The Smithsonian Institution

Regents of the University of Michigan

Early Paintings of the Goddess in Nepal

Author(s): Pratapaditya Pal
Source: Ars Orientalis, Vol. 12 (1981), pp. 41-48
Published by: Freer Gallery of Art, The Smithsonian Institution and Department of the
History of Art, University of Michigan
Stable URL: http://www.jstor.org/stable/4434248
Accessed: 19/11/2009 00:12

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=si.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

The Smithsonian Institution and Regents of the University of Michigan are collaborating with JSTOR to
digitize, preserve and extend access to Ars Orientalis.

http://www.jstor.org
EARLY PAINTINGS OF THE GODDESS IN NEPAL
BY PRATAPADITYA PAL
THE MOTHER GODDESS, IN HER MYRIAD MANIFESTA-
tions, continues to enjoy great popularity in Nepal.
There is reason to believe that her cult is among the
most ancient in that country. Her most common
appellation is "Bhagavatli," a term almost uni-
versally applied to any image of a goddess, but in
particular to that aspect known as Mahisasura-
mardini, or "destroyer of the buffalo-demon."
Curiously, representations of Mahis-suramardinT
cannot be dated much earlier than the tenth century,
which is also the date of the earliest surviving
manuscript of the DerCimihitmya, a text glorifying
Durga as the destroyer of Mahisasura and other
asuras.'
For a Sakta, a follower of Devi who is also known
as Sakti (meaning "power" or "energy"), the
Devimiihitmya has the same significance as the
Bible has for a Christian. It is customary for a devout
Sakta to read the text himself or to have it recited by a
brahmin. Often, a devotee would also commission a
manuscript for a special occasion and donate it to a
brahmin. The text is recited every day for ten days
during the autumn festival of the Goddess. The
Newark Museum possesses a fine illuminated copy
of a DevTm?ihiitmyamanuscript that was commis-
sioned on one of these holy days in 1477 by Prince
Rdyamalla, very likely to ensure his success in
battle.2 In the Devimhiihtmya, the Goddess explicitly
says that the person who reads this text will not
experience fear from enemies, robbers, or kings
(satruto na bhayam tasya dasyuto va na rdjatah//
chap. 12, v. 6a). Elsewhere we are told that the
Goddess protects her devotee from fire, on the
battlefield, and from great difficulties and disasters.
A recitation of the text also protects the devotee from
all sorts of psychological problems, from evil
influences of the stars, and from bad dreams.
Children are protected from evil eyes and from evil
spirits.3 But perhaps the most important reason for
commissioning, reading, or reciting the text was to
express piety, thereby securing a place in heaven.
It is also interesting to note that the manuscript of
the Devizm7ihaitmya(or CanEyf,as it is also called) is
often worshipped directly in place of an image of the
Goddess. In certain parts of the country, such as
Gujarat and Bengal, the Goddess is frequently
worshipped as a book. As the NTlamatapurihnatells
us, "O twice-born, books should be worshipped in
the temple of Durga. "4
The earliest and most beautiful manuscript illus-
trations of Sakta themes known to date portray the
Seven Matrkas, or Mothers, along with Gane'a (fig.
1). These figures occur on the inside of two wooden
covers now in the British Library.5Their diminutive
size notwithstanding, the depictions are excep-
tionally graceful and vibrant. Unfortunately, some
of the figures are more damaged than others, but
those colors which remain have preserved their
freshness remarkably well, considering the age of the
painting.
The Seven Mothers are led by Gan.esa,whose white
figure is silhouetted against a blue aureole and a red
halo. His trunk is outlined in pink; he sits gracefully
with his weight partially supported by his gray rat.
This is no ordinary rat, however, but is of gigantic
proportions. Obviously, the illuminator found it
incongruous to depict a rat of normal size carrying
such an enormous burden. Among Ganes'a's attri-
butes we can recognize the rosary, the battleaxe, and
his broken tusk.
Like Ganesa, each Mother is seated upon her
respective mount. Mahe'vari and IndrdnTsit with
their legs loosely crossed at the ankles; the others sit
in the more relaxed and graceful posture known as
lalitasana, where one leg is pendant. Each goddess
has four arms; the principal two hands hold a skull-
cup (right) and display the gesture of exposition
(left). This particular combination of the attribute
and the gesture with the two principal hands is a
feature common to Tantric images of the Goddess in
Nepal (fig. 5). It is also rather curious to note that the
upper right hand of each goddess holds the same
object, viz. a damaru, or "kettledrum." The distin-
guishing attribute of each is placed in the upper left
hand. Of those that can readily be recognized,
Mahesvari holds a trident, Kaumarfiaspear, Vaisnavi
a mace, V5r5hTan elephant goad, Indran.rTa thunder-
bolt, and C5mund- a staff that may be a khatvdhga.
As to their mounts, Brahman.i has her goose,
M5hesvari a yellow bull, Kaumari a peacock,
Vaisnav1 Garuda, V5r5hi a man, Indranii the ele-
phant, and C-munda a man. The complexion of
Brahman. is yellow, MWhesvarTwhite, Kaum5rT,
V5r5hi, and C-munda- are red, Vaisnavi dark green,
and Indra-ni.orange. IndranT.is further distinguished
by the third eye placed horizontally on her forehead;
M5hesvari does not appear to have a third eye.
Aesthetically, the illustrations are rendered in the
finest Nepali tradition. The outline of the figures is
firmly drawn, enclosing the elegant forms with a soft
42 PRATAPADITYAPAL
but sensuous grace. Very little shading is employed
to delineate the form, yet a remarkable sense of
volume is achieved solely in terms of line and color
contrast. Each deity is set off against a background of
a different color. Thus, the yellow Brahmd7ii, red
Kaum5ri, green Vaisnavil, and orange Indrdn7iare
presented against white; the white Mahesvari and red
VTr5hTagainst olive green; while Cmuunda's red
gains added intensity by being placed against a deep
green aureole. The haloes, too, are painted in
different shades of red, green, and mauve. In
addition, darker hues are applied to reinforce the
outlines of the aureoles, providing even greater relief
to the figures. Thus, a perceptible sense of depth is
created simply by manipulating the various color
surfaces.
In addition to the rich complexions of the
goddesses, the use of variegated drapery designs
confers both a sense of volume and an impression of
liveliness upon the various deities. All of these
(except Camunda) wear garments of beautifully
printed material of diverse colors. C-munda alone
wears an exquisitely rendered tiger-skin which is
more suitable to her depiction as an ascetic with an
emaciated body.
The earliest known painted representations of the
Goddess in her Mahis-asuramardini form occur in
three stray palm leaves which may have belonged to a
DevTmdhitmya manuscript and which are now in
the British Library (fig. 2). Only a single sentence is
written on the back of one of the folios (fig. 3), and it
may be read as follows:
om bhujendra [or bhujahga] kutilakara gajendra kadall-laya tha
salstra-madhye pradhanena likhitah kinna mucyate //
The inscription is written in poor Sanskrit and is
difficult to understand. I am indebted to Dr.
Gauriswar Bhattacharjee of the Berlin Museum for
the following comments and translation. It is not
quite clear whether the first word is bhuljendra or
bhujanga. If it is bhujendra, then it must have been
abbreviated by the author from bhujagendra, the use
of which would have disturbed the meter. A tentative
translation, as suggested by Dr. Bhattacharjee, is as
follows: "[Though] lord of the serpents [yet he has a]
shape, [and though] lord of the elephants [yet he has]
his resort in the banana plant [which is so weak] [and
therefore the person] who has written the foremost of
the scriptures, will he not be liberated?"
The inscription indirectly informs us that the text
must have been of the DevTmadhtmya, as this text
alone would be described as "foremost of the
scriptures" by a S-kta. We can also surmise that the
text was written by a literate scribe who, however,
was no Kalidasa. The style of writing, which is quite
beautiful, does help us to corroborate the date of the
pictures arrived at by stylistic analysis (see discussion
below).
In spite of the effaced condition of one of the
illustrations, it is clear that all three representations
of the Goddess are identical except for complexion
color; these are, respectively, red, green, and blue.
Each image is placed within a circular aureole of red
flames and is framed by red and yellow borders. The
space between the edge of the aureole and the yellow
border is painted deep blue. In each instance, the
Goddess is nimbate, has twelve arms carrying
identical attributes (mostly weaporns), and stands
astride her lion-vdhana and the decapitated buffalo.
No human being, however, emerges from the
buffalo's neck; instead, in each illustration the
Goddess releases two serpents which wrap them-
selves around the necks of two asuras. An identical
formula has been employed to represent each of the
asuras. Their militancy is apparent in their posture.
Each asura wields a club with one hand and, with the
other, attempts to loosen the coils of the deadly
snake. It is interesting to note the way in which the
figures of the asuras spill over the frame of the
illustration. In each of the representations, the
Goddess wears a fine, transparent dhoti that reveals
her firmly modeled legs. In addition, a jacket
designed like a cuirass hugs her voluptuous body.
The red goddess wears a blue jacket, while the green
and blue goddesses wear red jackets. On one of the
folios, a male and female devotee are seen worship-
ping the Goddess. They may represent the donor of
the manuscript and his wife.
This particular formula for representing Ma-
hisasuramardini has remained popular in Nepal.
Several representations in stone and bronze are
known. The interesting feature of all these de-
pictions is the convention that shows the Goddess, in
addition to killing Mahisasura, invariably attacking
two other asuras as well. It is difficult to ascertain
whether they depict a particular pair of generals,
such as Canda and Mun.da, or Sumbha and Nisum-
bha. In such representations, Mahisasura is usually
seen emerging from the buffalo's neck, but in these
particular illustrations, the artist has omitted the
human form (as is also the case with another version
of the subject rendered on a gilt-bronze manuscript
cover).6At any rate, it seems clear that this particular
iconographical variation, as well as the composi-
tional formula, was a peculiarity of the Nepali
artistic tradition.
Although diminutive, the illustrations are re-
markably animated and vibrant. They exhibit the
elegance of manner characteristic of better-known
Buddhist manuscript illuminations. They are, in
 EARLY PAINTINGS OF THE GODDESS IN NEPAL
fact, stylistically comparable to some fine Pani-
caraksii illuminations in the Los Angeles County
Museum of Art (fig. 4). There, too, we notice similar
compositions which depict kneeling figures repre-
senting forces of disease and evil, who tum away as if
unable to bear the wrath of the Goddess. In both
manuscripts, the drawing is exceptionally fine and
the compositions lively. The donor couple in the
Dev-m&hiitmya folio is especially engaging. Less
than an inch in size, they are fully modeled figures, at
once dignified and elegant. The details of their
garments and physical features, as well as the
offerings before them, are precisely articulated. In
general, the luminous colors of these small pictures
reflect the intensity and richness of early Buddhist
manuscript illuminations.
The most richly illustrated Nepali manuscript of
the Dezimdhdtmya is in the Bharat Kala Bhavan
(figs. 5-15). In terms of both narrative intent and
iconographic feature, this is the most elaborately
illuminated manuscript known to date. The pic-
tures, which are generally placed in the center of the
folio, illustrate literally the themes narrated in the
text. A few of the illustrations depict hieratic
representations of the goddess Durga, or Ambika. In
addition to these, the artist has provided some
remarkably lively renderings of combat scenes.
A typical hieratic representation (fig. 5) shows us
three devotees, wearing dhoti, ornaments, and tiaras,
engaged in adoring a white-complexioned goddess
who is seated on a blue lion. The four hands on the
right hold a skull-cup, an elephant goad, an arrow,
and a sword. Those on the left display a shield, a
bow, a noose, and the gesture of exposition.
Although the image does not correspond to any
particular description in the text, it may represent
the Goddess as Mah5sarasvati, the presiding deity of
the third part of the text. The red background is
strewn with flowers; these are seen in the majority of
the illustrations. The three devotees, shown in three-
quarter profile, stand or kneel quite naturally. Two
of them may represent King Suratha and the
merchant Samadhi, for whom the text was narrated
by the sage Medhas.
In another hieratic representation, two goddesses,
one white and one black, are shown seated in an
identical manner on a single lotus (fig. 6). The skull-
cup and the gesture of meditation are common to
both, but the white goddess carries a bow and an
arrow and the black goddess a sword and a noose.
Once again, the images do not agree with any of the
descriptions in the text but are obviously two
different manifestations of the Devi.
In two other folios (figs. 7, 8), two very spirited
images are shown being worshipped by devotees.
43
Although both assume the same militant posture,
there are striking differences between the two figures.
One of the goddesses is double-complexioned (fig.
7). Her face, torso, shoulders, and upper arms are
painted white, but her forearms, hips, and legs are
black. As far as I know, deities with two complexions
are encountered in the Vajrayana Buddhist pan-
theon, but there the two complexions are divided
along a vertical axis. Because of the damaged
condition of the illustration, it is not clear how many
arms the Goddess has, but the number appears to be
eighteen. In the second of these two images (fig. 8),
the Goddess is uniformly dark and is given at least
ten heads and probably an equal number of legs and
arms.
The half-white, half-black goddess with her
eighteen arms may represent the universal, or
visvariipa, form of the Goddess. The DevTmdhdtmya
tells us that she is to be worshipped in her eighteen-
armed form despite the fact that she has a thousand
arms (as!ddasabhujii pu-jya sai sahasrabhuij satU/).
The other figure with multiple heads and limbs is
almost certainly Mahakll, who is the presiding
deity of the first part (prathamacaritram) of the text.
She is described as follows:
I resort to Mahakali, who has ten faces, ten legs and holds in her
hands the sword, disc, mace, arrows, bow, club, spear, missile,
human head and conch, who is three-eyed, adorned with
ornaments on all her limbs, and luminous like a blue jewel and
whom Brahma extolled in order to destroy Madhu and Kaitabha,
when Visnu was in [mystic] sleep.7
Far more visually exciting are the narrative scenes
in which the Goddess (or another deity) is seen in
combat with various asuras. One of the folios (fig. 9)
illustrates a fight between a four-armed deity and
two asuras beside a water tank. Menacing serpents
coil around the feet of the deity, who has four arms
and carries a lotus, a conch, a club, and a wheel. The
divine figure is attacked by the two asuras. While it
may initially appear that this scene represents
Visnu's battle with Madhu and Kaitabha, who
emerged from the god's ears while he was engaged in
cosmic sleep on the ocean, the form of the deity is in
fact unusual. The figure seems to be bisexual. The
face with the third eye (unusual for Visnu) appears to
be identical with the faces of other images of the
Goddess. There is also a slight indication of a breast
on the left side of the chest, although this is not as
fully delineated as it is in the female figures.
However, unlike the Goddess, this figure wears a
short dhoti, which would indicate that, despite a few
feminine traits, the figure does indeed represent the
god Visnu. The composition is enlivened by the
fluid interaction between the figures and the ser-
pents, and the ocean is depicted conceptually, using
44 PRATAPADITYAPAL
a formula that is typical of contemporary Indian
paintings.
One of the curious features of the Devzmdhdtmya
is that the Goddess frequently fights a pair of asuras,
such as Sumbha and Nisumbha, or Canda and
Munda. Another is that the climax of the text is not
the death of Mahis-sura; rather, it is the destruction
of Sumbha and Nisumbha.
Can.da and Mund.a, the two generals of Sumbha
and Nisumbha, spot the Goddess in the mountains
and report back to their masters, praising her
incomparable beauty. Sumbha sends Sugriva to
persuade the Goddess to be his wife. In a charming
illustration (fig. 10), the Goddess, poised on a rock,
converses with the messenger, who has doubtless just
finished extolling his master's masculine prowess
and abilities. The Goddess is visualized as a demure
country girl who is out tending her goats and has met
a handsome young soldier. In a second composition
(fig. 11), we encounter the juxtaposition of two
different scenes. The seated Goddess appears to be
watching herself or one of her emanations destroy
the two asuras who had dared to insult her. Her
wrath is clearly expressed by the vigorous manner in
which she pulls the asura by his hair and simul-
taneously thrusts her trident into his body. The
second asura has already been slain. Other illus-
trations (figs. 12, 13) depict the Goddess engaging
two asuras, once with a sword and again with the
trident. These conflicts between goddess and asuras
are quite credible. The protagonists are the same size,
and the Goddess is shown without her divine
attributes. She is, in fact, essentially human-a
charming young girl who sometimes has a temper
tantrum and kills an asura or two. Not even a nimbus
is provided to suggest her divinity.
In two remaining battle scenes, however, the
Goddess is provided with additional arms so that we
are left in no doubt about her divine nature. In one of
these paintings (fig. 14), she is shown impaling the
asura Raktablija. Raktabija was a privileged asura:
every time he bled, another asura would be created
from each drop of his blood. Durga, therefore,
ordered K-arlto stretch her tongue out and lap up the
blood before it could fall upon the ground. In the
illustration, K-arl only her bust is shown-stretches
out her enormous tongue to swallow not merely a
drop of blood, but an entire asura.
The last combat scene (fig. 15) depicts the final
battle between Durg-aand Mahisasura. The Goddess
is now equipped with eight arms. She stands astride
the shoulder of the decapitated buffalo and the
attacking lion-vdhana. With one of her left hands,
she pulls the buffalo by its tail; with another, she
clutches Mahisasura's hair. The lion is about to take
a chunk from Mahis-sura's shoulder. The demon
looks around helplessly as he attempts to unsheathe
his sword. The composition, which is characterized
by considerable drama and action, continues the
basic formula which we have encountered in the
earlier representations of the theme (fig. 2). Here,
however, the human Mahisasura is included, and the
two additional asuras are excluded. In a sense,
therefore, the earlier illustrations encapsulate in a
single composition all of the various battles de-
scribed in the DeviTmdhdtmya.
The background of the Bharat Kala Bhavan illus-
trations is invariably painted red and is speckled
with flowers, symbolic of the divine presence. Most
of the figures are modeled by means of outline, with
the exception of the buffalo, whose form assumes
greater plasticity due to light shading and the
reinforcing of its outline. The figure types have
undergone slight modifications. The Goddess is
always shown with a naked torso, and her features
are also somewhat different. Of particular interest,
however, are the more convincing (because more
naturalistic) representations of the battle scenes in
the Bharat Kala Bhavan manuscript (as opposed to
their more hieratic depiction in the British Museum
folios). The artist has made obvious attempts to
enliven the different battles by showing both the
Goddess and the asuras in various postures and
positions. The compositions, as a result, are more
variegated, fluid, and visually exciting. Noteworthy,
too, are the colorful and decorative rocks with their
cubelike shapes, a hallmark of early Nepali paint-
ing.
To establish the dates of isolated book covers and
manuscript pages is always a difficult task. For-
tunately, we are on rather secure ground here because
of the enormous amount of dated Buddhist material
that has survived and because of the state of the
knowledge of Nepali paleography.
The three isolated folios representing the three
Durgas can be dated with fair certainty on both
stylistic and paleographical grounds. The script in
which the short inscription is written is described by
Bendall as the "early hooked Nepalese hand."8 The
paleography closely follows the script found in
various datable manuscripts in the Cambridge
University Library, including that of a Kurukullia
Kalpa, dated to 1179, and the Sddhanamdla--tantraof
1165.9 As we have already pointed out, the paintings
are stylistically close to the beautiful Paficaraksa
illustrations of the first half of the thirteenth century
(fig. 4) as well as to the painted covers of a
Sivadharma manuscript of 1139 in the Cambridge
University Library.'0 Thus, a date in the second half
of the twelfth century for these three isolated leaves
 EARLY PAINTINGS OF THE GODDESS IN NEPAL
would not be inappropriate. Stylistically, the covers
with the Seven Mothers are so close to the Durga
paintings that it seems obvious that they were
painted at about the same time.
I have elsewhere dated the Bharat Bala Bhavan
Devzmdh&tmyapaintings to the fourteenth century.
A more careful analysis of the paleography seems to
favor a date in the first half of that century. The style
of writing seems generally to conform to that seen in
a manuscript of Riama-nka-Natikia,written in 1360
and now preserved in the Cambridge University
Library." However, some of the letters continue to
use earlier forms typical of manuscripts of the
thirteenth century.
The stylistic elements of the Devzmdhihtmya
illustrations are closer to the Indian Museum
Astasiihasrika Prajniipiiramita manuscript'2 of 1367
than to such mid-thirteenth-century paintings as
those of the Los Angeles Panfcaraksa. While the
figures of the hieratic goddesses still retain the soft
sensuousness of the earlier representations, the
subsidiary figures, including the asuras, are not
quite as elegant as those in the Panicaraksii illus-
trations (fig. 4). The modeling seems somewhat
more attenuated and brittle, and both the propor-
tions and the facial features have changed sub-
stantially. However, the remarkable continuity of
the tradition is demonstrated by the fact that the
artist has used the same formula for delineating the
rocks as did his predecessor in the well-known
illustrations of the Astasaihasrikii Prajfnipdramita
manuscript of 1015 in the Cambridge University
Library. 'I
A painting on cloth from about the same period is
now in the collection of the Los Angeles County
Museum of Art (fig. 16). Stylistically, this cloth
painting does not differ from the numerous Bud-
dhist paintings surviving from this period. It is,
however, a rare art-historical document, for no
earlier painting of the Goddess has survived, even in
India. The painting has been damaged at the top, but
along the bottom are three panels, exactly as we find
in Buddhist paintings, which represent the per-
formance of various rituals by a priest, a lively scene
with dancers and musicians, and a row of donors.
The painting represents a mandala consisting of a
central figure of the Goddess destroying Mahis-sura;
both are enshrined within the pericarp of a lotus.
Eight other goddesses, identical except for their
complexions, are portrayed on the outer petals of the
lotus. Each is given eighteen arms; hence, we can
surmise that the effaced central figure must also have
had eighteen arms carrying different weapons and
attributes. Each of the goddesses pulls an asura by his
hair and thrusts a trident into his chest. There can be
45
little doubt that, together, the nine goddesses
constitute the Navadurga mandala. The central
lotus is superimposed on two intersecting squares
which produce eight triangles at the corners. These
triangles and the eight interstices contain sixteen
representations of the Mother Goddesses ($odas'a
Matrka). The octagon is surrounded by an outer
circle with a flaming perimeter. This circle is
inhabited by the eight guardians of the directions
(dikpala), each seated within a gateway and flanked
by a pair of goddesses of whom some are dancing, but
most are seated. The exact identification of this
group of sixteen figures is uncertain. Beyond the
fiery fringe, in the two lower corners, are two more
terrifying goddesses flanked by animal-headed com-
panions. Each goddess stands in pratyadi4ha posture
on a corpse; while one is multiarmed, the other holds
a chopper and a skull-cup with her two hands.
Presumably, there were two other similar groups in
the upper corners of the mandala.
The group of Navadurg-asis described twice in the
Agnipuriina, once in the Bhavisyapurd-na,and once
in the Kiranaigama.14 In all three texts, the nine
goddesses are said to be identical except for their
complexions and to possess eighteen arms. The
attributes and weapons, however, differ in the
different lists. The iconography here seems to
correspond closely to the description given in
chapter fifty of the Agnipura-na.
The nine Durgas have the following names:
Rudracan.d, Pracan.d.a, Can.dogra, Candanayika,
Candd, Candavatd, Candar-upa, Aticandika, and
Ugracanda-." Of these, Ugracan.da is the principal
deity and is represented in the center of the mandala.
Each goddess stands in alidha posture with the right
foot on the lion. With one hand, she pulls the asura
by the hair as he emerges from the buffalo and, with
another, drives the trident into his chest.
Nothing is said in any of these texts about the
other deities to be included in the mandala. Thus,
apart from being the earliest example, iconograph-
ically this fragmentary painting represents the most
elaborate mandala of Navadurgas known so far. In
fact, as de Mallmann observed, Navadurga repre-
sentations are so rare that she could cite only one
damaged example now in the Rajshahi Museum.'6 It
is also interesting to note that all nine names of the
goddesses contain the word canda, leaving no room
for doubt that they are the angry manifestations of
the Goddess. And yet, we are told in the Agnipuraina
that consecration of the Navadurgas increases sons
and other things for the devotee (navadurgaIhsyuh
sthdpyadhputrdddi-vrddhaye).'7
Despite its condition, there seems no doubt that
the mandala was executed sometime near the end of
46 PRATAPADITYA PAL
the fourteenth century. Stylistically, it can be placed
somewhere between the Vasudhara mandala of
about 1367 and the Los Angeles Visnu mandala of
1420.18This becomes particularly clear if we com-
pare the dancers and musicians in this mandala with
those in the Vasudhara mandala. In both paintings,
these slim figures are remarkably alike and are
characterized by a lively elegance. In general, the
drawing in the Navadurga mandala is particularly
fine; one can easily surmise how bright and
luminous the original colors must have been.
As is to be expected, there is abundant material for
the study of paintings related to the Goddess after the
fifteenth century. But the early paintings described
above are particularly important, not only for their
rarity, but also because they push the history of such
paintings back to at least the twelfth century. When
taken together with other known earlier examples of
Hindu paintings that have survived in Nepal, it is
evident that the tradition of mandala paintings and
manuscript illumination was not confined only to
the Buddhists. Meager though they may be, these
early Nepali paintings have particular relevance for
the study of Indian painting.
Few paintings depicting Hindu themes, either in
manuscript form or on cloth, have survived from
pre-fifteenth-century India. Most of the extant
paintings areeitherBuddhistor Jaina. Yet,it would
be difficult to imagine that manuscripts of the
Devimhiitmya and the Bhagavatapuriina were not
copied and illustrated by pious Hindus all over
medieval India. It would be wrong to assume that
Hindus in Nepal were more interestedin painting
thanwereHindusin India.The factthatso fewof the
early Hindu illuminated manuscriptshave survived
on the subcontinentis no reasonto believethat they
did not exist.
The recentacquisition by the Simla Museumof a
richly illustrated manuscript of a Devimiihdtmya,
purportedlybelonging to the sixteenthcenturyand
painted in the Panjab hills, is a case in point.'9
Simply because no example of sixteenth-century
paintings from that hilly region had come to light
previously, scholars interested in the field were
almost convinced that painting originated in the
area only in the seventeenthcentury,very possibly
due to Mughal influence. Such assumptionsare, of
course,absurd.20It would be worthwhileforscholars
of Indian painting to takea closerlook at Nepal. It is
this areawhich may provideinterestingclues to the
lost history of early Hindu painting in India.
 EARLY PAINTINGS OF THE GODDESS IN NEPAL
Notes
1. The manuscript is preserved in the National Library in
Kathmandu. In an article written in 1966,1 pointed out that
a manuscript of the Devimiahdtmya in the museum, which I
dated to ca. 1400, was the earliest known illustrated
manuscript of the text. See my "Paintings from Nepal in
The Prince of Wales Museum," Prince of Wales Museum
Bulletin, no. 10 (1967), pp. 4-6, figs. 4, 5. In view of the
evidence brought forward by the present article, the date of
the Prince of Wales manuscript must now be moved
forward to the sixteenth century.
2. For a detailed discussion of this manuscript, see my The
Arts of Nepal, 2 vols. (Leiden, 1974-78), vol. 2: Paintings,
pp. 33, 60; fig. 12.
3. Swami Jagadiswarananda, Devi Mdhdtmya (Madras, 1953),
p. xiv. The verse is as follows:
agnina- dahyam-nastu satrumadhyagata- rane /
visame durgame caiva bhaya-rtah saranam gata-h //
Most of chapter twelve of the Devr-mdhdtmyais devoted to
the benefits derived from chanting the text in the words of
the Goddess herself.
4. Ved Kumari, The Nflamata Purina (Srinagar, Jammu,
1968), p. 164. durgagrhe pustakana-m pija- kairya-tatha-
dvija. Also, lingstha-m puijayeddevrm pustakastham tath-
aiva ca, from the Yoginitantra, in Brhattantrasa-ra,ed.
Krishnanda Agamvagish (Calcutta, n.d.), p. 425.
5. The photographs of these two covers, as well as those for
figures 2 and 3, were kindly supplied by Mr. Simon
Mathews of London. The illustrations are rarely more than
two inches high.
6. Pal, The Arts of Nepal, vol. 1: Sculpture, fig. 14. The tying
of the asura with a serpent by the Goddess is described in the
Devfpurina (tarjjamianam hatam murddhni nagapa-s'ena
vestitam // chap. 32, v. 37).
7. Jagadiswarananda, Devi Maihatmya, p. 2.
8. For a discussion of the paleography of Nepali palm-leaf
manuscripts, see Cecil Bendall, Catalogue of the Buddhist
Sanskrit Manuscripts in the University Library, Cambridge
(Cambridge, 1883), Introduction.
9. Ibid., Table of Letters. Mss. ADD. 1686 and 1691.2.
10. Pal, The Arts of Nepal, vol. 2, fig. 54.
11. Bendall, Buddhist Sanskrit Catalogue, Table of Letters, Ms.
ADD. 1409.
12. Pal, The Arts of Nepal, vol. 2, figs. 28, 29.
13. Ibid., vol. 2, figs. 14, 15.
14. For a description of the Navadurgas, see Marie-Therese de
Mallmann, Les Enseignements Iconographiques de l'Agni-
Purana (Paris, 1963), pp. 147-49. The Devzpurina (chap.
42, v. 10) lists an important place of pilgrimage called
Navadurgasthala but says nothing about the shrine of the
image, except that the goddess was called Trimunda.
15. A different list of names of the Navadurga-sis found in the
Ma-rkandeyapurana (see Jagadiswarananda, Devi Mahat-
mya, p. xiv). They are: Sailaputri, Brahmacarini, Can-
draghanta-,Kusmanda, Skandamata, Ka-tya-yani7, Kalara-tri,
Mahagauri, and Siddhidatri.
47
16. Mallmann, Enseignements Iconographiques, p. 149.
17. Agnipurina, chap. 50, vv. 10-12.
18. Pal, The Arts of Nepal, vol. 2, figs. 1, 72.
19. This manuscript was discovered in the autumn of 1977 and
still awaits publication.
20. That mandalas were painted and manuscripts written and
perhaps illuminated by the Hindus prior to the eleventh
century is evident from the Devipurdna, a text of immense
importance for the Sakta religion. In chapter ninety-three,
in connection with the worship of the goddess Nandadevi,
Siva says the following about religious paintings on cloth:
patasya laksanam vaksye yatha sidhyanti sddhakdh /
granthikesavihi-ne tu ajtrne samatantuke // v. 148
asphitite achidre tu sthalenaiva samalikhet /
mangalaruipini karya jayddyaih parivarita // v. 149
vrddhena bhavate vrddho vyadhite vyadhito bhavet /
kuriupena kurupastui muirkhena tu na puijyate // v. 150
lekhakasya ca yadruipam citte bhavati teidrsTam
// v. 151
I shall now tell you the characteristics of patas or cloth
paintings with which the adept achieves his goals
immediately. The cloth for the painting should have no
knots or loose threads [literally, hair] and it should not
be old; all the threads should be smoothly woven. One
should rest it on a place without cracks or holes and then
paint the images of the auspicious Nanda along with
Jaya and other attendants. If the painter is old, then the
image also looks old; if he is indisposed, then the image
looks ill; if he is ugly, then the picture is without beauty.
One should never worship a painting drawn by a fool.
The form of the image will always reflect the condition
of the artist's mind.
In an earlier chapter (ninety-one), the author of the
purana discusses the importance of donating books to
brahmins. A detailed discussion of this chapter must be
postponed for another occasion, but some of the informa-
tion is relevant for us. Rather than quote the original text, I
will briefly summarize the pertinent portions. We are told
that the manuscript should consist of palm leaves and
should have two wooden boards, to be tied together with
either red or black thread. One should write very carefully,
making sure that the letters do not crowd one another and
that the writing is legible. When finished, the manuscript
should be placed on a decorated cloth and worshipped
along with the painting of the deity invoked in the text.
Although we do not know the exact date of the DezT-
purina, that chapter ninety-three is earlier than 1050 is
evident from the fact that almost the entire chapter is
quoted by Hema-dri and Laksmi-dhara,both of whose dates
are known. It is probably earlier than the Kalikapuriana,
which is usually dated to the eleventh century, and was
probably compiled in its present form sometime between
the sixth and tenth centuries. In both chapters ninety-one
and ninety-three, the word pustaka ("book") has been used.
This is a loan word from Old Persian and could not have
been introduced before the sixth century, as it is not
included in Amarasim.ha's Amarakosa, generally regarded
as a sixth-century lexicon.
In any event, these references to both painted images and
manuscripts are among the earliest in Sanskrit religious
literature. Especially significant is the discussion of the
48 PRATAPADITYAPAL

patacitras, which confirms the suggestion made above that painting (among other objects) as suitable images for the
painted images of the Goddess in India must have existed worship of the Goddess (pustakasthiim mahiidevi piaduke
much earlier than the present evidence indicates. Another pratimaisu ca / citre va trizikhe khadge jalasthaim vapi p2u-
passage in the text prescribes both a manuscript and a jayet // vv. 93, 105). See also n. 4 above.
 PLATE 1
PAL

I~~~~~~~~~~~~~~~~~~~~~~~~P 7-

'~ ~ Jw .\ * gv4'.'~

Ai

FIG.I1. Covers with the Mothers;co o w

FIG. 1. Covers with the Mothers; colors on wood. Twelfth century. The British Library.
PAL PLATE 2

I10U'_w __:

1 14

FII. 2. Folios representingIMahisisuramardini; colors on wood. Twelfth century. The British Library.
b_a
~~~~~~~~M

_ .*, _ .. s * t se, ,. W~~~~~~~~~~~~~~Z

l- ...

xe ' t' '

s , iv \.,,,e, ,S;tMr

| E [ S ' \ L i 4s 5, |~771
4 t w _
PAL PLATE 3

4t

-4i~~~~~~

FIG. 3. Reverse of a folio in fig. 2, with inscription.

F. 44

Rvre an AoliceHermn
Th.3 Nas ecCollewthincrition.
i rl_ _ A r- 'n I T v %A<__r
...... r, A-p

.%1|J~4 ? :;j
PAL PPLATE 4

FIG. 5. Mahisarasvati (?), folio from a Devimihitmya ms.; colors FIG. 6. Durga and Kill (?),same ms. as fig. 5.
on palm leaf. Fourteenth century. Bharat Kala Bhavan, Varanasi.
Photographs of figs. 5-15, courtesy, American Institute of Indian
Studies, Varanasi.

4MG.
8

FIG. 7. Universal form of Durga, same ms. as fig. 5. FIG. 8. Maha-k1il, same ins. as fig. 5.
 PLATE 5
PAL

I77U

wilw~~~~T~

sI~~~~~~~~~~~~~~~~~~~~i~~~~~I

FIG. 9. Visnu fights Madhu and Kaitabha, FIG. 10. Devi conversing with Sugriva, same ms. as fig. 5.
same ms. as fig. 5.

FIG. 11. Devi destroying asuras, same ms. as fig. 5. FIG. 12. Devi destroying asuras, same ms. as fig. 5.
 PLATE 6
PAL

~-I- 1

tAuI

FIG. 13. Devi- destroying asuras, same ins. as fig. 5. FIG. 14. Devi- and Kal11fight with Raktabi-ja, same ms. as fig. 5.

FIG. 15. Devi fighting Mahisasura, same ms. as fig. 5.

PAL PLATE 7

FI.1.Naaug mnaa clr ncotn 37-40 LsAgls onyMuemo At h NsiadAlc eraaekColcin

FIG. 16.Navad mandal; colorson.cott ..on. -40