, which takes place with the appearance of dawn. In the R̥ gvedic
period the dakṣiṇā, the reward of cattle to the priests, was also given at the Morning
Pressing. The cattle that come to the priests thus reflect the advent of the cattle and
dawns in the world.
Although the two major mythological narratives with Indra as protagonist are
the Vr̥ tra and the Vala victories, he figures in many other episodes—too many to
mention here—which are often fragmentarily attested and poorly understood.
We may start with his parentage. Although the identity of Indra’s mother is not
clear, in the occasional mentions of her she is a vivid character—as in the snatches
of dialogue between him and his mother in the famous birth hymn IV.18, where she
tries to persuade him not to pursue an unnatural exit from her womb. Elsewhere she
offers him soma to drink directly after his birth, soma that he stole from his father,
named as Tvaṣṭar (III.48; also IV.18.3). And in an even more enigmatic snatch of
dialogue (VIII.45.4–6 ≅ VIII.77.1–2) she seems to reassure the just-born Indra that
he will ultimately prevail. As just noted, Indra seems to participate in a rivalry
with his father, who may be Tvaṣṭar. The unnatural birth and the rivalry with the
divine father are of course well-nigh universal attributes of “the hero”; the many
prodigious feats attributed to Indra just after his birth are also typical of heroic
biography.
Two minor but intriguing myths pit Indra against the two most important
forms of celestial light, the Sun and Dawn. We find the merest allusions (primarily
IV.30.8–11) to a puzzling episode in which Indra crushes the cart of Dawn and she
runs away. Alluded to just a bit more (primarily V.29.5, 9–10; V.31.11; I.121.13) is
the chariot race in which Indra bests the chariot of the Sun, apparently by tearing
the wheel off his chariot. This latter myth is somehow connected with one that is
better attested, though hardly better understood, in which Indra and a sidekick
Kutsa drive on the same chariot, drawn by the horses of the Wind, to the house
40 The Rigveda

of Uśanā Kāvya (a name with a shadowy attestation also in Avestan mythology),

to receive some aid or advice, preliminary to slaying Śuṣṇa, the often-mentioned
opponent of Kutsa.
Indra has a number of other named adversaries. In one striking whiff of a narra-
tive, Indra, aided by Viṣṇu in some versions, shoots a boar named Emuṣa, enabling
him to acquire a special mess of rice porridge hidden in or behind a mountain (see
esp. VIII.77; also VIII.69.14–15, I.61.7). This myth is further developed in Vedic
prose. Another myth with more presence in later texts involves Indra’s slaying of
Namuci by beheading him (e.g., V.30.7–8). In the later versions Indra accomplishes
this by trickery, and part of the trick (using the foam of the waters as weapon) is
already mentioned once in the R̥ gveda (VIII.14.13). The names of other victims
of Indra include Śambara, Pipru, Dhuni and Cumuri, and Varcin, inter alia. The
details of these battles are too sketchy to provide much in the way of narrative
mythology.
Like a number of other gods, Indra has his characteristic draft-animals, and
his are especially prominently featured in the text. His pair of fallow bays (hárī)
conveys him everywhere, especially to and from the sacrifice. They have their own
food offered to them at the sacrifice (roasted grain, see, e.g., III.35), and hymns were
even devoted to a libation made when the pair was hitched up for the return journey
after the sacrifice (see I.61–63). The mention of the fallow bays is sufficient to signal
that Indra is present in the context, and hárivant “possessing the fallow bays” is a
standing epithet of Indra.

E. Agni “Fire”

The word agní is both the common noun meaning “fire” and the name of the god
who is deified fire. As with sóma (see below), it is often difficult to draw the line
between these uses. The sacrificial system of the R̥ gveda (and later Vedic texts), like
that of the cognate Old Iranian Avestan texts, is focused around the ritual fire. The
sacrificial ground is defined by the presence of sanctified fire(s), oblations are made
into them, and the gods and priests gather round them. Thus, first and foremost,
Agni is the god always present at our ritual performances and the immediate recipi-
ent of our offerings. He is the most prominent of the R̥ gvedic gods after Indra, and
all the Family Books and most of the smaller bardic collections open with their
Agni hymns.
Agni as ritual fire is both recipient of oblations in his own right and the conduit
of oblations destined for other gods, which are offered into the ritual fire. He is
therefore regularly called the mouth of the gods, and his role as the middleman
between the human offerers and divine recipients is often emphasized. The flames
and especially the smoke of the fire carry the oblations to heaven, but also, perhaps
more often, serve as a means for the gods to come to earth to our sacrifice: Agni is
said to be the conveyor of the gods many, many times in the text. He is a middleman
in another sense, as a god who nonetheless dwells intimately among mortals. For us
 Introduction 41

he is both ally and messenger to the more distant gods, and since he is not one of us
but a divinity, he is viewed as and often called our guest.
But the ritual fire is not the only form of Agni. The poets emphasize both the
divine aspects of Agni and his purely physical form, often intermingling references
to different forms of fire in the same hymn. As a god he is often identified with the
sun, the celestial form of fire: blazing hot, shining bright, and appearing at the same
time of day, namely dawn when the sun rises and the ritual fire is kindled. But the
fire on our sacrificial ground is also clearly kin to the fire on our domestic hearth;
indeed in later śrauta ritual the fire from which the other ritual fires are taken out
is called the Gārhapatya or “householders’ (fire).” Agni is therefore also praised for
his contribution to daily life and the pleasures of home and family.
The potentially destructive aspects of fire are not forgotten, however. Many of
the most inventive descriptions in Agni hymns are of the wild, uncontrollable ram-
pages of forest fire, spreading across the land and “eating” everything in its path.
We seek to harness this destructive power of fire, to turn it against our enemies
and other threats to our safety, and Agni, sometimes with the epithet rakṣohán
“demon-smasher,” is urged to turn his relentless flames against opponents we name.
A subtype of destructive fire is the funeral fire, the “flesh-eating” fire of cremation,
which is both welcomed and feared (see esp. X.16).
The paradoxical nature of physical fire also provides some part of the god
Agni’s personal qualities. That fire is fueled by plants, especially wood, contributes
to the belief that Agni lives concealed within the plants, even very juicy ones, until
his birth. Agni also comes to be identified with a minor divinity going back to
Indo-Iranian times, Apām Napāt “child of the waters,” who was probably origi-
nally separate—a glowing fiery being concealed and nurtured in the waters, prob-
ably configured in part as lightning.
The creation or birth of the ritual fire from the kindling sticks, his parents, is
a major subject in Agni hymns, with intricate descriptions of the first stirrings of
flame and smoke as the friction of the kindling sticks produces sparks that finally
catch. The just-born Agni is depicted as a tender babe, who quickly grows to become
stronger than his parents and to devour the plants from which he was born.
Many aspects of Agni are expressed through the variety of names and epithets
applied to him. Agni is Jātavedas as the fire established at the beginning of the rite
that continues to its end. As an unbroken presence in the ritual, Agni Jātavedas
also oversees the succession of generations, ensuring that a family’s lineage will
continue. Agni Vaiśvānara is the fire become the sun. As the sun, this fire sees every-
thing and governs everyone. This form of Agni is especially associated with the
king, who like the sun stands above and reaches all beings. The word vaiśvānará
means the one “relating to all men.” Agni is also Tanūnapāt and Narāśaṃsa. One or
another of these names—or sometimes both (I.13.2–3)—appear in the Āprī hymns,
which are recited in an animal sacrifice, and they both occur outside of the Āprī
hymns as well. The word tánūnápāt describes Agni as the “son of himself,” and
nárāśáṃsa as the one “who embodies men’s praise” of the gods. As Agni Kravyād,
42 The Rigveda

the “flesh-eating fire,” Agni is the fire of the funeral pyre that consumes the body
of the deceased and transports it to heaven. Mātariśvan is sometimes identified as
Agni himself, but he is more properly the one who brought the fire from heaven.
Agni participates in almost no narrative mythology, in strong contrast to Indra.
Besides the very sketchy account of Mātariśvan’s theft of fire from heaven, there is
one, ritually connected, tale—of Agni’s flight from the sacrificial ground and his
self-concealment in the waters, to avoid his ritual role as bearer of oblations to the
gods. The gods find him in his hiding place and coax him back by promising him a
share of the oblations. This myth is treated most fully in the late sequence X.51–53,
but there are glancing mentions of it elsewhere. The story may have in part been
generated by the conflation of Agni with the originally distinct divine figure Apām
Napāt “Child of the Waters,” on which see II.35.

F. Soma

Like Agni, Soma is both a god and a crucial ritual substance, and the boundary
between them is not always clear. As has already been discussed, the juice of the
soma plant (whatever that may have been), pressed from the plant and elabo-
rately prepared, is the chief offering of the most important complex of rituals, the
soma sacrifice. This sacrificial substance and its ritual preparation go back to the
Indo-Iranian period, since Avestan attests to the substance haoma, an exact cognate
to Sanskrit sóma, and to its pressing and offering (see especially the so-called Hōm
Yašt, Y 9–10). In both traditions the substance is also deified.
The “Soma Maṇḍala” of the R̥ gveda, Maṇḍala IX, contains 114 hymns dedi-
cated to Soma Pavamāna “Self-Purifying Soma.” These hymns focus entirely on a
single ritual moment, the pressing of the plant, the straining of the juice by pour-
ing it over a sheep’s fleece to trap the impurities (twigs and the like), the mixing
of the juice first with water and then with milk, and the pouring into containers
prior to offering it to the gods, especially Indra. These actions are often presented
metaphorically, with Soma conceptualized as a king making a royal progress across
the filter and into the cups, a progress that can be compared to the conquering of
territory. Or as the Sun in his journey through the cosmos. Or, quite often, as a bull
racing to mate with a herd of cows, who represent the milk with which the juice will
be mixed. Soma is thus regularly presented as having agency in the many descrip-
tions of the purification of the liquid.
Besides this dynamic deification especially characteristic of the IXth Maṇḍala,
there is little narrative mythology involving the god Soma. The most important
tale is the theft of Soma from heaven, where he was confined in a citadel guarded
by an archer called Kr̥ śānu. A falcon stole him and brought him to earth, success-
fully evading serious injury from Kr̥ śānu’s arrow, to deliver him to Manu, the first
sacrificer. This exploit is mentioned a number of times in the text, but is most fully
described in IV.26–27.
Though one characteristic of Soma in later texts, a commonplace already in
middle Vedic, is his identification with the moon, this equation is only attested in
 Introduction 43

the very late R̥ gveda. It is clearly found only in the Wedding Hymn (X.85), whose
first verses depict the wedding of Soma and Sūryā, daughter of the Sun. The bride-
groom Soma in this hymn has clear lunar qualities, which are distinguished from his
identity as an earthy ritual substance.

G.  Aditi and the ā d ityas

As a group, the Ādityas generally represent the powers that order human society.
This function is most evident in the three principal Ādityas, Varuṇa, Mitra, and
Aryaman. In addition to these three, however, there are minor deities who are also
called Ādityas—Dakṣa, Bhaga, and Aṃśa—and a number of other gods, such as
Savitar and Sūrya, who may be called Ādityas when they exercise functions like
those of the major Ādityas.
The Ādityas are sons of the goddess Aditi, whose name means “offenselessness”
or “innocence.” She embodies obedience to the principles of right social behavior
that her sons represent. Later, the motherhood of Aditi becomes central to her
identity and she becomes a mother to other deities.
The most prominent of the Ādityas is Varuṇa, whose name is related to vratá
“commandment” and who therefore is the god of commandments. While all the
major Ādityas are kings, Varuṇa in particular represents the authority of the king.
In IV.42, as in the later Rājasūya, the king becomes both Varuṇa and Indra; that is,
as Varuṇa, the king is a judicial authority governing the actions of his subjects, and
as Indra, he is a leader in war. Accordingly, the divine acts of Varuṇa were often
reflected in the functions of the R̥ gvedic king. Like the king, Varuṇa watches over
his subjects by means of his spáśaḥ “spies” (e.g., I.25.13). One of the responsibili-
ties of the king was to ensure the prosperity of his subjects by providing sufficient
water for animals and crops. Therefore, the divine king Varuṇa brings rain (V.85.3–
4) and controls the waters, causing them to flow according to his commandment
(II.28.4). As the king orders the human world, so Varuṇa orders both the human
world and the world at large: the moon and stars appear and disappear according
to his commandment (I.24.10), and he makes a place and a path for the sun in the
sky (I.24.8, V.85.2, VII.87.1, 5). The king maintained the social order by punishing
wrong-doers, and, likewise, poets fear Varuṇa’s anger and his fetters (pā́ śāḥ), with
which he binds those who violate his commandments (e.g., I.24.15, 25.21). Varuṇa
is the master of the truth that governs the actions of things, as the king must be as
well (II.28.6). Given that his kingship complements Indra’s, we might have expected
Varuṇa to have had a greater presence in the R̥ gveda than he does. However, the
R̥ gveda emerged primarily from the soma rite, and the soma rite belongs to Indra.
In the R̥ gvedic period there probably were other rites dedicated to Varuṇa or to
Varuṇa and other Ādityas—there is such a rite in the classical tradition—but these
left little trace in the R̥ gveda.
In most hymns Varuṇa is closely connected to Mitra, with whom he shares most
of his royal functions. Unlike váruṇa, the meaning of mitrá is reasonably certain.
A mitrá was an ally or an alliance, and Mitra is the god of alliances. While Varuṇa
44 The Rigveda

governs relations in which one person has authority over another, Mitra governs
relations defined by mutual obligations. These two kinds of relationships overlap
with one another, so it is not surprising that the functions of Mitra and Varuṇa
likewise often coincide and that the two gods are so often paired. Only one hymn,
III.59, is dedicated to Mitra alone. As the god of alliances, Mitra governs peace
agreements between different people, ensuring that they will take their proper
places (III.59.1, 5; cf. VII.36.2) and remain in them (III.59.6). When other gods
have functions similar to Mitra’s, they may be identified with him. In particular,
Agni is sometimes called Mitra (e.g., III.5.4) or creates a mitrá, an alliance, when he
appears at dawn. The alliance to which such passages refer is the sacrificial alliance
between gods and mortals. Humans offer the truth in their hymns and offer soma,
milk, ghee, and the like as their oblations. In this way, they empower the gods, and
the gods in turn provide what is necessary for human life.
The last of the major Ādityas is Aryaman, the god of the customs of the Āryas.
He therefore represents a third social principle, the customary rules that govern
relations among Vedic tribes and peoples. This principle was especially essential in
a society where the authority of the ruler would not have penetrated deeply into
the daily lives or the households of his people. Among the spheres in which custom
determined behavior was marriage, which created a new social bond between unre-
lated families. Since marriage depended on the recognition of custom, marriage
fell within Aryaman’s governance. While we have presented Aryaman as the god
of customs, Thieme (1938, 1957) and other scholars following him have preferred
to see Aryaman more narrowly as the god governing the rules of hospitality. In the
absence of a state, the Vedic peoples needed to expect Ārya strangers to recognize
and to act according to the customary norms of hospitality. Such norms were criti-
cal in creating the possibility of relations among Āryas and therefore in unifying
them. Aryaman does not often appear apart from Varuṇa and Mitra and shares
their broader roles in maintaining the natural as well as the social world.
Although relatively minor presences, three other gods, Bhaga, the god of for-
tune, Aṃśa, the god of the share, and Dakṣa, the god of (priestly) skill, are also
called Ādityas. Bhaga ensures that people will receive an appropriate portion of
the goods of life. He is often linked with Aryaman and with the expectation for the
prosperity of a marriage. Aṃśa ensures that people will receive the share of goods
owed them, and therefore he is concerned with inheritance. In both cases, the two
gods bring goods to people according to their behavior and family identity, and
that function brings them within the sphere of social principles represented by the
major Ādityas. Like the major Ādityas, Dakṣa is also concerned with right behavior,
but in his case, it is the skilled actions of sacrificers. For further on the Ādityas, see
Brereton (1981).

H. Savitar

Sometimes linked to the Ādityas and especially to Bhaga is the god Savitar. He is
the god who “impels” or “compels” beings—and these can include mortals, gods,
 Introduction 45

animals, and objects. He especially acts at the beginning of night, when he sends
beings to rest (cf. I.35.2, IV.53.3, VII.45.1). But he also commands the end of the
night and the beginning of the day, when he brings forth the sun (I.35.9) and impels
beings to action. Because he is associated with the night, he is also connected with
the generation of offspring, who would be conceived during the night. Savitar is
bright, with golden eyes, golden arms, and golden hands. He stretches out his arms
in a gesture of command (II.38.2). Falk (1988: 17–22) reasonably suggests that his
brilliance during the night and his outstretched arms point to the Milky Way as a
manifestation of Savitar.

I. SŪ r ya

If the Milky Way is the celestial embodiment of Savitar, Sūrya, the Sun, comes
close to being that of Mitra and Varuṇa. The Sun is their eye, for his gaze is wide
(VII.35.8) and falls on everyone (I.50.2). The Sun watches over the good and evil
deeds of humans (VI.51.2, VII.60.2–3) and, so the poet hopes, declares the inno-
cence of the sacrificers to Mitra and Varuṇa (VII.60.1, 62.2). The Sun is the felly
that rolls toward Mitra and Varuṇa (V.62.2) or the chariot that the two gods set
in heaven (V.63.7). Since he is so closely linked to the Ādityas, he himself is called
an Āditya (I.50.13, 191.9; VIII.101.11). His link to the Ādityas is also a link to the
king, who oversees his subjects the way that the Sun oversees all beings (X.121 and
Proferes 2007: 137–41).
However, Sūrya is not associated exclusively with the Ādityas. He is a form of
Agni, Agni Vaiśvānara, and the face of Agni. Not only the Ādityas (IV.13.2) or
Mitra and Varuṇa (V.63.4), but also Varuṇa and Indra (VII.82.3), Agni (X.3.2),
Soma (VI.44.23, IX.86.22), and Indra and Viṣṇu (VII.99.4) are said to have given
birth to the Sun, to have caused him rise to heaven, or to have established his
brilliance.
A number of images depict the movement of the sun through the heavens. The
Sun flies through the air on a chariot pulled by seven horses or seven mares (I.50.8,
9, IV.13.3, V.45.9), or the Sun is a wheel pulled by only one horse, Etaśa (VII.63.2).
The Sun is also the “reddish eagle” (V.77.3) or a falcon (V.45.9), or he flies like a fal-
con (VII.63.5). However, there are relatively few narratives concerning the Sun. One
repeated but mysterious story is that Indra stole or tore off the wheel of the Sun.
He did so in order to help his ally Kutsa in Kutsa’s battle against Śuṣṇa (I.130.9,
175.4, IV.30.4, V.29.10). What exactly Indra accomplished by doing this and how
this helped Kutsa remain unclear.

J. UṢ a s “Dawn”
Dawn is one of the few female divinities in the R̥ gveda and the most prominent
among them. Twenty-one hymns are dedicated to her alone (every maṇḍala but
II, VIII, and of course IX containing at least one), many of them displaying high
poetic artistry and beauty of imagery, and she is mentioned hundreds of times in
46 The Rigveda

the text. She also has an Indo-European pedigree, being cognate with the Greek
goddess Eos and the Latin goddess Aurora.
The femininity of Dawn is one of her defining characteristics. She is generally
depicted as a beautiful young woman, flirtatious and scantily dressed. Since she
embodies the first light of day, she is gleaming and covered with bright ornaments,
and her appearance thus strongly contrasts with that of her dark sister Night, a
much less prominent goddess, though the ceaseless alternation of Dawn and Night
is often remarked on. Her dispelling of the darkness and of fears of night is much
appreciated, as she awakens and rouses everyone to their daily activities. Dawn is
also, not surprisingly, associated with the god Sun, Sūrya, who is often depicted as
following her as her suitor or husband. She is also said to be the mother or possessor
of cows—the cows being the milky sky and rays of light at early dawn (see Watkins
1987 and 2009 for the Indo-European trope of “the milk of the dawn cows”).
Her associations are not all positive, however. Because she heralds every new day,
she reminds men of the unstoppable passage of time and of the aging process, as
well as of the generations of men who used to view the dawn but have passed away.
Dawn’s daily rebirth as an ever-young beauty presents a cruel contrast to the human
condition of change and decay. The poets also often reflect on the paradox that each
Dawn is new but each is the same as the one before and the one that will come after.
The characteristics of Dawn mentioned above are reflections of the universal
nature of dawn, but she also displays culturally specific qualities relating to Vedic
ritual. Dawn ushers in the sacrificial day, especially the kindling of the ritual fire pre-
paratory to the early-morning rites, and the interplay between the natural sources
of light—dawn and the sun—and the man-made one—fire—is often described as
complex and co-determined. Moreover, Dawn is regularly associated with wealth
and its distribution to the sacrificial participants and is urged to give generously to
them. This association between wealth and dawn has no naturalistic source, but
arises from the fact that in R̥ gvedic ritual the dakṣiṇās or “priestly gifts” were dis-
tributed to the priests and poets at the early-morning rites (rather than at midday,
as in classical śrauta ritual).
Despite the vividness of her depiction, Dawn participates very little in narrative
mythology, though there is a briefly alluded to (primarily IV.30.8–11) and extremely
enigmatic tale in which Indra smashes the cart of “evilly angry” Dawn, and she
runs away. Why Indra should turn against this emblem of benevolent femininity
is unclear, but the story is also associated with Indra’s stealing the wheel off the
Sun’s chariot, and both may have to do with the perturbation of regular temporal
sequences.

K. VĀ y u / VĀ ta “Wind”

As his name indicates, Vāyu is an ancient god of the Wind, although verses to Vāyu
that refer to the phenomenon of wind are somewhat rare and oblique. For example,
 Introduction 47

the roar of Vāyu echoes the sound of the wind, his hundredfold (I.135.3) or thou-
sandfold (I.135.1) team reflects the wind’s speed, and the Maruts, who personify
thunderstorms especially the monsoon storms, are born from his belly (I.134.4).
Such characteristics show Vāyu’s close connection to the wind, even though he does
not represent the wind directly.
On the soma-pressing day, Vāyu is the first of the gods to receive the soma
(I.134.1, 6; VII.92.1; cf. II.11.14), which he drinks unmixed (I.134.5, VII.90.1–2).
But Vāyu also arrives with Indra on the same chariot, and the two of them share the
first drink of soma. Just how both Vāyu and Vāyu and Indra have the first drink of
soma is unclear, but, following a suggestion of Oberlies (1999: 155), perhaps Vāyu’s
first drink reflects soma’s symbolic descent through the midspace as it is filtered,
and the first drink of Vāyu and Indra is the first soma libation.
The ordinary word for the wind is vā́ ta, and unlike Vāyu, the god Vāta closely
reflects the character and activity of the wind. He goes shattering and thunder-
ing, raising the dust; he moves through the midspace and is the companion of the
waters. The symbolic features of Vāta likewise reflect the wind. Vāta is the breath
(ātmán) of the gods (X.168.4), and as the lifebreath, he is the father, brother, and
companion of the man whom he makes live (X.186.1–3). Like the Sun and the
Dawn, therefore, Vāta, the Wind, is completely transparent to the natural phenom-
enon to which his name refers.

L. AS ́ v ins

The Aśvins, the two “Horsemen,” are old Indo-Iranian or even Indo-European dei-
ties who have been brought into the soma rite. They are also called Nāsatyas, a
name of obscure meaning and etymology, found already in an ancient Near Eastern
Hatti-Mitanni treaty dating from the fourteenth century bce (in the form Na-ša-at-
ti-i̯ a) and in the Avestan cognate, Nā̊ŋhaiθya. It is probably the older name of
this pair, with the lexically transparent aśvín originally an epithet. The Aśvins are
connected with honey, mádhu, and while soma comes to be called “honeyed” and
“honey,” mádhu was likely in origin a different offering to the Aśvins. They are also
connected with the Pravargya rite and the offering of gharma, hot milk. Because
they are two, the Aśvins find a place particularly in the morning soma offerings,
which are primarily dedicated to the dual divinities Indra and Vāyu and Mitra and
Varuṇa. Reflecting their association with the Morning Pressing, the Aśvins appear
in the early dawn: they come at the break of dawn (I.157.1, VII.72.4), follow the
chariot of Dawn (VIII.5.2), or accompany the dawn (X.61.4). However, they also
receive the last soma offerings in an Atirātra, or Overnight Soma Ritual. Therefore,
even if they were secondarily grafted onto the soma rite, that graft was a strong
one. They are the fourth most frequently invoked deities in the R̥ gveda after Indra,
Agni, and Soma.
As “horsemen,” the Aśvins are chariot riders and drivers, rather than horse
riders. Their chariot is an object of special attention for the poets. It is often
48 The Rigveda

threefold, with three chariot-boxes, three wheels, three turnings (I.118.1–2), and
three wheel-rims (I.34.2). The sacrifice with its three soma-pressings is compared to
a chariot, so the Aśvins’ threefold chariot may represent the sacrifice. Their chariot
is also swift—“swifter than a mortal’s thought” (I.118.1) or than the wink of an
eye (VIII.73.2). Their chariot is drawn by various animals including bulls, buffa-
loes, and horses, but also by birds (I.119.4), geese (IV.45.4), or falcons (I.118.4).
Their chariot flies to many places and makes the Aśvins present in many spheres: in
heaven, earth, and the sea, in the flood of heaven (VIII.26.17), among plants, and
at the peak of a mountain (VII.70.3). The Aśvins’ speed and mobility are essential
for them, for they are gods who rescue people from various dangers and difficulties
in various places and circumstances.
The story of the Aśvins that the poets mention most often is their rescue of
Bhujyu, the son of Tugra, whom his father had abandoned in the sea (e.g., I.116.3).
They also rescued Rebha from the waters, when he was bound, confined, and left for
dead (I.112.5, 116.24, 119.6). They raised up Vandana (I.118.6), although exactly
from what is not clear. They rescued Atri from an earth cleft (V.78.4) and from
threatening heat (I.112.7). They found Viṣṇāpū, who was lost, and restored him to
his father, Viśvaka (I.116.23, 117.7). They restored the youth and vigor of Cyavāna,
who had grown old (I.117.13, 118.6; VII.71.5), and of the aging Kali (X.39.8). They
brought Kamadyū, the daughter of Purumitra, to be a wife for Vimada (I.116.1,
117.20; X.39.7, 65.12) and gave a son to Vadhrimatī, a woman “whose husband is
a steer” (I.116.13, 117.24; X.39.7, 65.12). They restored the sight of R̥ jrāśva, who
had been blinded by his father (I.116.16, 117.17, 18). They replaced the lost foot of
the mare Viśpalā with a metal shank (I.116.15) and made the cow of Śayu give milk
(I.116.22, 117.20, 118.8). They gave a swift, white horse to Pedu (I.116.6, X.39.10),
and they set a horse’s head on Dadhyañc, the son of Atharvan, in order for him to
reveal the honey to them (I.116.12). Not only do they arrange marriage or bring a
child to a marriage, they themselves wed or woo Sūryā, the daughter of the Sun.
While sometimes the husband of Sūryā is Soma (X.85) or Pūṣan (VI.58.4), else-
where she chooses the Aśvins as her husbands (I.119.5, IV.43.2, 6, VII.69.3–4) and
rides with them on their chariot (I.116.17, VIII.8.10).
What the Aśvins do has been relatively uncontroversial. Why they do it and
what is their character have been more difficult questions. Early scholars tried to
place them in the natural world: Yāska cites interpreters who understood them to
be heaven and earth, day and night, and the sun and the moon. Such interpreta-
tions have been largely and rightly abandoned. Early on, Western scholars observed
their similarity and therefore possible genetic relationship to the Greek Dioskouroi.
Both pairs ride or drive horses; both are young men (kou͂roi in Greek, yúvānā in
Sanskrit); both are sons or, in the case of the Aśvins, perhaps grandsons of Heaven
(divó nápātā); both rescue people in trouble; and both are called twins. Focusing
on the last characteristic, Zeller (1990) sought to show that the Aśvins’ acts reflect
above all the fact they are twins. So, for example, she explains their concern with
sexuality and rescue as partly due to their birth. Because they have one mother but
 Introduction 49

two fathers, they themselves are endowed with a greater sexual potency, and because
one of their fathers is mortal, they are closer to humans and inclined to help them.
The circumstances of their birth are not very clear in the R̥ gveda, however, and
it is not certain that they were often considered twins or that their twinship was
their central feature. Along somewhat similar lines, Oberlies (1993) suggests that
the Aśvins as dual divinities can extend between opposites. They are essentially
gods of the intermediate sphere, who facilitate movement between spheres: between
childlessness and birth, death and life, old age and youth, non-marriage and mar-
riage, and so forth. This is a reasonable explanation of the Aśvins, which might be
extrapolated from the R̥ gvedic evidence, but it is not expressed in it.

M. Maruts

The Maruts are a troop of male gods. Though they lack individual identities, they
are quite prominent as a group: over thirty hymns are dedicated to them alone and
several more to them in conjunction with Indra, and they are frequently mentioned
elsewhere. Their character has both naturalistic and social aspects. On the one
hand, they are the embodiments of the thunderstorm, especially of the monsoon,
and many of their aspects reflect this natural phenomenon: like lightning, they are
brilliant and flashing, bedecked with ornaments and glittering weapons; like thun-
der, they are excessively noisy on their wild chariot journeys, causing the earth to
shake with fear, bending the trees and even the mountain; like thunderclouds, they
are shape-shifting and sometimes clothed in gray; and they are accompanied by
floods of rain. The terror they inspire is more than balanced by the fructifying rains
they bring. All these physical aspects of the Maruts often inspire the poets to vivid
and imaginative language.
As a social phenomenon, the Maruts represent the Männerbund, an association
of young men, usually at a stage of life without significant other social ties (such as
wife and children), who band together for rampageous and warlike pursuits. The
violence of the thunderstorm is akin to the violence of these unruly age-mates, raid-
ing and roistering. It is not unlikely that Vedic society contained and licensed such
groups among its young men, given the frequent warfare depicted in the R̥ gveda,
and the divine Maruts provide the charter for this association and behavior.
The Maruts are not, however, entirely without social ties. Their parentage is
clear, though the manner of their birth problematic and disputed—and often
alluded to as a mystery. Their mother is a dappled cow, Pr̥ śni, who can display
androgynous characteristics and behavior; their father is Rudra, and they are often
themselves referred to as Rudras. Moreover, they have a female companion, Rodasī.
When the word ródasī appears in the dual number, it refers to the two world-halves,
but as a singular (also accented rodasī́ ) it is the name of the Maruts’ consort, a
beautiful young woman who accompanies them on their chariot. Their normal
location in the midspace between the two world-halves is presumably responsible
for her name.
50 The Rigveda

Perhaps the Maruts’ most important companion is Indra, for whom they serve
as a sort of posse: marútvant “accompanied by the Maruts” is one of Indra’s stand-
ing epithets. Their major role in dynamic mythology was to provide support and
encouragement to Indra before the Vr̥ tra battle, an episode also treated in Vedic
prose narratives. But, according to one of the most striking hymns in the R̥ gveda,
I.165, a dialogue among Indra, the Maruts, and the seer Agastya, Indra disputed
the extent of their aid at that time. In this hymn Indra and the Maruts argue over
their respective rights to a sacrifice offered by Agastya; Indra asserts his rights in
part because he claims the Maruts abandoned him to fight Vr̥ tra alone, though
elsewhere in the R̥ gveda (and later) there is no doubt about their supportive role in
that combat.
This mythological contretemps has its reflection also in ritual, in fact to a rit-
ual change occurring during the R̥ gvedic period. Although in some of the Family
Books Indra alone is the recipient of the offering at the Midday Pressing, in
Maṇḍalas III and VI, in scattered mentions elsewhere, and in the classical śrauta
ritual, the Maruts share the Midday Pressing with Indra. The tense negotiations
among Indra, the Maruts, and the sacrificer Agastya in I.165 and I.170–171 sug-
gest that the change in recipients of the midday oblation was a contested topic for
R̥ gvedic ritualists and the inclusion of the Maruts needed and was given mythologi-
cal underpinning.

N. Heaven and Earth

One of the most remarkable and satisfying phrasal equations across the older
Indo-European languages is that of Vedic dyaúṣ pitā́ “father Heaven” with Greek
Zeus Pater and Latin Jupiter, thus attesting to a deified paternal Heaven for
Proto-Indo-European as well as the older daughter languages. Ironically perhaps,
the Vedic god, the meaning of whose name is still transparent and lexically addi-
tive, is far less important in the Vedic pantheon than his correspondents in the
Classical languages, where the original semantics have become attenuated or have
disappeared entirely.
In the R̥ gveda, Heaven as a divinity is generally paired with the female Earth,
who is frequently referred to as “mother,” with the two a complementary paren-
tal pair. They are normally grammatically joined in a dual dvandva compound
(dyā́ vā-pr̥ thivī́), and several hymns are dedicated to this couple. If Heaven and
Earth are the archetypal parents, who are their progeny? This is mentioned less
than one might expect, but in a few hymns it is clearly stated that the gods are their
children and especially the Sun. A less beneficent aspect of Heaven’s fatherhood is
found in a myth, obliquely but vividly referred to a few times in the R̥ gveda (I.71.5,
8; X.61.5–7) and told more clearly in Vedic prose (though with Prajāpati substitut-
ing for Heaven)—namely his rape of his own daughter.
Heaven and Earth also give shape to and encompass the cosmos, providing a
safe enclosure within which life can flourish. The separation of the two to create
 Introduction 51

this space is the primal cosmogonic moment, and Indra’s accomplishment of this
separation by propping them apart is endlessly celebrated.

O. TvaṢ Ṭ ar

The meaning of Tvaṣṭar’s name is fully transparent:  he is the “Fashioner,” who

“adorned all the creatures with their forms” (X.110.9) and who knows all living
things (IV.42.3). But his role is more complex than his name might imply. He is the
father of Indra (III.48.2–4, IV.18), whom Indra displaced and from whom Indra
stole the soma (III.48.4). On the other hand, Tvaṣṭar fashioned Indra’s distinc-
tive weapon, the mace, for him (I.32.2, V.31.4, VI.17.10, X.48.3). Tvaṣṭar is also
sometimes called the father of Agni (III.7.4), and he is the god who has begotten
offspring (III.55.19; cf. III.4.9). His role as a father perhaps explains his close rela-
tionship with the Wives of the Gods (I.22.9, 161.4; II.1.5, 31.4, 36.3; VII.35.6), with
whom he is ritually joined.

P. R̥ b hus

In some forms of the soma sacrifice, the three R̥ bhus have a significant role as prin-
cipal soma recipients in the Third Pressing. But despite that role, they have a limited
presence in the R̥ gveda itself. Only ten hymns are dedicated to the R̥ bhus, together
with one other that invokes the R̥ bhus along with Indra. Nonetheless, despite their
decidedly low profile in the R̥ gveda, their principal actions emerge clearly. The
R̥ bhu hymns repeatedly return to five great deeds for which the R̥ bhus are famed.
They took a soma cup made by the god Tvaṣṭar and fashioned it into four cups
(III.60.2; IV.33.5, 35.2, 3, 36.4). They made a chariot, sometimes identified as the
chariot of the Aśvins (I.20.3, 111.1, 161.3; IV.33.8). They created the two fallow bay
horses of Indra (I.20.2, 111.1; III.60.2; IV.33.10, 34.9, 35.5). They fashioned a cow,
or made a cow give milk, or carved up a cow (I.20.3, 110.8, 161.7, 10; IV.33.4). And
lastly, they rejuvenated their aging parents (I.20.4, 110.8, 111.1; IV.33.3, 35.5, 36.3).
Significantly, as a result of these creative acts, the R̥ bhus are said to have attained
immortality or to have become gods.
Their skillful acts are essentially priestly, and their great deeds reflect ritual acts
or, more specifically, ritual acts at the Third Pressing. The four soma cups they cre-
ated are the cups of the four principal soma-drinkers: Indra and the three R̥ bhus.
As mentioned above, the Aśvins’ chariot can represent the sacrifice, and therefore
the chariot they made could be the sacrifice in general. The creation of the fallow
bays of Indra is reflected by a special soma offering in the Third Pressing that marks
the departure of the two horses of Indra. The cow over which they work may rep-
resent the soma stalks from the previous soma-pressings, which are pressed again
at the Third Pressing. The R̥ bhus cause these depleted “cows” to release even more
milk, which is the soma juice. Their last deed, the rejuvenation of their parents, is
more mysterious, but it might represent the return or “rejuvenation” of the Aśvins
52 The Rigveda

at the end of the sacrifice in an Overnight rite or it could reflect the rejuvenation
of the sacrificer and his wife, since the fertility of the sacrificing couple is a theme
of the Third Pressing. For a more detailed discussion of the acts of the R̥ bhus and
their meaning, see Brereton (2012).

Q. PŪ Ṣ a n

Although Pūṣan is a minor god in the R̥ gveda, with only eight hymns dedicated
to him alone and several more shared with more prominent divinities (Indra and
Soma), his idiosyncratic characteristics and the special diction used in his hymns
attract more than his share of attention to him. Of the bardic families, only the
Bharadvājas of Maṇḍala VI favor this god; they dedicate five hymns to him (VI.53–
56, 58) with a further one to Pūṣan and Indra (VI.57) and a significant portion of
the composite hymn VI.48; the three other hymns exclusive to him are found in
I and X.
The characteristics ascribed to him are humble and somewhat countrified: his
draft-animals are goats, his tools generally an awl and a goad, his food of choice is
porridge, and the skills he deploys for us are especially the protection of the roads
and the finding of lost articles, particularly cattle. The level of discourse is often
colloquial and lively, though he is occasionally celebrated in a register more appro-
priate to loftier divinities.
One striking feature does not fit this profile: Pūṣan in several passages is said to
be the husband or consort of Sūryā, the daughter of the Sun, who is the archetypal
bride in the R̥ gveda, and he is also said to be the lover of his sister and the wooer of
his mother (VI.55.4–5), though this apparent incest provokes no blame. The tangled
family relations thus alluded to are not treated in any detail, so we are left with only
tantalizing clues.

R. ViṢ Ṇ u

In the middle Vedic period, Viṣṇu became a central figure as the embodiment of the
sacrifice itself and therefore of a power that can exceed even the might of the gods.
In classical India, of course, he finds an even greater destiny. There is little sign of
those futures of Viṣṇu in the R̥ gveda, since it has only a half dozen hymns dedicated
to Viṣṇu or to Indra and Viṣṇu (I.154, 155, 156; VI.69; VII.99, 100). Viṣṇu appears
alongside Indra fairly often, especially in his battle with Vr̥ tra (IV.18.11, VI.20.2,
VIII.100.12), and he is also Indra’s partner and ally generally. The only acts that are
especially his are his three strides or three steps. With these strides Viṣṇu encom-
passes the earth, and with his third step he disappears into a realm where none
can follow (I.155.4–5). Or he enters into heaven where there is the “wellspring of
honey,” the source of soma (I.154.5), or the highest cattle-pen (III.55.10). He is
therefore the god who is wide-ranging (urugāyá) and wide-striding (urukramá). The
purpose of his strides is to create space and a place for people to live and move
(I.155.4, VI.49.13, VII.100.4). This purpose could explain Viṣṇu’s close connection
 Introduction 53

with Indra in the fight with Vr̥ tra, since Vr̥ tra represents what confines and hinders,
and Viṣṇu’s strides what opens and frees. The strides of Viṣṇu in the R̥ gveda antici-
pate the strides that Viṣṇu takes as Vāmana, the dwarf avatar of classical Hinduism,
and also in the middle Vedic literature Viṣṇu as the sacrifice is a dwarf (ŚB I.2.5.5).
However, there is no direct evidence that Viṣṇu already has the form of a dwarf in
the R̥ gveda.

S. Rudra

Although Rudra, under his euphemistic epithet Śiva, “the kindly one,” has, of
course, an extraordinarily great future in classical Hinduism, in the R̥ gveda he has
a very circumscribed role, with only three complete hymns dedicated to him. He
has two major and complementary characteristics:  on the one hand, he is fierce
and malevolent, with an often inexplicable anger that needs to be appeased; on
the other, he is a healer, who controls the remedies for disease. He is also, as noted
above, the father of the Maruts, who are much more prominent in our text.

T. Sarasvati ̄ and the Rivers

Another divinity with a great future ahead but little prominence in our text is the
goddess Sarasvatī. Again the R̥ gveda provides little or no evidence for her later role
as patron of learning and the arts, though a number of scholars have attempted
to find it. Instead she is celebrated, in the three hymns dedicated to her and in
other mentions in the text, simply as a physical river with a powerful flow, which is
sometimes destructive but which also provides fecundity with its fructifying waters.
Other rivers are also praised in the R̥ gveda, especially the Sindhu in a hymn dedi-
cated to the rivers in general.

U. VĀ c “Speech”

One last goddess should be mentioned, Vāc or deified Speech (a noun with feminine
gender). Although this goddess figures in a number of mythological narratives in
the middle Vedic period, and although she ultimately seems to have become con-
flated with Sarasvatī, thus amplifying the status of both goddesses, in the R̥ gveda
she appears as a clear personage very rarely, most prominently in a late hymn
(X.125), which is a 1st-person self-praise (ātmastuti) spoken by Vāc herself.

VI.  R̥gvedic People and Society

A. R̥ g veda as History

A number of scholars have treated the question of the historical background of the
R̥ gveda and to what extent we can extract historical information from the names
of kings and peoples, the events they are depicted as participating in, and the place
54 The Rigveda

names where these events are depicted as occurring or where the kings and their
retinues are depicted as living. We do not intend to enter into these issues and will
simply make global reference to the many works of Michael Witzel (e.g., Witzel
1995a, 1995b), who has been especially active in discussing these questions in recent
decades.
We will only note here that the R̥ gveda contains a large number of certain or
likely personal names (see Mayrhofer 2003). Besides the poets and their ancestors,
many of the names belong to royal patrons and are therefore especially common
in dānastutis. Some kings have a presence outside of dānastutis, however; particu-
larly noteworthy are the Tr̥ tsu or Bharata King Sudās (“Good Giver”), the victor
in the famous Battle of the Ten Kings treated in VII.18, whose forces also crossed
a formidable barrier of rivers, as depicted in the dialogue hymn III.33, and King
Trasadasyu, whose Royal Consecration may form the subject of the dialogue hymn
IV.42. Others belong to human enemies of the poet and his group, whose defeat the
gods (generally Indra) aid our side in effecting, or to clients of various gods, espe-
cially the Aśvins, who receive help from these divine patrons. Needless to say, the
“reality” of those named, especially in the last two groups, cannot be ascertained.

B. Ā r yas, DĀ s as, and Dasyus

The people of the R̥ gveda refer to themselves as Āryas, which probably meant the
“civilized” ones or something similar. Under this term they define their own group
as the people who sacrifice to the gods, who adhere to Vedic customs, who speak
Indo-Aryan languages, and who in other ways identify themselves with Vedic cul-
ture. They also refer to themselves as mā́ nuṣa and mānavá, the “sons of Manu” or
the “peoples of Manu,” for the legendary Manu (mánu simply means “man”) was
the one who first instituted the sacrifice and was therefore the founder of Vedic
religious culture. They also called themselves the “five peoples” (páñca jánāḥ,
carṣaṇáyaḥ, or kr̥ ṣṭáyaḥ), who lived in the “five directions” (páñca pradíśaḥ)—the
cardinal directions and the center—or in the five lands (páñca bhū́ mā; cf. Proferes
2007: 62). Corresponding to this world of five peoples in five lands, the R̥ gveda also
mentions five major Ārya tribes or tribal federations—the Pūru, Yadu, Turvaśa,
Anu, and Druhyu—who can, at one time or another, be allies or enemies of each
other. Even these major tribes may not have been stable social units, and they are
not the only social groupings mentioned in the R̥ gveda. Including the major tribes,
Witzel (1995a:  313)  lists around thirty social groups named in the R̥ gveda, but
notes that it is not clear which of these were extended families or clans or tribes or
confederations.
The primary social units that made up larger tribal units were the víś. There has
been much discussion about how to characterize the víś (cf. Proferes 2007: 15–16).
Generally speaking, scholars either see the víś as a “clan,” which was composed
of related lineages, or a “settlement” of a kinship group. The former emphasizes
the perceived relation of the people belonging to the víś; the latter, their shared
 Introduction 55

locality. We have translated víś as “clan,” although our primary intention has not
been to adjudicate between these two views but to find a consistent translation. The
head of the clan was the viśpáti, the “clanlord,” who led his clan and represented
it within larger social units, such as a tribe. The clan was itself composed of dif-
ferent extended families, themselves led by a gr̥ hápati or dámpati, a “houselord.”
Sacrificers generally came from the ranks of these clanlords and houselords. Larger
social units composed of several clans were led by a rā́ jan, a “king” or, as others
prefer, a “chieftain,” who was chosen from among the viśpatis by the viśpatis of the
clans that formed these units. These groups of clans then formed parts of larger
tribes or confederations, also led by a king. The R̥ gvedic saṃrā́ j or “sovereign king”
was likely a king whose rule included other social units that also had their kings.
The saṃrā́ j was differentiated from the svarā́ j “independent king,” who ruled with-
out interference from other lesser or greater rulers. While there is good reason and
good evidence to believe that kings were selected by clanlords or lesser kings, there
is also evidence for the lineal descent of kings, at least of the kings of major tribal
confederations (cf. Witzel 1995a: 330, 336).
The nature of a king’s rule was also affected by cycles of settlement patterns.
Periods of fixed settlement (kṣéma) alternated with periods of movement (yóga,
lit. “yoking up”). During the former the clans tended their cattle and raised crops,
more or less in peace, in more or less fixed habitations. During the latter they trav-
eled into new areas to gain new lands or to take cattle from other clans or tribes,
or they confronted others moving into their territories. These two periods of settle-
ment and movement may have been fixed according to the seasons of the year. In
periods of mobilization the clans were governed by a king who could lead them in
battle. This king was an embodiment of Indra, a war-king, a svarā́ j, who required
obedience from his subjects. During times of settlement the king was an embodi-
ment of Varuṇa or of Mitra and Varuṇa, who maintained the peace among his
people. He could be a saṃrā́ j, who ruled, perhaps more loosely, over other rulers.
The “war-king” and the “peace-king” might have been two different people, but
more likely these represent two roles that a king might or must play. According to
our understanding, in IV.42 King Trasadasyu is both Indra and Varuṇa, the king
both in war and in peace.
The Āryas fought among themselves, but their enemies were often groups
of non-Āryas, called Dāsas or Dasyus, who may, or may not, have been
non-Indo-Aryans. The opposition between Āryas and Dāsas or Dasyus was not
an unbridgeable divide. There are many people, clans, and tribes in the Veda who
have names without likely Indo-European derivation. Witzel (1999: 359–60) gives
a “fairly comprehensive list” of Vedic “tribal and (some) clan names” that includes
names from the R̥ gveda. Of these he counts twenty-two that are non-Indo-Aryan
names. The evidence is rough, but it suggests that at some point in their histories
these people had adopted Vedic culture and had become part of the Ārya com-
munity. The distinction between Āryas and Dāsas or Dasyus, therefore, was essen-
tially a cultural and political one. The Dāsas and Dasyus were people who had
56 The Rigveda

not adopted or not yet adopted the customs and behaviors of the R̥ gvedic Āryas
and therefore were not part of the Ārya community. Exactly who the Dāsas and
Dasyus were—as opposed to who they were not—is a more difficult problem. They
must have been people and cultures either indigenous to South Asia or already in
South Asia—from wherever or whenever they may have come—when the carri-
ers of R̥ gvedic culture and religion moved into and through the northwest of the
subcontinent.
According to the evidence of the R̥ gveda the Dasyus are regularly the enemies of
the Āryas, and the poets repeatedly ask the gods’ help against them. R̥ gveda X.22.8
lays out the character of the Dasyu according to the R̥ gvedic poets. He is akarmán
“of non-deeds,” that is, he does not perform the sacrificial rites. He is amantú “of
non-thought” because he does not know the truths formulated in the Vedic hymns
and therefore is unable to articulate these truths. He is anyávrata, one “whose com-
mandments are other” than the commandments of the gods. And he is ámānuṣa
“no son of Manu” and therefore one who does not belong to the Vedic peoples. The
Dasyus are not only other than the Āryas, they are hostile to the Āryas. The poets
accuse them of having cunning tricks or wiles (māyā́ , IV.16.9, VIII.14.14, X.73.5)
that they use against the Āryas, and they call on the gods, especially Indra but also
Agni and Soma, to strike the Dasyus down (VI.29.6), drive them off (V.31.7), or
blow them away (I.33.9, X.55.8). Such Dasyus are human, although some of them
may have been demonized humans or beings on the way to becoming demons.
There is a great degree of overlap between Dasyus and Dāsas, since both names
can be used of the same beings (I.103.3, IV.28.4, V.30.9). Like the Dasyus, the
Dāsas are also humans and usually they are enemies of the Āryas. Indra destroys
them (IV.30.15, 21;VI.20.10, 47.21, X.120.2) and their fortresses (II.20.7, IV.32.10).
However, the use of Dāsa in the R̥ gveda is more complex than that of Dasyu.
Since the greatest enemy of Indra, Vr̥ tra, is a Dāsa (I.32.11, II.11.2, IV.18.9) but
not a Dasyu, the Dāsas apparently penetrated further into the nonhuman realm
as demonic beings. Such a nonhuman Dāsa occurs also in X.99.6, where Indra
“subdued the mightily roaring Dāsa with his six eyes and three heads.” However,
dāsá can mean “servant, slave” already in some R̥ gvedic passages. According to
VIII.56.3, a man named Dasyave Vr̥ ka, “Wolf to the Dasyu,” has given to the poet
“a hundred donkeys,” “a hundred wooly ewes, a hundred slaves (dāsá), and garlands
beyond that” (cf. also VII.86.7, X.62.10). These dāsás were obviously not enemies
of the Āryas, at least not as long as they were subordinate to them. The R̥ gveda also
shows less insistence on the Dāsas’ cultural difference from the Āryas than on the
Dasyus’—Dāsas are not described as akarmán, amantú, anyávrata, ámānuṣa, and
the like. However, the poets sharply distinguish between Āryas and Dāsas (V.34.6,
VI.25.2, X.86.19) and worry that the Dāsas have wealth that should belong to Āryas
(II.12.4). Yet they also can have ties to the Āryas. In VIII.46.32, a dānastuti verse,
the poet mentions a wealthy Dāsa named Balbūtha Tarukṣa, from whom he says
he received a hundred camels. Although Balbūtha’s name is not Indo-Aryan and
although he is called a Dāsa, he had apparently employed the poet, presumably to
 Introduction 57

compose hymns and to sacrifice for him. Therefore, he must have had one foot in
Ārya culture, if not quite in the Ārya community.
In summary, the Dasyus and Dāsas are overlapping categories of peoples
opposed to the Āryas, and the poets call on the gods to defeat them for the sake
of the Āryas. However, sometimes Dāsas may have been rivals to the Āryas or may
even have been at the fringes the Ārya community rather than inevitable enemies of
Āryas. For a thorough discussion of the attestions of dásyu, dāsá, and dā́ sa in the
R̥ gveda and later Vedas, see Hale (1986: 146–69). The above summary is very much
indebted to Hale’s work, but Hale is inclined to see a racial distinction between the
Āryas and the Dasyus or Dāsa that is not justified by the evidence.

C. Social and Political Organization

The R̥ gveda is the product of an elite segment of society, and it no doubt reflects
only a small part of even elite religious life. For example, the life-cycle rituals that
are so important in the late Vedic period, chronicled in the gr̥ hya sūtras, are only
tangentially treated in the R̥ gveda, and we get only glimpses of the dharmic pre-
scriptions that later ordered daily life, as preserved in the dharma sūtras of the
late Vedic period and the dharma śāstras that followed them. The religious beliefs
and practices of the non-elite are completely absent from the text, except perhaps
in obscure and slighting references to practices that do not conform to R̥ gvedic
standards.
Outside of the religious sphere we have almost no direct evidence of social or
political organization and very little information about how people, ordinary or
elite, spent their days. We learn a little about the various trades plied by members
of society, mostly by way of incidental similes or poetic images. Most of what we
learn in this sphere is about stock-raising, since the cow and the bull are both such
powerful symbols in the poetry. We also learn a bit about the leisure pursuits of the
elite, especially dicing and horse racing.
All of this is to say that it is unwise to use the evidence of the hymns uncriti-
cally to speculate on Vedic society. Not only does the text concern a very small
percentage of the population, but even in that population its focus is very narrow.
Moreover, everything we learn is shaped by the pragmatic purpose of the hymns as
well as by the poetic sensibilities of their composers. Nonetheless, we can venture
some very general remarks about social organization.
There is no evidence in the R̥ gveda for an elaborate, much-subdivided, and over-
arching caste system such as pertains in classical Hinduism. There is some evidence
in the late R̥ gveda for the fourfold division of society into varṇas, the large social
classes so prominent in the later legal texts. But even this system seems to be embry-
onic in the R̥ gveda and, both then and later, a social ideal rather than a social real-
ity. The clearest evidence for it is found in the so-called Puruṣasūkta or “Hymn of
the Man” (X.90), in which the body parts of the Ur-man correspond to the four
varṇas, hierarchically arranged (vs. 12): the brahmin is his mouth, the kṣatriya (there
58 The Rigveda

called rājanya) his arms, the vaiśya his thighs, and the śūdra is produced from his
feet. But this hymn is generally considered to have been a quite late addition to the
text, perhaps to provide a charter myth for the varṇa system after it had taken more
definite shape. Otherwise, the late R̥ gveda provides some evidence for the begin-
nings of a formal contrast between brahmaṇic and kṣatriyan powers; for example,
in the final verses of VIII.36 and VIII.37, which are identical save for bráhmāṇi
“priestly formulations” in VIII.36.7 and kṣatrā́ ṇi “lordly powers” in VIII.37.7; the
hymn preceding this pair contains a tr̥ ca (VIII.35.16–18), in which each of the three
verses calls for blessings appropriate to one of the three upper varṇas (though they
are not named as such).
The rest of the R̥ gveda does attest to a division of labor and complementary and
reciprocal relationship between rā́ jan “kings” (whatever form this kingship took at
this period) and the poets and priests who performed their sacrifices and composed
the accompanying poetry (as well as the secular royal encomia whose existence we
can posit though we have no direct evidence for them [see Jamison 2007: chap. 4,
esp. 146–48]). This poet–patron relationship is especially on view in the dānastutis
that were already discussed above. The status and pursuits of the “producers” or
vaiśyas are barely and glancingly alluded to in the text, primarily in similes and
the like.
We might here spare a few words for the creatures who are otherwise invisible in
the social and political structures, namely women. As is quite common for ancient
societies, we don’t know much, but the few female figures that appear in the text
tend to be quite vivid. Given their general absence from the R̥ gveda, females appear
disproportionately as speakers in dialogue hymns—both divine and semi-divine fig-
ures such as Indrāṇī, wife of Indra (X.86) and the Apsaras Urvaśī, once married to
the mortal Purūravas (X.95), or human or semi-human women such as Lopāmudrā,
wife of the legendary seer Agastya (I.179), or Yamī, the first (almost) mortal with
her twin Yama (X.10). The females in these hymns are quite outspoken, usually
about sex, and their male conversation partners tend to look weak and helpless in
comparison. For possible linguistic features of women’s speech as represented in the
R̥ gveda, see Jamison (2008a, 2009b, 2009c).
But none of these female speakers is depicted as a real, contemporary woman,
and what we know of that class is extremely limited. Beautiful sexy women are
sometimes recorded by the poet as a particularly appealing feature of the gift cel-
ebrated in his dānastuti, and Dawn is often compared to everyday women—either
good, eager wives or not-so-good, eager courtesans. One must make allowance for
male fantasy at this period, as in so many others. Mothers are tenderly described,
but in generic fashion, and we also learn something about the contracting and sol-
emnizing of marriage, both in the wedding hymn (X.85) and in the numerous men-
tions of the wedding of Sūryā, daughter of the Sun, which appears to have been
of a Svayaṃvara (“self-choice”) type, familiar from later Sanskrit literature, par-
ticularly the weddings of Damayantī and Nala and of Draupadī and the Paṇḍava
brothers in the Mahābhārata.
 Introduction 59

The Anukramaṇī attributes a few hymns to females, for example XIII.91 to Apālā
Ātreyī and X.39–40 to Ghoṣā Kākṣīvatī, but these ascriptions are derived from the
personnel depicted in the hymn itself. There is no reason to assume that the poet
was female in these cases. A particularly egregious example is the attribution of the
very interesting X.109 to Juhū Brahmajāyā, or “Sacrificial Ladle, Wife of Brahma [/
the brahmin],” based on the appearance of both those terms in the text.
Although there is no real evidence for female poets, there is evidence in the late
R̥ gveda for women in a ritual role, that is, as Sacrificer’s Wife (patnī). This is a standard
and required role in the classical śrauta ritual of the middle Vedic period (discussed
extensively in Jamison 1996a), and it appears to be a ritual innovation, much dis-
puted, in the late R̥ gveda (see discussion in Jamison 2011 and forthcoming a and b).

VII.  Language and Poetics

A. Grammatical Structure and Language Use

Sanskrit is an inflectional language with an especially rich morphology, and most of

the grammatical information that is carried in English by word order and by separate
elements such as prepositions and auxiliary verbs is coded in Sanskrit on the word
itself. R̥ gvedic poets glory in their grammar and are skillful in exploiting not only the
many distinctions it provides but also grammatical ambiguities and neutralizations
of grammatical distinctions. Moreover, since basic information, such as the identity
of the grammatical subject and object, is coded on the word, the poet is free to use
word order for rhetorical purposes, placing particularly significant words in emphatic
positions such as initial in the verse line. When possible, we have tried to reproduce
these effects, but all too often, given the very different structure of English, the result
of a well-intentioned effort to follow the Sanskrit word order is at best awkward and
artificial, and at worst unintelligible. This is discussed further in section VIII below.
Even the briefest sketch of Vedic grammar would be out of place here. We refer
the interested reader to Jamison (2004b) for a very skeletal account and to any
standard Sanskrit reference grammar for more details. Macdonell’s Vedic Grammar
for Students (Macdonell 1916)  is admirably clear and informative, and the clas-
sic Sanskrit reference grammar is Whitney (1889). However, in some of the hymn
introductions we do make reference to certain grammatical facts, and some of these
we will mention here. The Sanskrit noun has eight grammatical cases, expressing
most of the syntactic relations pertaining in a sentence, including subject (nomi-
native case), direct object (accusative case), possessor (genitive case), and so on.
The noun also has three numbers, singular, dual, and plural, and three genders,
masculine, feminine, and neuter. The verb is even more complex. Its basic unit is
a nine-member grid of three persons (1st [I/we], 2nd [you], 3rd [he, she, it, they])
and three numbers (matching those of the noun) of the subject of the verb form
(Table I.1).
60 The Rigveda

Table I.1.  Sanskrit Verb Forms.

I we two we (all)
you (sg.) [“thou”] you two you (all)
he, she, it they two they (all)

This basic unit is deployed in a plethora of tenses and moods, including at least
three ways to express the past tense (the imperfect, aorist, and perfect tenses). Of
these three, the aorist is often used to express the immediate past (in English, “has
[just] done” vs. “did”) and is therefore frequently encountered in ritual situations,
in which the poet announces a sacrificial act as just completed (like the kindling
of the fire) or a poem just composed. The verbal system also has a special cat-
egory called the injunctive, which has no formal marking for tense or mood and
therefore can be employed in a variety of functions—an ambiguity that the poets
often exploit.
Such is the structure of the language in general. We should now consider how
the poets utilized their language and what stylistic choices they made among the
many possibilities afforded them by the grammar. We should first remark that
the language we encounter in the R̥ gveda was almost surely not the standard
everyday idiom of the poets themselves. Instead, they composed in a deliberately
archaic and deliberately elevated register appropriate to the poetic tradition they
belonged to and the solemn nature and high sacred purpose of their hymns.
Such reaching for the archaic and the elevated is common across religious tradi-
tions; one need only glance at modern prayer books and liturgies, even those
supposedly updated to reflect contemporary language, to encounter the same
phenomenon. The problem with regard to the R̥ gveda is, of course, that we pos-
sess no control sample of the “standard everyday” language of the poets, though
occasional forays into a lower register as well as phonological and morphologi-
cal forms embedded in the text that show developments characteristic of later
forms of Sanskrit and Middle Indo-Aryan give us some hints of what everyday
language might have been, and the language of the only slightly later text, the
Atharvaveda, may be closer (though certainly not identical) to what the poets
spoke “at home.”
Although most of the R̥ gveda is couched in very-high-register language, the
poets sometimes, sometimes quite abruptly, slip into what appears to be a collo-
quial, even slangy, register—a switch that almost always has a dramatic purpose.
These passages are especially found in dānastutis, which are often filled with puns,
often obscene, and obscure terms, and characterized by “popular” phonological
and morphological forms. When women’s speech is represented in the text, it also
appears to belong to a lower register, and the technical terms of Vedic pastimes
like dicing and horse-racing and occupations like stockbreeding and agriculture
introduce us to lexical levels different from the high style of praise poetry and again
 Introduction 61

presumably closer to ordinary language. Not surprisingly, it is harder to determine

the meaning and reference of the words in these low-register passages than those
in the elevated discourse that generally prevails in the R̥ gveda, and many of our
translations of such passages are provisional. Insofar as possible we have tried to
signal the change in register by using more idiomatic and slangy English. (Needless
to say, our translations of these parts will soon be dated. Such is the fate of slang,
both now and in antiquity.) For further discussion of low register in the text, see
Jamison (2008a, 2009b, 2009c).
The poets’ manipulation of language was not limited to change of register. The
most significant and salient feature of the poets’ relationship to language is their
deliberate pursuit of obscurity and complexity. The strong privileging of obscurity
is found in all aspects of R̥ gvedic poetry, and we will mention only a few aspects
here. To begin with mythology, the poets seldom (we would venture to say never)
relate a myth fully and in chronological order; instead they shatter the narrative line
by ignoring chronology, omitting key plot elements, and obsessively focusing on a
few moments, alluding briefly and obliquely to the myth and deliberately failing
to identify the actors in the myth. The same is true for ritual: well-known steps in
ritual procedure (like the kindling of the fire or the mixing of soma with milk) are
shrouded in metaphor and figurative language and do not always appear in their
ritual order. The fondness for riddles and enigmas, to be discussed below, places the
puzzle-to-be-solved at the heart of the intellectual and poetic enterprise.
And so it is also with their use of language. The poets push their syntax to
the limits of intelligibility (and, at least for us, sometimes beyond) by permuta-
tions of word order, radically breaking up constituents and scattering their words
through a verse, omitting key constituents (like the direct object or even the verb),
and violating expected case frames and other grammatical conventions. They
delight in confecting variant morphological forms, and the vocabulary they use is
simply enormous relative to the size of the text: Grassmann’s lexicon of the R̥ gveda
is a very fat volume, with 1685 pages of lemmata—a giant word-hoard, full of
rarities and hapax legomena, slang and colloquialisms. Determining the mean-
ing and reference of R̥ gvedic words has always been one of the greatest challenges
in Vedic studies, going all the way back to the indigenous lexicographer Yāska in
the mid-first millennium bce. Even words that are continued into later Sanskrit
often have radically different values in the R̥ gveda from their later representatives.
(Especially cogent examples are the resonant terms dharma(n) and brahma(n),
which cannot be read with their later values.) But many words simply do not
appear in later Sanskrit or in Middle-Indo-Aryan. Much help is given by cognates
in other Indo-European languages, especially closely related Avestan, and from
Indo-Iranian and Indo-European root etymologies. But of course cognates and
root etymologies are not always a reliable guide to synchronic semantics, as mean-
ing changes over time and space, and some words are simply isolated, attested only
once (the technical term for this is hapax, or more fully hapax legomenon) or only
a few times in non-diagnostic contexts. We have often signaled such words in the
62 The Rigveda

introductions to individual hymns, although we generally give at least a provisional

translation. Again, the poets clearly revel in their lexical riches, employing what
seem like dozens of synonyms and near-synonyms for key concepts (like “shine” or
“sing”) and seeking out rarities and archaisms.
Why is obscurity so highly prized in the R̥ gvedic aesthetic? On the conceptual
level it has to do with the audience—or the most important members of the audi-
ence—as well as the target of the composition, namely the gods. The aim of the
poets is to praise the gods at the sacrifice. But it can’t be just any praise, tired repeti-
tions of already hackneyed formulae—for the gods are connoisseurs. They of course
already know their own fine qualities and formidable deeds, and they have attended
countless sacrifices where these have been recounted. In order to ensure that the
gods will come, and keep coming, to our sacrifice, not that of some competing sac-
rificer, we have to trick out the same underlying material—“you are great, and your
deeds and gifts are great”—in ever novel fashion, with imagery and turns of phrase
to embellish and transform the same underlying message. For, as prose texts not too
much later than the R̥ gveda regularly say, “the gods love the obscure” (or literally,
what is “out of sight”). Thus it is the gods’ taste for the recondite that shapes the
humans’ poetic products. The poets regularly boast that they are producing “a new
hymn”—for example, in the felicitous phrasing of III.31.19, “I make new (the hymn)
born of old” (návyaṃ kr̥ ṇomi. . . purājā́ m). This statement indirectly draws attention
to another reason for the prizing of obscurity: the poets are working in an old tradi-
tion, and each poet takes pride in making something new using inherited techniques.
It is the “born of old” feature that gives the answer to a question we have only
implicitly posed: if the poets are pushing their language to the very limits, how can
it be intelligible? It is because of the very stereotyped and limited nature of the sub-
ject matter and the shared universe of discourse among the poets, the human audi-
ence, and the putative divine audience. Everyone knows the great deeds of Indra
or the step-by-step procedure for purifying the soma juice, and they can recognize
their plain outlines under the fancy-dress embellishments and recherché conceits
in which the poet has clothed them. The audience can complete a truncated phrase
with the proper missing word because the full phrase is always in their heads. They
can unscramble the syntax and rechronologize the narrative and identify the mythic
actors indicated only by masking pronouns. A wider range of poetic subjects would
in fact have constrained the poets’ verbal exuberance, because they could not so
fully have relied on the unconsciously shared background knowledge of their hear-
ers. For us this background knowledge is hard and consciously won and will be
forever incomplete, but we must try to read the poems with an awareness of the
performance context in which they would have been received.

B. Hymn Types and Structuring Devices

There is no single, universally applicable template to which all, or even most,

R̥ gvedic hymns conform. At best we can state the uselessly vague obvious:  most
 Introduction 63

hymns praise a god or gods, often with reference to their attributes and deeds, and
explicitly or implicitly request goods and services from the divinities in return for
this praise. This exchange is often effected during the sacrifice at which we hope they
are present, either because they are part of the sacrificial paraphernalia (like Agni
and Soma) or have come as invited guests. Within these extremely loose param-
eters, the poets take many different approaches, often emphasizing one element
of the overall program while backgrounding or ignoring others. Indeed sometimes
even the “praise” portion of the overall genre “praise hymn” seems to have been
entirely elided. Because of the multifarious nature of the hymns, we have provided
an introduction to each individual hymn, describing its particular preoccupations
and structures and its particular deployment of shared tropes and themes. In what
follows here, we will first just pick out a few especially common tropes that can
dominate single hymns—this is a representative, not exhaustive selection. We will
then discuss formal devices that provide structure to whole hymns. Our underlying
assumption throughout is that hymns should be approached as hymns, not as mere
unordered collections of loosely linked verses, and that it behooves the investiga-
tor to seek structure and coherence even when the hymn seems on the surface to
lack them.
In hymns that do roughly conform to the sacrificial model just outlined, we
can single out two salient aspects, which sometimes carry all or most of an entire
hymn: the invitation and the journey. Many hymns begin with an invitation to the
god to come to our sacrifice, and some hymns make this invitation the focus of the
whole. Many other hymns focus on the god’s journey to the sacrifice (e.g., VII.24)—
the hitching up of horses and chariot, the progress from heaven through the mid-
space and across the earth, often passing over other sacrificers on the way (see, e.g.,
III.35). The epiphany of the god, his arrival at our ritual ground, can be the climax
of the hymn (for a superb example of this genre, see the Marut hymn I.88). Not
surprisingly, the invitation and the journey are often combined in a single hymn.
The progress through the sacrifice, itself a kind of journey, is also a frequent
organizational device. Sometimes this organization is quite precise, as in the
hymns that follow an ordered series of oblations, like the Praügaśastra (I.2–3), the
R̥ tugrahas (I.15, II.36–37), or, especially, the Āprī litany of the Animal Sacrifice, ten
versions of which are found scattered through the R̥ gveda. More often the ritual
progress is less formalized—for instance, the dawn sacrifice, first signaled by the
approach of the goddess Dawn, the rising of the sun, and the kindling of the ritual
fire, followed by the arrival of the gods who receive the offerings at the Morning
Pressing, especially Vāyu, Indra, and the Aśvins, and the distribution of the priestly
gifts or dakṣiṇās (see, e.g., VII.78).
A favored verbal conundrum, the riddle, not infrequently furnishes the subject
matter for whole hymns, in which each verse provides a puzzle, to which the answer
may or may not be given within the verse itself. The most famous example of this is
I.164, an All God hymn frequently referred to as “the riddle hymn,” but many other
hymns have the same basic structure—for example, “the weapon hymn” (VI.75)
64 The Rigveda

and the delightful All God hymn VIII.29. Numerology is often an important part
of these riddle hymns. Riddle hymns are generally formally structured as lists (see
below).
A number of hymns, especially, although not exclusively, late hymns, reflect on
the nature or performance of the ritual and the functions of poets and priests.
The Puruṣasūkta just mentioned anticipates the interpretations of the sacrifice in
the Brāhmaṇas, for the “man” whose parts become the world represents the sacri-
ficial offering (X.90.6) or the sacrifice itself (cf. X.90.7). A few hymns (e.g., IV.5,
VI.9) take as their subject the poet’s meditation on the craft of poetry and on his
acquisition of it, and R̥ gveda X.129 links the creativity of poets to the creation
of the world. Again anticipating comparable discussion in the Brāhmaṇas, still
other hymns enter into debates about the performance of the sacrifice. In telling the
story of Agastya, the Maruts, and Indra, for example, R̥ igveda X.165, 170, and 171
appear to defend traditions that favor offering soma to both Indra and the Maruts
at the Midday Pressing instead of to Indra alone as some traditions held. Likewise,
R̥ igveda X.109 provides justification for including the Sacrificer’s Wife in the sacri-
ficial performance and thereby defends against critics of her inclusion, such as the
poet of VIII.33. Although later employed in the ritual, such hymns apparently had
a meta-ritual function and were probably composed for an extra-ritual context.
Another famous category of R̥ gvedic hymns is the dialogue or ākhyāna type,
already mentioned above, in which two or more speakers, generally divine or
semi-divine, trade verses with each other, often in a fraught or agonistic fashion.
Generally one of the speakers is a female, and sexual tension is on display—as in
the dialogue between the legendary seer Agastya and his wife Lopāmudrā (I.179),
the twins and first humans Yama and Yamī (X.10), or the three-way discussion
among Indra, his wife Indrāṇī, and a monkey (the Vr̥ ṣākapi hymn, X.86).
Such are a few of the thematic organizing principles of R̥ gvedic hymns. At
least equally important are various formal means of organizing hymns or parts
of hymns. As one of us has treated this at some length in several publications (see
especially Jamison 2007: chap. 2; also 2004a, 2006), we will provide relatively brief
discussion here.
One of the simplest and most effective ways to impose structure is by repeti-
tion, a procedure that the poets constantly employ. It can be as straightforward as
repeating the same word (often a personal pronoun such as “you” or the name or
epithet of a god) at the beginning of every verse (e.g., forms of the 2nd sg. tvám in
II.1); Sanskrit grammar facilitates such patterned repetition because its elaborate
case system allows flexible word order, and therefore whatever their grammatical
function, key words can be positioned in initial position. The poets often introduce
complications into their repetitive schemata. Sometimes a repetitive pattern takes
awhile to become established in a hymn, with the first few verses providing several
variants that settle down into a frozen pattern somewhat later (see, e.g., I.112); con-
versely, strict repetition earlier in the hymn may loosen up in the last verses (e.g.,
II.1, in its final verses 15–16). Moreover, many cases of repetition consist not of a
 Introduction 65

single repeated word, but of several (such is actually the case with II.1, where the
pronoun tvám is followed by vocative agne), and the pattern may be established on
the basis of grammatical categories, not simply words (e.g., preverb pronoun... ).
Mention should also be made here of refrains. In some hymns every verse ends
with a repeated phrase, which is often the length of a full pāda (e.g., III.55), a hemis-
tich, or even longer (see VIII.36, in which only the first pāda of a six-pāda verse
contains new material; similarly VIII.37 and to a lesser extent VIII.35), but can be
shorter (e.g., the famous refrain of II.12 “he, o peoples, is Indra,” which occupies
the last [post-caesura] six syllables of a triṣṭubh line). Sometimes the refrain is syn-
tactically integrated into the verse in some parts of the hymn and not in others (e.g.,
I.96). Sometimes some or all subdivisions of the hymn have refrains; for example,
in the tr̥ ca hymn VIII.12 the last four syllables of the final pāda of each verse form a
refrain, syntactically integrated in the verse, and each tr̥ ca has a different refrain. (In
the translations we will generally mark refrains with a preceding dash.) A special
type of refrain is the family or clan signature: in some of the Family Books many
of the hymns (in the right meter) end with a pāda that marks the hymn as a product
of that bardic family—for instance, the Gr̥ tsamada refrain of Maṇḍala II, “May we
speak loftily at the ritual distribution, in possession of good heroes,” found at the
end of most, though not all, of the trimeter hymns of that book.
In addition to repetition of a single word or phrase in every verse, or almost
every verse, repetitions can knit one verse to another in a chain, a procedure we
might call concatenative repetition. There a word or phrase from one verse will be
repeated in the next verse; then a different word from that verse will be repeated in
the following one, and so on through the hymn (see, e.g., I.85, VI.55, X.84).
As was noted above at the mention of riddles, hymns are frequently structured
as lists, with each verse representing a separate item in the list. This structural prin-
ciple is well suited to All God hymns, many of which treat a series of gods, one
per verse (see, e.g., VI.49). Another common application of the list model is in the
recounting of a series of divine deeds. A number of Aśvin hymns have this shape,
with each verse treating a different (and often quite obscure) rescue or kindly act
for a series of named persons (e.g., the Aśvin hymns of Kakṣīvant, I.116–119).
Many list hymns are reinforced, their list shape called attention to, by syntactic
parallelism and by repetition. For example, the famous Indra hymn II.12 consists
of a series of definitional relative clauses (“[he] who. . . ”) recounting deeds and attri-
butes of Indra, with each verse ending with a main clause refrain: sá janāsa índraḥ
“he, o peoples, is Indra.” A list by itself is undramatic and has no built-in trajectory
toward climax, but the R̥ gvedic poets are adept at finding ways to inject forward
momentum into the static list pattern. For example, VIII.29, already mentioned, is
a riddle hymn in which each verse refers to a different god or gods; the list builds on
increasing numbers, from “one” (six verses) to “two” (two verses) to “some” (one
verse), by way of “three” (once explicitly, once implicitly). The rising number leads
toward climax, while the strict placement of each number in second position in the
verse provides a rigid list skeleton.
66 The Rigveda

A different kind of repetition is found in the well-known and often discussed

phenomenon of ring composition (see, e.g., Watkins 1995:  33–37 and passim,
with further literature). In ring composition the beginning and end of a poem or
of a unified section therein is marked by repeating at the end, either verbatim or,
more usually in the R̥ gveda, a variation on, material found at the beginning. In
the R̥ gveda this generally involves the reappearance of lexical items or derivational
variants of them, or phrases with partial agreement. For example, in the Agni hymn
I.59, verse 1 contains the phrase víśve amŕ̥tāḥ “all the immortals” while the last
verse (7)  describes Agni with the bahuvrīhi compound viśvákr̥ ṣṭiḥ “belonging to
all (human) communities.” The repetition of víśva- “all” provides the ring, and the
contrastive nouns it modifies (immortals, [human] communities) express Agni’s role
as mediator between gods and men. The word víśva- is not otherwise found in the
hymn, except in the derivative vaiśvānará, an epithet of Agni meaning “belonging
to all men,” a form of which occurs in every verse. The ring thus reinforces the mes-
sage implicit in the epithet. Ring composition frequently demarcates smaller seg-
ments in a larger hymn, and paying attention to these clues often allows us to make
structural sense of apparently sprawling hymns.
The term “ring composition” focuses attention on the beginning and end of
the section demarcated, but a particularly elaborate form of ring composition
instead defines and focuses on the center of the section. We call this the “ompha-
los” (navel) structure (see Jamison 2004a as well as 2007: 80–89). Here a series of
concentric outer and inner rings isolate the middle verse or verses, which contain
the mystery or the message of the hymn and are often phrased in mystical and
complex fashion. Good examples of omphalos hymns include the famous cos-
mogonic hymn X.129 (see Brereton 1999), enigmatic compositions like I.105 and
X.28, dramatic presentations like the Vasiṣṭha-Varuṇa hymn VII.86, and hymns
like VII.76 whose content is relatively conventional but which nonetheless show
formal omphalos structure. Omphalos structure is also prominent in the Old
Avestan Gāthās and has been more discussed there (though not by that term)
than the corresponding phenomenon in the R̥ gveda (see Jamison 2007: 86–89 with
references).
We might make brief mention here of initial and final verses as especially privi-
leged sites in the architecture of a hymn, though they are not “structuring devices”
per se. The first verse often poses a problem or sets the theme that the rest of the
hymn will seek to work out. Probably the most famous Indra hymn, I.32, begins
índrasya nú vīryā̀ ni prá vocam “Now I shall proclaim the heroic deeds of Indra,”
after which the poet does just that—though it must be admitted that he only relates
one of them, the Vr̥ tra battle. Final verses can be of two types, integral to the hymn
or extra-hymnic. The latter includes the clan refrains discussed above. A different
type of extra-hymnic verse is what we might call the meta-summary. The poet often
ends his hymn with an announcement of what he has just done, often expressed in
the aorist of the immediate past and breaking the mood of immersion in the divine
world. For example, the final verse of the exquisitely crafted Indra hymn I.61, which
 Introduction 67

is full of praise for Indra’s deeds and attributes, begins with the brisk hemistich,
“Thus have the Gotamas [the poet’s poetic clan] made you sacred formulations for
the ‘Fallow-bay-yoking’ (oblation),. . . o Indra.” The sacrifice is complete, their job
is done, and they point out their achievement to the god to spur his benevolence
as they return to the mundane world. Final verses that are integral to the body
of the hymn often subtly break patterns established earlier in the hymn to bring
the composition to a climax. Such for example is the final verse (10) of VIII.29,
a hymn mentioned several times above. Other examples are discussed in Jamison
(2007: 79–80 and passim).
One last technique to be mentioned, “poetic repair” (Jamison 2006), is not a
structuring device but a method for producing forward momentum. The poet sets
a problem—lexical, syntactic, or thematic—earlier in the hymn and then “repairs”
this problem later in the hymn by substituting the expected word, syntactic con-
struction, or thematic element for the problematic one. The audience is thus first
put off-balance by a disturbance in the poetry and then rebalanced when the super-
ficial solecism is fixed. An audience used to this type of repair will build expecta-
tions that propel them through the poem.

C. Imagery, Metaphors, and Similes

The R̥ gveda is rich in imagery: the procedures of sacrifice, the exploits of the gods,
the activities of men, and the elements and functioning of the cosmos are con-
stantly presented in images of something else, images based on similarity and paral-
lelism. On the formal level by far the most common way of expressing these images
is in a simile—the most common poetic device in the R̥ gveda, as it is in Classical
Sanskrit poetry. Although to Indologists the dominance of the simile may seem too
predictable to be worth noting, in fact this is one of the features of R̥ gvedic style
that looks forward to the classical era, and seems to represent something of a break
from the stylistic parameters of the poetic tradition from which R̥ gvedic practice
emerged. The Avesta contains very few similes—none in the Old Avestan Gāthās
to which the R̥ gveda is otherwise so akin—and the so-called “Homeric simile” of
ancient Greek epic, with its elaborately imagined world expressed in verb phrases,
is structurally very different from the Sanskrit simile.
Both in Vedic and in Classical Sanskrit poetry the simile is essentially nomi-
nal: that is, in a syntactic structure nominal elements are compared with each other,
while the verb is held constant. An English example would be “Indra attacked
the enemy, like a lion a sheep,” where the verb “attack” serves for both frame and
simile, while Indra=lion and enemy=sheep provide the comparisons. In a case lan-
guage like Sanskrit both “Indra” and “lion” will be coded as nominative case, and
“enemy” and “sheep” as accusative. Neither Vedic nor Classical Sanskrit regularly
has similes of the type “Indra attacked the enemy, as a lion devours a sheep,” with
two different clauses constituting the comparison and a difference in verb. The
“like” of the English example has overt expression in Sanskrit as well: by iva or ná
68 The Rigveda

in the R̥ gveda, and just iva in Classical Sanskrit. For further on the structure of the
simile and the ways that R̥ gvedic poets exploit it, see Jamison (1982).
Although the simile is ubiquitous in the R̥ gveda, it is not the only vehicle of
imagery in the text. Implicit identifications of disparate elements are another ines-
capable stylistic feature, and, as was discussed above, regularly recurring identifica-
tions (bandhus)—the fire as sun, the chariot as sacrifice, and so forth—provide the
conceptual structure of the Vedic cosmos. Thus, poetic style coincides here with
the shared notions of the world that shape “the Vedic mind.” However, in mak-
ing identifications the poets do not confine themselves to these shared and stable
associations, but often make bold and superficially puzzling equations. Decoding
the shared features that allow such equations to be made is one of the intellectual
challenges that the poets posed to their own audience and that engages us, and
often eludes us, to this day.
Although the poetic foundation on which the R̥ gveda rests was an ancient one,
the imagery of its poems comes from the immediate world of its poets. As already
remarked, the R̥ gveda is fundamentally a collection for the soma rite in which Indra
is the principal deity. Indra is a warrior, and therefore images of battle, war, con-
test, and conflict provide the background for a great many R̥ gvedic hymns. In the
IXth book, for example, the ritual process of creating soma can represent a war
campaign by King Soma. The dripping of soma as it is pressed is the beginning of
Soma’s attack or raid. The flow of soma over the woolen filter and into the soma
vessels is the destruction of Soma’s enemies. With the mixing of soma with milk,
Soma wins cows and other goods and distributes them to his subjects, who are the
sacrificers (cf. Oberlies 1999: 167–206).
Other items associated with war and periods of mobilization also loom large in
the poems, perhaps none more than the chariot. The chariot was one of the most
visible cultural symbols of the Āryas, for it was likely unique to them among the
peoples of ancient South Asia, and it was critical to their success in battle and their
mobility. In R̥ gvedic poetry the chariot becomes the hymn that travels to the gods
or the sacrifice that brings the gods (II.18.1). When they perform the ritual or com-
pose the hymns, priests become the fashioners of the chariot (I.61.4, V.2.11), and
just as a chariot brings booty from war or a winning chariot the prize of a race, the
sacrifice carries goods from the gods to humans (X.53.7). In I.129.1 Indra is asked
to lead a chariot that is both an actual, racing chariot and a metaphorical chariot,
the sacrifice. The chariot is also the vehicle of the gods, by which they come to the
sacrifice. Distinctive animals pull the chariots of different gods: the two fallow bay
horses of Indra, the dappled mares of the Maruts, the mares of the Sun, the ruddy
cattle of Dawn, the goats of Pūṣan.
More abstractly, the imagery of war is also implicit in the frequent symbols of
expansion and confinement. This imagery is most evident in the principal Indra
stories. Vr̥ tra represents what hinders and blocks, and when Indra kills Vr̥ tra, he
shows himself to be the power that can destroy any other obstacles to life and pros-
perity. The Vala cave enclosed and entrapped the cattle, and Indra must break open
 Introduction 69

Vala to free the cows, the dawns, and the light. Finding open and well-watered pas-
ture lands was essential for the Āryas, since their cattle, horses, and other livestock
depended on them. The second half of IX.113 is a poetic vision of heaven, where
there is “inexhaustible light” (7) and “youthfully exuberant waters” (8), where one
moves “following one’s desire” (9), and where there is “independence and satisfac-
tion” (10). It is the heavenly vision of a pastoral people, longing for a place of
freedom and abundance.
Pastoral imagery dominates R̥ gvedic poetry at many turns. Cows are every-
where, especially as the symbols and substance of wealth. The attention of the gods,
which will bring rewards to humans, is a cow (II.32.3), and rain is milk from heaven
(V.63.5). In II.34.8 the Maruts are compared to a cow, since they “swell” with rain,
the way cows swell with milk. In VI.45.7 Indra, as the god who inspires poetic
formulations, is a cow whose milk is the hymns. Or again, in VIII.1.10 Indra is
both the milker of the cow and the cow whose milk is “refreshment” for sacrificers.
Speech is a cow that gives the forms of speech as her milk in VIII.110.10, and in
X.64.12 the insight that the gods have given the poet should swell like a cow with
milk. VI.48.11–13 combines several images of the cow, beginning with an actual
cow, whose milk is the milk that is mixed with soma but including also the cow that
represents poetic inspiration and prosperity. In X.133.7 prosperity brought by the
gods is a cow giving her milk “in a thousand streams.” The dawns, which are linked
with the dakṣiṇā, the sacrificial reward, and with riches more generally, are cows
(IV.1.16, 52.2–4).
While the poets have particular fondness for cows, male animals too figure signif-
icantly in the hymns. Bulls and buffaloes embody strength and virility, and therefore
they represent mighty gods, potent sacrifice, and strong men. Agni is a bull with a
strong neck (V.2.12), horns (V.1.8), and a powerful bellow (X.8.1). In V.40.1–4 the
bull-like pressing stone and bull-like soma are prepared so that the bull Indra will
join together with his bulls, the Maruts. Indra is “the bull overcoming the power-
ful, the tempestuous king, smasher of Vr̥ tra, soma-drinker” (vs. 4). Parjanya, the
Thunderstorm, roars like a bull (V.83.1), has the powers of bull (vs. 2), and pours
his fertilizing streams as a “bullish stallion” (vs. 6). Indeed, anything associated with
their ideal of masculinity is likely to be bull-like for the R̥ gvedic poets.
The cultural role of horses—racehorses and warhorses—was obviously central
to the Āryas and, as a result, so was their poetic role. Agni is praised like a steed
(III.22.1), for he is a horse that brings rewards (I.27.1). When he is kindled, he is
a hungry horse, which breaks free of his enclosure (VII.3.2). Soma is a racehorse
groomed by the fingers of the priests and running over the woolen filter (IX.6.3,
5; cf. 13.6). The waters too are like racehorses that should run forever (IV.3.12).
A sacrificer harnesses himself to the sacrifice like a horse (V.46.1), and the sacrifice
is brought to success like a horse (IV.10.1). The significance of the horse is perhaps
most obvious in the Dadhikrāvan hymns, IV.38–40. Dadhikrāvan was likely the
actual horse of King Trasadasyu, but Dadhikrāvan also represents the rule of the
Pūrus, the tribe to whom Trasadasyu belonged, and the sun, which can represent
70 The Rigveda

the king. Similarly, in X.178, Tārkṣya is a protective deity of chariot drivers and
perhaps a deified racehorse himself.
Although the poets are focused on pastoral life and the herd animals they
know best, wild animals also occasionally appear in R̥ gvedic hymns (see Jamison
2008b). Birds are frequently mentioned (I.164.21; X.80.5, 123.6), especially since
the gods fly like birds (I.166.10). Unlike other birds, the falcon is not only fast but
also can be trained. Mythologically the falcon is particularly significant, since it
brought the soma from heaven to Manu (IX.48.4, IV.26–27). The wolf is a recur-
ring symbol of lurking danger (I.42.2, 105.11, 18, 120.7; II.28.10), as is the snake
(VII.38.7). The most dangerous creature of all is Vr̥ tra, who is a gigantic cobra
(I.32). More benignly, the inseparable Aśvins are compared to a pair of cakravāka
ducks (II.39.3), and in I.64 the Maruts are not only bulls and buffaloes, but they
also roar like lions and devour trees like elephants. Perhaps the broadest array of
animals appears in X.28, which is built around various animal fables (also Jamison
2008b).
Various human pursuits play significant roles in poetic imagery. The poets fre-
quently mention weaving (e.g., I.115.4), which is similar to the intricate patterns
of hymn composition and sacrifice (VI.9.2–3, X.101.2, 130). This is an inherited
Indo-European trope. Given the pastoral symbolism elsewhere, it is not surprising
that cattle-tending is a major source of poetic imagery. Indra is like a herdsman
who separates his flocks from those of others (V.31.1; cf. VI.19.3). The poet too
can be a herdsman driving his praise to Indra (VI.49.12). Or we have the reverse
image: in VII.18.10 enemies run helter-skelter like cows without a herdsman. In
addition to animal husbandry, the Āryas also raised crops, such as barley, but agri-
cultural imagery does not figure much in the R̥ gvedic poetry. Again, the reason
may be the connection of the soma rite to Indra and to the period of mobiliza-
tion (yóga). In the period of settlement (kṣéma), agriculture would have had a
larger role.

D. Riddles

Much of the R̥ gveda is enigmatic, not only because of our distance from the time
of its creation, but also because the poets meant it to be enigmatic. They valued
knowledge, especially the knowledge of the hidden connections (the bandhus dis-
cussed above) between the visible world, the divine world, and the realm of ritual.
They embedded that knowledge in hymns that were stylistically tight and elliptical,
expressively oblique, and lexically resonant. As a result, many hymns of the R̥ gveda
can appear to us as riddles. However, there are also hymns purposefully composed
to cloak their subjects or to withhold them until late in the hymn. In the later Veda
the ability to solve the riddles they pose became a formalized demonstration of the
knowledge of the priests taking part in the sacrifice. In the R̥ gveda the riddle hymns
challenged the interpretive ability of their hearers and demonstrated the cleverness
of their poets.
 Introduction 71

One kind of riddle hymn is represented by V.47. Through its first six verses, it
describes various deities but does not name them. The descriptions point toward
various gods, but never so plainly that hearers could be sure of their interpretations.
At the very end of the hymn, in its last verse, the poet twists his strategy and in the
very first pāda names the gods he addresses, gods who may or may not be among
those he described earlier. The clarity of that last verse throws into greater relief
the difficulty of the preceding verses. The poet of VIII.29 uses exactly the reverse
strategy. This is a list hymn, in which each of its verses identifies a god or gods by
attributes or behavior, but not by name. In this hymn, however, the identity of the
gods it describes is not mysterious at all—or is not until the last verse. At that point,
when the confidence of hearers to interpret the hymn would be greatest, the poet
presents a real riddle, in which the identity of the subject is not at all clear. In III.55
the riddle of the identity of the gods described in its verses points toward an even
deeper mystery. This mystery is laid out in its refrain: it is that the many gods share
a single lordship (asuratvám ékam).
Other poets treat the sacrifice as a mystery. In X.114 the poet describes the con-
stituents of the sacrifice in symbols that make the identification of those constitu-
ents difficult or impossible. It leaves the sense that the sacrifice can be penetrated
only so far, that it is beyond full understanding, except, perhaps, for poets who cre-
ate such hymns. Another characteristic of X.114 is that each of its verses contains
a number—small numbers at the beginning (2, 3, 4, 1), larger numbers toward the
end, and an ironic plural of the word for “one,” éke “such ones,” in the last verse.
Other hymns employ numerical riddles, in which a number is key to their solution,
if there is a solution. One of the most obscure hymns in the R̥ gveda is X.27. Verse
15 describes the convergence of numbered groups of heroes, none of whom are
identified. There are seven who come from the south, eight from the north, nine
from the west, and ten from the east, but we are left to puzzle about the significance
of these numbers and directions.
Perhaps the most famous riddle hymn in the R̥ gveda is I.164, a very long poem
(52 verses, one of the longest hymns in the R̥ gveda) that moves in different direc-
tions as it unfolds. The hymn makes both implicit and explicit reference to Vedic
ritual, including rites other than the soma ritual. In referring to these rites, the
hymn suggests hidden links between ritual objects and acts, realities and processes
of the natural world, and constituents and functions of the human body. These
hidden connections continue to be hidden, for while we can see the general pattern
of the hymn, the interpretation of its specific elements often remains difficult, and
many competing solutions have been offered. This is, of course, often the case with
these riddle hymns.

E. Metrics

The most conspicuous and ubiquitous formal feature of the R̥ gveda is meter. The
hymns are composed in a variety of meters, but all of the meters are syllable-counting
72 The Rigveda

and quantitative. That is, they consist of lines containing a fixed number of syllables,
arranged in patterns of “heavy” and “light” syllables (the “quantity” referred to by
the term “quantitative”). Heavy syllables contain a long vowel (a class that includes
e and o) or diphthong or a short vowel followed by two consonants (which need not
belong to the same word) and are symbolized in Western analysis by a macron (¯).
Light syllables contain a short vowel followed by, at most, a single consonant and
are symbolized by a breve (˘). (Aspirated consonants [those written with two roman
letters, the second of which is h, e.g., th] count as single consonants.) The final syl-
lable of a line is metrically indifferent and symbolized by x. The distinction between
heavy and light metrical syllables simply formally enshrines patterns inherent in
the language itself, where various linguistic processes are sensitive to distinction
in syllable weight and the difference between short and long vowels is lexically and
morphologically crucial.
This type of metrical structure was inherited from the Indo-European poetic
tradition, most clearly evident in ancient Greek meter, especially the Aeolic meters
utilized by Sappho and Alcaeus, which are also syllable-counting and quantitative;
these Aeolic meters have long been considered cognate to Vedic meter. The meter
of the Old Avestan Gāthās is also closely akin: though the quantitative aspect has
been lost, Gāthic lines have a fixed number of syllables.
R̥ gvedic meter is also identical in its structural principles to most of the meters
encountered in Epic and Classical Sanskrit (with the exception of the ārya and
related types). The major difference between Vedic meter and Classical Sanskrit
meter has to do with the regulation of quantities. In the earlier parts of the line
R̥ gvedic meter has relatively unfixed quantities; it is only toward the end of a line
(the cadence) that the quantity of each syllable is fixed (especially in trimeter meter,
on which see below). By contrast, most Classical Sanskrit literary meters regulate
the quantity of each syllable in the line; in other words, the relative flexibility of
R̥ gvedic meter has become frozen. The exception is the eight-syllable epic anuṣṭubh,
or so-called “śloka” meter—the overwhelmingly predominant meter in the two
great epics and the workhorse meter of non-literary Classical Sanskrit verse texts
such as Manu’s lawcode—whose quantities are precisely fixed only in the second
half of each line.
R̥ gvedic meters are generally divided into “dimeter” and “trimeter” types. The
former consists of eight-syllable lines, which can be conveniently considered to con-
sist of two equal segments, with the second half tending more toward fixed quanti-
ties, generally in an iambic pattern. However, it is important to keep in mind that
there is no fixed caesura (word-break) in dimeter meter, and even the four cadential
syllables are not rigidly fixed in quantity. Trimeter meter is more complex. It gener-
ally consists of lines of either eleven or twelve syllables (triṣṭubh or jagatī respec-
tively), characterized not only by a fixed cadential sequence of four (triṣṭubh) or
five (jagatī) syllables but also by a strong caesura after the first four or five syllables
in the line. The caesura and the cadence thus effectively divide the line into three
parts—the opening (the first 4–5 syllables before the caesura), the “break” (the 2–3
syllables following the caesura), and the cadence.
 Introduction 73

Examples of the three most common types of lines, dimeter, trimeter (triṣṭubh), and
trimeter (jagatī) follow, with heavy and light syllable scansion given below the text. We
have provided two examples each for the trimeter lines, one with four-syllable opening
and three-syllable break, one with five-syllable opening and two-syllable break.

Dimeter

eight-syllable dimeter
I.1.1a   agním īḻe puróhitam
¯   ˘  ¯  ¯        ˘  ¯   ˘   x
Note the iambic rhythm of the last 4 syllables.

Trimeter

Trimeter cadences:

triṣtubh:    ¯ ˘ ¯ x
jagatī:     ¯ ˘ ¯ ˘ x

Note that twelve-syllable jagatī can be seen as a one-syllable extension of triṣṭubh,

with an extra light syllable inserted right before the end, resulting in a five-syllable
cadence. The structure of the rest of the line is the same.

eleven-syllable trimeter: triṣṭubh—opening of four

I.32.1a   índrasya nú vīríyāṇi prá vocam
¯ ¯   ˘ ˘ | ¯ ˘   ¯    ¯  ˘   ¯    x
The obligatory caesura (word break) after the opening is marked by a slash; the
break is three syllables (¯ ˘ ¯) and is not followed by a caesura, while the cadence is
the proper triṣṭubh one (¯ ˘ ¯ x).

eleven-syllable trimeter: triṣṭubh—opening of five

I.32.1b   yā́ ni cakā́ ra prathamā́ ni vajrī́
¯   ˘ ˘ ¯ ¯ | ˘ ˘ ¯  ˘  ¯  x
Note that the rhythms of the openings of the two triṣṭubh lines are quite distinct.
Here the break is two syllables (˘ ˘), and the cadence the same as the previous line.

twelve-syllable trimeter: jagatī—opening of four

II.1.16b   ágne rātím upasr̥ jánti sūráyaḥ
¯ ¯ ¯ ˘ | ˘ ˘  ˘  ¯   ˘  ¯  ˘  x
The opening is different from either of the triṣṭubh openings above; the break is
three shorts (˘ ˘ ˘), unlike the short/long/short (¯ ˘ ¯) three-syllable break of the
quoted triṣṭubh line. The cadence is the standard jagatī five-syllable one (¯ ˘ ¯ ˘ x).

twelve-syllable trimeter: jagatī—opening of five

II.1.16c   asmā́ ñ ca tā́ ṃś ca prá hí néṣi vásya ā́
¯  ¯ ˘ ¯ ¯ | ˘ ˘ ¯ ˘  ¯  ˘  x
74 The Rigveda

Yet another opening pattern. The two-syllable (˘ ˘) break is identical to that of the
triṣṭubh with opening of five, as shorter breaks have fewer possible patterns. The
cadence is the standard jagatī cadence. Note that though there happens to be a word
break between the break and the cadence in this line, it is not an obligatory one, and in
fact none of the other three trimeter lines quoted shows a word break at this position.
The lines just described are called pādas, a term we will employ throughout
this work. A  R̥ gvedic verse ( r̥ c) consists of a group of pādas, generally three or
four, though meters with fewer than three pādas or more than four are also found.
Lower-case roman letters are used to refer to the pādas in a verse (so, e.g., “5b” will
refer to the second pāda in the fifth verse of a hymn). Again, we will regularly use
this shorthand in identifying lines. There is another significant division within the
verse: the hemistich or half-verse. In four-pāda verses this consists of two two-pāda
units; in three-pāda verse, the first two pādas are considered the hemistich, with a
single pāda following.
The most common dimeter meter is gāyatrī, a collection of three eight-syllable
pādas. Approximately one quarter of the R̥ gveda is composed in this meter. The
other common dimeter meter is anuṣṭubh, which contains four such eight-syllable
lines and is the ancestor of the Epic and Classical śloka mentioned above. The
most common trimeter meter, in fact the most common meter in the R̥ gveda, is
triṣṭubh, which accounts for approximately 40% of the text. It consists of four
eleven-syllable pādas, while jagatī likewise consists of four pādas, though of twelve
syllables each. Jagatī is the third most common meter, after triṣṭubh and gāyatrī.
Besides these mono-type meters, some hymns are composed in what are referred to
in general as “mixed-lyric meters,” whose verses consist of combinations of eight-
and twelve-syllable pādas. There are a number of different combinations, each with
a different name. (For details, see especially the table of meters in van Nooten and
Holland 1994, referred to below.)
The details just given are not relevant only to professional metricians, for the
rhetoric and semantic structure of the R̥ gveda are strongly driven and shaped by
meter. Syntactic constituents often occupy single pādas, for example, and metrical
boundaries (the beginning and end of the line, as well as the position immediately
after the caesura) are favored sites for positioning emphatic elements. The hemistich
is a particularly salient unit, dividing the verse into syntactic and semantic halves.
In fact, we have discovered that it is almost always possible, and generally desirable,
to render the hemistich division in English—that is, to translate the first half and
the second half of the verse as separate units. This is reflected in the physical layout
of our translation, with the second hemistich starting a new line. It is remarkable
how faithful it is possible to be to the Sanskrit hemistichs without significantly com-
promising the English. (The hemistich division is less important and more often
syntactically breached in gāyatrī, since the division results in uneven parts:  two
pādas followed by one, but even in gāyatrī the third pāda is often independent of
the first two.)
 Introduction 75

The verse is the most significant unit in a R̥ gvedic hymn—hence the name
R̥ g-veda, or Veda of verses (r̥ c). It is almost always a self-contained syntactic con-
struction, and even when that construction is not entirely independent syntactically
(e.g., when it is a relative clause, dependent on a main clause in a verse following or
preceding), it will be internally unified. There are almost no examples of syntactic
enjambment between verses. This focus on the internal unity and syntactic indepen-
dence of the verse is continued in later Sanskrit poetry, where it reaches its defining
limit in so-called muktaka verses or single-verse poems.
Nonetheless, some hymns are structured into larger groupings of two to three
verses, which are sometimes referred to as “strophes.” The tr̥ ca or “triplet” is the
most common such grouping, consisting of three verses, generally in gāyatrī,
though other meters are also found. Sometimes the tr̥ ca unit is strongly defined by
shared lexicon or a shared refrain, or by parallel syntactic structures, or by a com-
mon theme; other tr̥ cas have only the faintest signs of unity in rhetoric or content.
Quite long hymns can be built from these three-verse units, which are especially
common in Maṇḍala VIII and, as Oldenberg argued (1888: 119–40), were the spe-
cial province of the Udgātar (singer) priest. Many of these strophes were borrowed
into the Sāmaveda, whose principal priest in classical śrauta ritual was the Udgātar
and one of whose major textual sources is R̥ igveda VIII. The other major strophic
type is the pragātha, consisting of two verses in two different types of “mixed lyric”
meters. The usual combination is br̥ hatī (8 8, 12 8)  and satobr̥ hatī (12 8, 12 8).
Again, pragāthas are especially common in the VIIIth Maṇḍala and frequently
taken over into the Sāmaveda. Thus the standard types of multi-verse groupings
tended to provide the lyric or sung portion of the ritual, as against the recited por-
tions associated with the Hotar priest.
A brief and clear account of R̥ gvedic meter is given in Macdonell (1916: appen-
dix II). For full details see Arnold (1905). Oldenberg’s Prolegomena (1888) is invalu-
able both for his detailed treatment of the meters of particular hymns and sets
of hymns and for his discussion of the historical implications of metrical details.
Van Nooten and Holland (1994) provide a hymn-by-hymn metrical commentary
in somewhat lapidary style, as well as a table of R̥ gveda meters as identified in the
Anukramaṇī (pp. xiv–xvi), which should be consulted for the names and character-
istics of meters not explicitly discussed here.

VIII.  Translation Principles

A. Sanskrit into English: Problems and Solutions

As has been repeatedly emphasized above, the R̥ gveda is a poetic text, structured
by intricate meters, driven by rhetorical principles based on this metrical struc-
ture, and crafted by skillful poets for a poetically aware audience. Nonetheless, we
have chosen to translate the text into prose, not verse—for several reasons. First
76 The Rigveda

and perhaps most important, we are not poets, and we would dishonor the highly
trained and highly inventive poets of the R̥ gveda by translating their artful cre-
ations into bad English verse. Moreover, the structures of the English language and
of English verse are entirely different from those of Vedic Sanskrit. Since English
lacks the elaborate morphology of Sanskrit, it is not as possible in English, with-
out awkwardness or, indeed, loss of sense, to use word order for rhetorical rather
than syntactic purposes. The stress-counting principle that regulates English blank
verse and the end rhyme characteristic of much English poetry are alien to R̥ gvedic
poetry. Thus the English poetry that resulted from a verse translation would not
replicate in any of its most salient features the structures of R̥ gvedic poetry.
We have, however, tried to retain the verse structure as much as possible. All the
translations reproduce the verse divisions found in the hymn, and within verses the
hemistich boundary is also always marked, with the second hemistich beginning
a new line. In fact, as noted above, it has almost always been possible to translate
hemistichs as units without breaching the boundary—an indication of how strong
a compositional element the hemistich was for the poets.
Translators who are also poets might succeed in rendering the R̥ gveda (or parts
thereof) into poetry that captures the spirit and flair of the original, but such ren-
dering would perforce (or so we think) distort or reinvent the literal meaning of the
text. This is of course a perfectly acceptable translational strategy—some think it
is the only acceptable strategy: that a literal translation is a fundamental betrayal
of the original. But we have chosen to hew as close to what we consider the literal
meaning of the text and its constituent words as we can. The text is multivalent, and
over the millennia it has received multiple, often incompatible, interpretations. By
translating the text literally, we hope to leave the interpretive opportunities open for
the readers, inviting them to participate in the act of interpretation though provid-
ing as much guidance as we can.
As was also noted above, the everyday language of the R̥ gvedic poets was almost
surely not identical to the language they used in their hymn compositions. They
may well have spoken a form of early Middle Indo-Aryan—judging from some
Middle Indic phonological features found in the hymns—or at the very least a more
stripped-down form of Sanskrit, with the limitations on morphological categories
and variant forms found in middle Vedic prose and in Epic and Classical Sanskrit.
Since they were therefore composing in a deliberately archaic style, we have aimed
for a fairly formal and old-fashioned English style on both the lexical and the syn-
tactic levels, with occasional whiffs of the archaic. (For example, one of us often
translates the morphologically opaque archaic frozen form śám with the equally
opaque English “weal,” to capture its linguistic isolation; the other of us prefers
“luck,” which does have the advantage of conveying more sense to the modern
English reader.)
The poets were, however, quite aware of linguistic registers, and are capable of
making wrenching shifts from, for example, the high-register solemnity of most
of their production to slangy and sometimes obscene humorous banter. (See, for
 Introduction 77

example, Indrāṇī’s discourse in X.86, which lurches from shockingly explicit vul-
garity [especially from a goddess] to the highest of high styles.) Such passages are
especially common in dānastutis. When we have been able to identify low-register
vernacular passages, we have attempted to render them into a similar English
register.
One feature where we have deliberately avoided English archaism is in the trans-
lation of the 2nd-person pronouns, singular versus plural (and dual)—rendering all
of them with “you,” although it would occasionally have been useful to disambigu-
ate between singular and plural. In part this is an English stylistic choice: a consis-
tent distinction between “thou/thee” and “ye/you” quickly becomes wearisome for
the reader. But it would also produce a misleadingly archaic impression; the distinc-
tion between the singular and plural [/dual] 2nd-person pronouns remains a feature
in Sanskrit throughout its history, and, though the dual was lost, the singular-plural
distinction was also a feature of Middle Indo-Aryan. Therefore this same distinc-
tion in Vedic Sanskrit carried no stylistic weight at all; it’s simply an unremarkable
feature of the grammar, and the obsolete English “thou” would distort the effect
of the text.
An area where the balancing act between R̥ gvedic style and English intelligibil-
ity most tests the translator is in the placement of words. The R̥ gvedic poet may set
a word at the beginning of a verse or hemistich to draw special attention to it, or
hold back words until the end of the verse to surprise his audience or cause them
to change their expectations about the meaning of the line. He may position identi-
cal or similar elements in the same place in a sequence of verses in order to build
larger rhetorical units or to emphasize the similarity of ideas in the verses, or may
scramble elements that underlyingly belong together in order to produce a mosaic
effect, whose contours only become clear over time. In short, the poet can use the
placement of words in order to unfold his ideas in a particular and deliberate way.
The freer word order that an inflected language and a shared poetic diction allow
gave the poets flexibility in arranging their verse. Our translation does not aspire to
be poetic, and our medium is the highly uninflected English language. Nonetheless,
we have been reluctant to surrender all attempts to follow the poets’ positioning of
elements, and have tried to follow the order of words and ideas as well as we can,
especially in cases where the verbal positioning seems to make a particular rhetori-
cal or conceptual point. The result is an English that is not quite natural. Apart
from preserving the structure of the verse and the unfolding of ideas, there are
other advantages we see in adopting this style. It reminds readers that the R̥ gveda
is elevated and solemn poetry, and it signals that even for the composers and their
audience the language of the R̥ gveda was not everyday speech and the expression
of ideas in the text was not designed to be direct and pellucid. A translation into
a too fluent and colloquial English would produce a very misleading impression
of the original. We hope that this style of translation will slow readers enough to
allow them to sense the style of the original, while not rendering the English too
impenetrable.
78 The Rigveda

One of the features of almost all previous translations of the text, as well as the
scholarly paraphernalia, like dictionaries, that have supported these translations, is
that the lexical range in the translation language is very narrow and tends toward
the abstract, the general, and the overlapping. It’s a standing joke among Vedicists
that all verbs mean “shine,” “sing,” “speed,” or “give”; all adjectives, “bright” or
“swift”; all nouns, “praise.” This monotony contrasts sharply with the variety and
color of the R̥ gveda’s Sanskrit. As we have already noted, Grassmann’s lexicon of
the R̥ gveda is enormous compared to the size of the text (there must be somewhere
in the digital world a metric for evaluating lexical diversity per text size). The “sing/
speed/shine” translational style severely flattens this lexical richness. It is as if all the
verbal exuberance of Shakespeare were squeezed into the neo-classical spareness of
a French dramatist like Racine.
This habit not only robs the translation of its verbal diversity but, perhaps more
important, of its metaphorical content. Translators are usually not semantically
wrong when they translate a verb as, say, “shine,” but they are often translating the
underlying sense of a metaphor rather than the more vivid, and often more jarring,
image evoked by a more precise word. To give a single example, there is a verbal
root √ribh that seems literally to mean “croak,” “creak,” “rasp,” “squawk”—used,
for example, of the sound of a creaky wagon, or of a crackling fire. It is also fre-
quently used of poets and praise-singers, and a nominal derivative of it, rebhá, is
regularly applied to poets. But—no surprise—the standard rendering in current
translations and dictionaries is “singer”—German “Sänger”—and this certainly
correctly identifies the referent of the word. But the particular voice quality, per-
haps “hoarse” or “husky,” is sacrificed to making a fluent and easily interpreted
translation—not scaring the horses, as it were, by suggesting that the singers were
anything but mellifluous.
The twentieth century saw much progress in limiting and pinpointing the mean-
ing of many R̥ gvedic words, and in appreciating their position in linguistic register
(e.g., solemn and hieratic vs. colloquial vs. technical vs. unmarked and general).
Incorporating these results makes for a much more vivid text, but also a “spik-
ier” one, with unsettling or discordant images or juxapositions of concepts. To
return to the √ribh example just given, a translation like “hear the poet as he rasps”
(VIII.37.7) is more likely to take the reader aback than “hear the praise-singing
poet” (Geldner’s “den lobsingenden. . . ”), but a simile in a passage also containing
this verb, IX.97.57  “the poets squawk like birds of prey,” supports the less easy
interpretation. In that passage Geldner’s attempt to rescue it so radically attenuates
the phrase that it becomes close to incomprehensible:  “the seers become audible
like vultures” (“werden die Seher lautbar wie die Geier”).
In our translation we have aimed, again, for the literal, and sometimes discor-
dant, sense of the word—even if the literal sense was a dead metaphor for contem-
porary speakers. We have, in fact, no way to determine whether a metaphor was
dead or alive at the time of composition, and in any case, even if at the time of com-
position it was dead, it had been previously a metaphor with real poetic heft. On the
 Introduction 79

other hand, we have not been able entirely to follow the austere and chaste principle
of Bergaigne and Thieme—to translate a particular word always in the same way,
no matter what the context. Attempts to do so often produce unintelligible English,
whereas some adjustment to context aids the reader to grasp the sense. Thus our
translation has been a balancing act between the flattening effect of rendering the
underlying sense as opposed to the literal surface and a too principled adherence to
a single literal meaning for every word regardless of context.
In representing the spikier and more difficult style that we see in the R̥ gveda,
in contrast to the translation tradition that has attempted to smooth out the dif-
ficulties, fill in the gaps, and aim for clarity above all, at the expense of the literal
meaning and obscure style of the original, we have been aided by the sheer chance
of the age we live in. As the discussion above has repeatedly suggested, newer ways
of looking at the text have the unintended result of making the text “harder.” The
images are more striking but also more obscure; the lexicon is more specific but the
combinations of words therefore less harmonious; the ritual underpinnings cannot
be glossed over, however bizarre and “unpoetic” the actions depicted may seem. We
think the result is truer to the spirit of the text than the easy flow of older transla-
tions and also truer to our age. It must have been difficult to appreciate the polished
obscurities of R̥ gvedic verse before the twentieth century enshrined deliberate dif-
ficulty in poetry as a preeminent sign of serious verbal art. So Ezra Pound and
T. S. Eliot begat our poets Vasiṣṭha and Viśvāmitra, in a paradoxical reversal of the
paternal relationship that a R̥ gvedic bard would relish.
Stylistically the R̥ gveda is a dense and compact text. This results from at least
two factors. On the one hand, the elaborate morphology of Sanskrit allows a
parsimony of words, as opposed to the diffuse nature of a language like English.
To choose one example, this from the Atharvaveda (III.20.8):  the two-word
clause áditsantaṃ dāpayatu necessarily translates into at least eleven English
words:  “Let him cause the man not wishing to give to give.” There is no way
to reproduce the verbal spareness of Sanskrit in a language like English, which
needs a plethora of function words to render the less assertive moving parts of
Sanskrit morphology.
But in addition to the purely linguistic aspect, we must reckon with the narrow
shared universe of discourse in the text. We have previously discussed the formu-
laic nature of R̥ gvedic rhetoric. The shared deep-structure verbal formulae and the
shared knowledge of the nature of R̥ gvedic gods, rituals, and religious principles
and beliefs allow the poets to refer to such knowledge with a kind of verbal short-
hand—truncated or twisted formulae, brief and recherché allusions, and so forth.
Deliberate ellipsis is an important stylistic feature of the text, with the audience
invited and expected to fill the ellipsis.
These issues present the translators with at least two problems (in addition to
figuring out what is actually going on). First, to try to reflect the density of the
text in a language that is not well adapted to this style—a balancing act through-
out the translation. Second, to determine when to supply material not found in the
80 The Rigveda

particular context, and what to supply. Although we began the translation with
the determination to supply nothing not found in the passage itself, this principle
came to seem not only unworkable but also contrary to the practice of the poets,
since they often rely on shared knowledge to allow their contemporary audience
to “fill in the blanks” of allusive expressions. But we have tried to avoid the regret-
table tendency of some translators to supply material without any methodologi-
cal controls. Instead, supplied material in our translation is either generated from
material recoverable from the passage itself (e.g., verbs or nouns in the same or
adjacent verses) or from parallel phrases elsewhere in the text. Here the study of the
parallel passages adduced by Geldner in his notes has been absolutely crucial for
our interpretation, as well as other types of parallel phraseology, the discovery of
which has been much aided by Lubotsky’s R̥ gvedic Word Concordance (1997). We
have thus attempted to allow the poets and the verbal connections they have forged
to determine what a passage needs for completion, rather than simply to follow our
own whims and a common-sense notion of what has been omitted, since “common
sense” is a culture-specific quality.
Most of the material we have supplied has been placed in parentheses, in order
to indicate that it is the result of our extrapolation and is not found explicitly in the
text itself. Although parenthetical additions can sometimes be distracting, the pos-
sibility of misleading readers about what is actually there and what is not, in a text
where almost nothing can be taken for granted, seemed to us a sufficient danger to
require major additions to be overtly signaled. However, some supplied material
has been silently slipped into the text proper, especially necessary English function
words that lack obligatory Sanskrit counterparts. For example, Sanskrit does not
obligatorily supply definite or indefinite articles (“the” vs. “a[n]‌”) and in fact does
not have either category as a separate grammatical marker. But English requires
them and we have supplied them; this often forces a choice on the translators as
to whether a noun is indefinite (“a horse”) or definite (“the horse”), but in most
such cases of ambiguity we have considered it a kindness to the reader not to indi-
cate the uncertainty with parentheses. Likewise for possessive pronouns: Sanskrit
regularly does not overtly mark pronominal possession (“his horse”) where English
does, and in cases where the possessor seems fairly clear we have often supplied the
possessive without parenthetical marking. We also regularly supply pronouns for
repeated objects and the like, of the type “praise you and beseech you” rather than
“praise you and beseech (you),” unless the construction and meaning are unclear
or disputed. And, since Sanskrit does not have mandatory surface subjects because
the verb encodes the person and number of its subject, while English requires
subjects to be expressed, we have had to supply pronominal subjects, which has
required us on occasion to choose the gender of the subject (“he” vs. “she” vs. “it”).
Occasionally too, while the elliptical style of the R̥ gveda might not repeat a verb or
other crucial word in proximate clauses, good English often must do so. In these
cases we have repeated the crucial word without indicating the absence of this rep-
etition in the Sanskrit text.
 Introduction 81

A rarely used sign in our translations is the asterisk (*), which indicates that
the translation rests on an emendation. Since we seldom resort to emendation, the
asterisk will be encountered only occasionally.

B. Format of the Translations

Perhaps the most unusual feature of our translation is our decision to avoid the use
of footnotes entirely. We did not begin with that idea, but several considerations
led us to it. First, the notes in Geldner’s translation, especially the parallels he cites,
are full and informative, and interested readers should consult them: knowledge of
German is not required for tracking down the parallels. Further, this translation
will appear approximately coincident with the new Witzel-Gotō translation into
German, which contains annotations to verses that build on Geldner’s work. Yet
another set of similar annotations seemed to us unnecessary and redundant.
Second and more important, however, this decision reflects our view of the text.
One of the aims of this translation is to demonstrate that the hymns are unified,
structured compositions and not merely collections of single verses randomly col-
lected into hymns. Our emphasis on the integrity and structure of R̥ gvedic hymns,
on the idea that the sūkta is the essential compositional unit, is to us the most per-
suasive reason to avoid footnotes, endnotes, or running commentary. If we were to
invite readers to turn away from the progress of the hymn toward notes of one kind
or another, we would risk causing them to lose precisely the thing we most want our
readers to have, the sense of the hymn as a compositional whole.
The starkness of our presentation is softened by several different strategies,
corresponding to different types of problems the text presents. First and most
important, we provide introductions to each hymn. These introductions treat the
overall structure of the hymn, its theme or themes, the development of the hymn,
and any special features it may have—special features being defined as whatever
strikes the translators as particularly noteworthy. These include intertextual con-
nections with other R̥ gvedic hymns, especially in the oeuvre of the same poet,
spectacular rhetorical flights or poetic devices, marked metrical features, out-of-
the-ordinary ritual procedures or models of rituals, little-known myths, interesting
twists on well-known myths, and so forth. We also discuss the conformity or non-
conformity of the hymn to the expected patterns of organization in the maṇḍala
or hymn collection when relevant—where verses or even whole hymns have been
added and hymns have been combined, divided, or relocated. Such changes reflect
the compositional history of the text and may be important for the interpretation
of the hymn.
On the other hand, the discussions in the introductions generally do not treat
technical linguistic details, except insofar as they impact the reader’s experience of
the hymn. However, we do not apologize for regularly mentioning grammatical fea-
tures of the hymn in such cases: the poets were, among other things, expert gram-
marians and often deployed particular linguistic forms to subtle semantic ends. To
82 The Rigveda

remain in ignorance of this linguistic dimension of the hymns is to miss many of

their most striking effects.
In the introductions we usually do not engage with the considerable secondary
literature on the hymn or the variant interpretations of other translators and com-
mentators, including Sāyaṇa. Although study of this material has deeply affected
our reading of the text, a detailed discussion of how and why we differ from other
interpretations would swell our already long work to unmanageable proportions.
We know of no other R̥ gveda translation that systematically provides such intro-
ductions for every hymn, rather than just for the famous or the famously diffi-
cult hymns. We hope that these introductions will give readers sufficient guidance
to proceed through the hymn without getting unduly mired in the difficult parts.
We also hope that the introductions will make clear how very different R̥ gvedic
hymns are from one another, even when they are “about” the same subject, and will
counter the general assumption that the liturgical material in the R̥ gveda, which of
course makes up the greater part of the text, is monotonously the same throughout.
Quite the contrary: the poets obviously sought novelty not only in individual details
of imagery and wording but also in their overall approach to well-worn topics,
and they thought out innovative ways to frame such topics. Even the sixty-seven
gāyatrī hymns to Soma Pavamāna that occupy the first half of the IXth Maṇḍala
and are extremely narrowly focused on a very few steps of ritual procedure manage
an astounding amount of variety within these metrical and topical limits. In fact,
they might serve as a textbook of how a R̥ gvedic poet deploys his skills to produce
multiple, very different versions of a simple and single message.
A second type of aid we provide to readers relates to a major stumbling block
they will encounter in approaching the translation directly: while the original audi-
ence of the R̥ gveda would have been able to identify the referent in a particular
verse on the basis of characteristic epithets, defining actions, or habitual associa-
tions, our readers will not have the cultural knowledge that the poet and his audi-
ence shared. Therefore, in many cases in which such references may not be clear to
our readers, we have inserted brief bracketed identifications to situate the reader
with as little fuss as possible. We have kept such bracketed additions to a minimum,
since we assume that the general introduction will familiarize our readers with com-
mon epithets, characteristics, and tropes, and therefore identifications of the type
“Vr̥ tra-smasher [=Indra]” will be unnecessary. These bracketed additions are to
be differentiated from material in parentheses, supplied from context and neces-
sary to complete the syntactic construction of the verse, as was discussed imme-
diately above. We should note that identifying every stray referent and decoding
every cryptic image, either in the introduction or the bracketed insertions, would
prove to be unduly distracting—not to mention sometimes impossible—and read-
ers are encouraged to let the momentum of the poetry carry them across these little
perplexities. Some bracketed additions signal possible double meanings (puns or
double references) of single words, expressions, or constructions in the text; the
alternatives will be separated by a slash.
 Introduction 83

Occasionally, for hymns in which density of reference or deliberate obscurity

makes each verse a challenge to decode and appreciation of the hymn as a whole
depends on following the unfolding of its cryptic train of thought, we will provide
a verse-by-verse exegesis in the introduction to the hymn. Again, we have kept such
discursive treatment to a minimum—not only for considerations of space, but also
on principled grounds: the R̥ gveda is, first and foremost, poetry, and detailed para-
phrase is a terrible crime to commit against poetry. It is especially a crime against
this poetry, where verbal execution, rhetorical ingenuity, and deliberate complexity
are central to its poetic style.
The heading of each hymn will provide the maṇḍala and hymn number, fol-
lowed, in parentheses, by its “Grassmann number.” Grassmann’s R̥ gvedic lexi-
con did not make use of the usual numbering system of maṇḍala plus hymn, but
through-numbered the entire text without regard to maṇḍala (or aṣṭaka) divisions.
Although this number will be irrelevant for most readers of the translation, for
those who use the translation in conjunction with the Sanskrit text and the various
relevant reference works, including Grassmann’s lexicon, the Grassmann number
will facilitate cross-reference. (The Grassmann numbers for the first maṇḍala are
not supplied, as they are identical with the hymn number in that maṇḍala.) The
deity (or other element) who is the dedicand of the hymn follows the Grassmann
number in the first line, almost always according to the Anukramaṇī, except when
the Anukramaṇī’s identification is plainly fanciful or wrong. On the next line is the
poet, again as given by the Anukramaṇī; the third line gives the number of verses in
the hymn, the meter(s), and strophic arrangement, if there is one.

C. Hymn Distribution by Translator

At the beginning of this project, we two translators each took primary responsibil-
ity for approximately half the hymns of the R̥ gveda and produced first and usually
second drafts of them. As the project continued, we consulted with each other at
every turn, passed hymns back and forth, and read each other’s translations criti-
cally. In the end, however, the final version of each hymn became the responsibility
of a single translator. The division of those final versions is as follows:

Maṇḍala I
JPB: 1, 24–25, 32, 46–47, 116–120, 157–158, 161, 164–165, 170–171
SWJ:  2–23, 26–31, 33–45, 48–115, 121–156, 159–160, 162–163, 166–169,
172–191
Maṇḍala II
JPB: 11–24, 27–28, 39
SWJ: 1–10, 25–26, 29–38, 40–43
Maṇḍala III
JPB: 1–7, 9–29, 58–60
SWJ: 8, 30–57, 61–62
84 The Rigveda

Maṇḍala IV
JPB: 33–37, 42
SWJ: 1–32, 38–41, 43–58
Maṇḍala V
JPB: 1–28, 62–78, 85
SWJ: 29–61, 79–84, 86–87
Maṇḍala VI: SWJ
Maṇḍala VII
JPB: 1–17, 48, 50–52, 60–74, 82–89
SWJ: 18–47, 49, 53–59, 75–81, 90–104
Maṇḍala VIII: SWJ
Maṇḍala IX: SWJ
Maṇḍala X
JPB: 71–72, 81–82, 90, 108, 121, 129–130
SWJ: 1–70, 73–80, 83–89, 91–107, 109–120, 122–128, 131–191