The Image of Maheshwara

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 50

The Image of Maheśvara:

An early example of the integration of


Hindu Deities in the Chinese and Central
Asian Buddhist Pantheon

Riccarda Gallo
567131
MA History of Art and Archaeology

This dissertation is submitted in partial fulfilment of the requirements for the degree of MA History of Art
and Archaeology of the School of Oriental and African studies (University of London).

September 16th 2013

9.991 words

1

Table of Contents


Acknowledgments .............................................................................................................................. i

Abstract .............................................................................................................................................. ii

Introduction ....................................................................................................................................... 1

Chapter 1

1.1 The Yungang Caves and the first Maheśvara figures……………………………………….7

1.2 Cave 285 at Dunhuang……………………………………………………………………..19

1.3 The votive tablets and paintings of Khotan………………………………………………..21

1.4 Kizil and the Umāmaheśvaramūrti iconography…………………………………………...27

Chapter 2

2.1 The influence of the Maheśamurti image on the Śiva figures of China and Xinjiang..........30

2.2 Wešparkar: Maheśvara’s Sogdian counterpart……………………………………………..40

Chapter 3

3.1 The process of assimilation of Hindu deities into Buddhism……………………………....42

3.2 Spatial location of the Dharma protectors at Yungang and Dunhuang………………….....44

3.3 Some historical and social elements linked to Maheśvara’s representation………………..46

Conclusion………………………………………………………………………………………….49

Bibliography…………………………………………………………………………………………

Appendix. List of Plates……………………………………………………………………………

2

Acknowledgements

First of all I would like to thank my supervisor Lukas Nickel for his support and precious advices

concerning the writing of this dissertation. I would also like to thank Dr. Luca Milasi for his

noteworthy help with the translation of the Japanese text Tonkô-ga No Kenkyû. Lastly I would like

to thank Matteo for his patience and sustain throughout this year of Master.

3

Abstract
This essay aims to analyze the process of integration of Hindu deities inside the Central Asian and

Chinese Buddhist Pantheon through the analysis of a particular divinity, Maheśvara, an epithet used

in Buddhist sūtras for the Brahmanical god, Śiva. This deity was depicted on early Buddhist visual

material from the Tarim Basin and China, and was characterized by iconographic uniformity during

the early period of the so-called “esoterification of Central Asia” and China (5th-8th century A.D.)

(Sørenesen1991/1992: 287). This image seems to testify the influence of Hindu iconography on

Central and East Asian Buddhist art. Moreover, the process of incorporation of the Hindu deity into

the Buddhist pantheon results in the attribution of a new role of protector to Maheśvara. The choice

of this deity was led by both its particular iconography, which allowed for easier detection amongst

other foreign divinities, and by its frequency in depiction.

4

Introduction

Since its origin, the Buddhist religion incorporated some of the Brahmanical deities and ideas.

Therefore, when Buddhism eventually reached Central Asia and China it brought these Hindu gods

with it. With regards to Śivaism, there is a legend attributing the merge of this religion with

Buddhism to the Buddhist philosopher Asaṃga (400 A.D. ca.). It is said that he tried to unite the

two divergent faiths by placing Śaiva deities into the Buddhist Pantheon, as devotees of Buddha and

Avalokiteśvara (Rhys Davids 1886: 208). The reasons for this integration were mostly social. On

one hand, it allowed for the easiest adoption of Buddhism by the foreign tribes, as they could still

worship their own deities. On the other hand, both religions had common characteristics: they were

ascetic and were supported by the merchant classes (Banerjee 1970: 282-283). Śiva was the most

represented Hindu divinity in early Buddhist sites of China and Xinjiang, and was designated by the

Sanskrit epithet Maheśvara (Chinese Daijizaiten) in the Buddhist sūtras (Iyanaga 1983: 714-715).

Despite Maheśvara’s particular iconography, showing a clear foreign origin, and its frequent

appearance in Central and East Asian Buddhist art, very few scholars published works exclusively

dedicated to him. The first researcher that dealt with Maheśvara’s iconography is the Japanese

scholar Matzumoto; in his publication Tonkô-ga No Kenkyû (1937), he studied the Dunhuang

paintings and manuscripts. He dedicated a short chapter mostly comparing later Maheśvara

depictions found at Dunhuang with the descriptions of the deity found in the Taisho tripitaka.

However, his work demonstrates that there is no evident connection between Maheśvara images and

the texts. Subsequently, Williams is the first scholar to briefly investigate the origin of the

Maheśvara figure in her particularly interesting work on the Khotanese iconography (1973). The

more recent research on Maheśvara’s image was conducted by Zhang (2009) who analysed the

guardian figures depicted on the west wall of cave 285 at Dunhuang, attempting to demonstrate

their foreign derivation, connecting it to the work of Sogdian artists. However, the scholars that

5

dealt with the image of Maheśvara in the Buddhist East Asian areas mostly worked on cataloguing

the figures or were concentrated on specific sites without giving an exhaustive iconographic

investigation.

This dissertation aims to analyze the first depictions of Maheśvara in Buddhist sites of

Xinjiang and China, from the 5th to the 8th century A. D. in order to retrace the early process of

acquisition of Hindu deities by the Central and East Asian Buddhist Pantheon. These initial

representations were characterized by iconographic uniformity. In fact, even if some attributes

changed according to the region or to the related type of Buddhism, Maheśvara was almost always

depicted with three differentiated heads, holding the sun and the moon with the upper hands and

sitting on his vehicle Nandi; all these characteristics reflect Northern Indian Śiva’s iconography.

After the 9th century A.D., it seems that the Maheśvara images started to lose this iconographic

uniformity, instead acquiring different forms characteristic of Tantric Buddhism, thus detaching

from their Hindu origins.

The first chapter will provide both a description and an iconographic analysis of early

Maheśvara figures found in Yungang, Dunhuang, Khotan and Kizil sites. The second chapter will

demonstrate the Hindu origin of these images, suggesting the influence of Maheśamurti’s concept

and sculptures on early Buddhist Śiva representations in China and in the Tarim Basin.

Furthermore, the hypothesis of Sogdian appropriation of the Maheśamurti iconography for the

depiction of the god Wešparkar will also be explored. Finally, the third chapter will investigate the

new role that Maheśvara acquired within the Buddhist Pantheon, both through the textual and the

visual materials, analysing the deity’s images in their own context.

6

Chapter 1

1.1 The Yungang caves and the first Maheśvara figures

The Yungang caves represent an essential document in order to understand the early Chinese

Buddhist iconography and its development (Knauer 1983: 28). Moreover, they house the first

depictions of Maheśvara outside the Indian subcontinent. The grottoes were constructed during the

Northern Wei Dynasty (386-534) and can be categorized in three phases. The first phase, namely

the Tan Yao Caves includes grotto 16 to 20; the second phase comprises cave 1 to 3, and 5 to 13;

and finally the third phase includes the remaining grottoes (Yi 2010: 1-2). It is very significant to

underline two characteristics linked to the images of Maheśvara at Yungang. Firstly, the deity

figures are only found in the second phase caves (12, 8, 10) except for one image carved on the

reveal of the window of grotto 35, belonging to the third phase. Secondly, the deity is always

accompanied by his son Kārttikeya (variants Kumara or Skanda), forming a constant iconographic

motif together. The second phase caves were probably built between 465 and 494, which

correspond to the Emperor Wencheng’s death and the relocation of the capital to Luoyang (Yi

2010: 53-54).

According to Yi’s re-examination of the Yungang chronology, the first representation of

Maheśvara was found in cave 12, which probably was constructed during the reign of Emperor

Xianwen (466-471) (Yi 2010: 59). The cave’s ceiling was decorated in the zaojing pingqi style,

namely a series of squared or trapezoidal panels (Yi 2010: 70) into which both Maheśvara and

Kārttikeya were depicted [Plates 1-2]. Both deities hold the sun and the moon with the upper hands,

have six arms, and one and five heads respectively. They are seated in the royal position next to

their vehicles: the bull and the peacock. This is the only case at Yungang where Maheśvara is

represented with one head, probably demonstrating that its iconography was not yet fixed.

7

Plate 1

Plate 2

8

However, the most well preserved Maheśvara figure is found in cave 8, which belongs to the

so-called “twin caves” built during Emperor Xiaowen (471-499) and Empress Dowager Wenming’s

reign (Yi 2010: 59). The deity is carved on the east reveal of the entrance arch [Plate 3]. He is

represented seating in lalitāsana on a crouching bull, Śiva’s vehicle Nandi. He is three headed and

eight armed. The lateral heads are undifferentiated, as opposed to the other representations of

Maheśvara found in China and Central Asia. They bear what appear to be Phrygian caps, a common

attribute of princes and deities in Kuṣāṇa art (Soykut 2009: 213, 217) [Plate 4]. These hats may be a

proof of the Central Asian influence on Yungang art, which is corroborated by the “many recorded

embassies, which for decades, arrived at the Northern Wei court” (Knauer 1983: 44) from

Gandhāra, Bactria and Sogdiana. On the other hand, the central head wears a crown with a crescent

moon on the top, the symbol of Śiva. Of the eight arms just five are still intact. The right ones hold -

from top to bottom - a cosmic symbol (either the sun or moon), a bow, an unidentifiable object with

a hole in the middle and a bunch of grapes. The grapes, are clearly not a Buddhist attribute,

according to Trombert, they attests the Sogdian presence at Yungang (Trombert 2005), which will

be investigated in Chapter 2. Moreover, the only remaining left arm is placed on Maheśvara’s lap.

9

Plate 3

Plate 4

10

On the west reveal of cave 8 is another figure that we identify with Kārttikeya, Śiva’s son

[Plate 5]. He is five-headed, eight-armed and sits on a sinicized peacock which turns his head

towards the deity and holds either a pearl or a ball in its beak (Baumer 2009: 178; Mukherjee 1987:

253). The arms are still intact except for the right middle one, which probably used to hold an arrow

as, the corresponding left one holds a bow. The upper arms sustain cosmic symbols such as the sun

and moon, while the lower left hand holds a cockerel at chest height. The right hand shows an

unknown mudrā and is placed on his thigh (Mukherjee 1987: 253). Both Caswell, Mizuno and

Nagashiro identify the figure as Vishnu, probably due to the fact that they exchange the peacock

with Garuḍa, Vishnu’s vehicle. However, the cockerel is a typical attribute of Kārttikeya [Plate 7]

which is totally absent in Vishnu images.

In fact, it is narrated that Kārttikeya received the cock by Aruna, and the peacock by

Garuda. Moreover, the fact that he is represented with youthful smiling heads can be associated

with the Viṣṇudharmottara, an iconographic text which describes the deity as «boyish in appearance

“Kumara, one who is eternally endowed with youth”» (Banerjee and Banerjee Sarkar 2008: 44).

This text also portrays the divinity as riding a peacock, holding a cock (kukkuṭa) and a lance (śakti)

(Mukherjee B. N. 1987: 250). In addition, the deity’s epithet in the Mahābhārata was Ṣaḍānana,

literary “having six heads” (five visible and one invisible) (Mukherjee 1987: 253). Kārttikeya was

adopted by the Buddhist pantheon as the “Commander-in-Chief of the army of the gods”, which is

why he rides a peacock, a symbol of his victory against enemies, and holds a bow and arrow

(Chandra 2009b: 165). Maheśvara and Kārttikeya’s Śivaite origin is corroborated by the presence of

the lower part of both east and west reveals of two attendants holding a triṣūla, a trident, which is a

typical attribute of Śiva (Daniélou 1964: 216) [Plate 6].

11

Plate 5

Plate 6

12

Plate 7

13

Maheśvara and Kārttikeya seem to be portrayed also on the lintel of the door that connects the ante-

room to the main shrine of cave 10 [Plates 8-9]. The scene represented is probably the Mount

Sumeru, formed by three rows of mountains, the upper level corresponding to 16 peaks (Mizuno

and Nagashiro 1938-1945, Vol. 7 text: 108). On both sides of the mount are depicted two multi-

headed and multi-armed deities. The one on the right is three-headed and has four arms, where the

upper two hold the sun and moon while the remaining hold a round object to the chest. The lateral

faces wear Phrygian caps, while the central one wears a crown. The left figure has five heads and

six arms, with the upper hands holding the sun and moon, the middle ones holding a bow and arrow

and the lower ones holding a ball or cintāmaṇi to the chest. The two figures look specular, sitting in

the royal position and with the lower left arms placed on their thighs. It is apparent that these

figures are Maheśvara and Kārttikeya (Baumer 2009: 178; Sørensen 2011b: 259). In fact, except for

the vehicles which are lacking, the other iconographical characteristics are very similar to the

figures depicted in cave 8. Moreover, the fact that they are represented together on a doorway’s

lintel supports this hypothesis.

14

Plate 8

Plate 9

15

The latest depictions of Maheśvara and Kārttikeya were found in cave 35, on the east and

west reveals of the main window situated above the entrance door [Plates 10-11]. An inscription

placed at the entrance of the grotto confirms the date of construction: “4th year of Yanchang” which

corresponds to 515 (Yi 2010: 101). The cave clearly belongs to the third phase, which is attested by

the sinicized representations of the deities. Unfortunately, the figure of Maheśvara is extremely

ruined, but the bull Nandi, on which the deity is sitting, can be recognized. The image of Kārttikeya

shows three heads, instead of the usual five, and six arms. These attributes are unrecognizable, but

the divinity is visibly riding a peacock. The sinicization typical of the third phase caves can be seen

in the faces, the headgear and the bird traits, which appear more elongated and diverge from the

representations in cave 8. These deities have been recognized to be the same as the divinities

depicted in cave 8, in fact they also acquire the same position: Maheśvara is depicted on the east

reveal and Kārttikeya on the west (Mizuno and Nagashiro 1938-1945, Vol. 15 text: 176; Yi 2010:

xi).

Plate 10 Plate 11

16

The iconographic motifs of Maheśvara and Kārttikeya were also found on a Buddha statue

housed in the Freer Gallery of Art in Washington D.C. [Plates 12-13]. This sculpture belongs to the

Sui dynasty (581-618) and represents a standing Buddha draped in robes with low reliefs portraying

the realm of existence (Howard 1984: 53). The cosmological scenes are divided in registers; on the

5th register, the central element of the composition is a to-pao t’a (“many treasures”) stūpa, which

symbolizes the Parinirvāṇa of the Buddha (Howard 1986: 19). This kind of stūpa originated from

the Indian Gupta tradition, and is surrounded by several beings who came to worship the Buddha

for the last time. Among them, on both sides of the stūpa, are Śiva and Kārttikeya who are holding

the sun and the moon with the upper hands. They are recognizable thanks to their vehicles: the bull

and the peacock (Howard 1984: 55). Apart from demonstrating that Hindu Gods were a common

motif during the Six Dynasties epoch, this statue specifically proves that Maheśvara and Kārtikkeya

were often represented together in Chinese art. Moreover, Howard points out that the

“representation of these gods follows the formula established at Yün-kang (which is faithful to the

original Gandhāran model)” corroborating the argument that Maheśvara’s iconography originated in

the Indian subcontinent (Howard 1984: 55).

17

Plate 12

Plate 13

18

1.2 Cave 285 at Dunhuang

Only one example of Maheśvara’s image was found in the cave complex of Dunhuang, depicted on

the west wall of Grotto 285 [Plates 14-15]. This particular grotto is one of the few caves in

Dunhuang to be accurately dated. It bears an inscription proving that the end of the construction of

the cave was 538-539 A.D. (Western Wei Dynasty). Moreover, it is the largest vihāra-type,

monastery, within the whole site and is characterized by a very innovative ornamentation reflecting

Central Asian and Indian influences (Whitfield 1995, Vol. 1: 284). The west wall, which is situated

in front of the entrance, is partitioned in three niches: the middle housing a Buddha seated in the

European way and the other two niches housing meditating monks. On both sides of the central

alcove are depicted two families of Hindu deities, each one accompanied by two guardian kings

representing the four directions (Zhang 2009: 35). On the Buddha’s right, there are Viṣṇu (in his

Buddhist form Nārāyaṇa) and Indra (in the form of Śakra), recognizable by the third eye; while on

the left there are Śiva (or Maheśvara), Kumāra riding a peacock and Gaṇeśa (or Nandikeśvara)

recognizable by his elephant head (Whitfield 1995, Vol. 1: 287). Here, Maheśvara, unlike the

representations found at Yungang, is depicted with three distinguishable heads: the left has a

ferocious expression while the right has a feminine appearance. Even the colour of the faces is

different: the lateral faces are pale pink while the main face and body are black. He is seated in the

lalitāsana position on the blue bull Nandi. With his upper hands he holds the sun and moon, while

with the middle hands he holds a bell and an arrow, and the lower arms are positioned in front of his

chest (Zhang 2009: 35). An interesting element is the wind god Vāyu, placed in the crown of

Maheśvara which was also found in cave 249 at Dunhuang and cave 39 at Kizil. This image

ichnographically originated in Central Asia, in fact, the first depictions are found on Kushan coins

(Zhang 2009: 38). There are no comparable representations of Maheśvara with Vāyu depicted in his

crown, but it is hypothetical that this iconography has an Indian origin, inasmuch as, the Indian god

Vāyu was said to house in Śiva’s headdress (Grenet F. 1994: 43).

19

Plate 14

Plate 15

20

1.3 The votive tablets and paintings of Kothan

Maheśvara figures found in the area of the Khotan reign were mostly recovered at Dandan-Oilik, a

very important fortress situated not far from the capital city Yōtkan (Whitfield 2004: 134).

Although it has to be noted that the date of Khotan sites is still being debated. Through the C14

analysis of a wooden pillar found in 1999, archaeologists established that Dandan-Oilik earliest

artefacts, such as the wooden tablets, belonged approximately to the 6th century A.D. (Whitfield

2004: 158). However, Chinese literary sources found at Dandan-Oilik and belonging to 781-789

A.D. testify that Khotan reached its apogee during the 8th century A.D. when its major cities

Dandan-Oilik, Khadalik and Balawaste formed one of the four Tang Garrisons of Anxi, together

with Kucha, Kashgar and Karashahr. The end of the apogee coincides with the Tibetan conquest of

the region in 790 A.D., and consequently the interruption of the artistic evidence (Williams 1973:

112). Therefore, according to Williams, the Khotanese paintings belong approximately to the lapse

of time between 7th and 8th century A.D. (1973: 109-110).

Maheśvara figures were mostly depicted on wooden votive panels, except for two wall

paintings found at Balawaste and in shrine D13. The first important site is the so-called DVII b, a

housing quarter where Stein discovered a significant votive tablet painted on both sides depicting

Maheśvara on the obverse and the Persian prince Rustam or the Silk God of Khotan on the reverse

(Stein D VII 6) (Baumer 2000: 79-80) [Plate 16]. The recovery of the votive panel in a house proves

that Maheśvara worship was quite widespread in Khotan and was of a familiar type. The fact that

the wooden tablet was painted on both sides demonstrates that it was probably not fixed to a wall

but just endorsed somewhere and was therefore moveable (Giès – Cohen 1995: 301). Moreover, it

seems that the panel had not only a votive role but was also an object of worship by itself (Mode

1991/1992: 180). The deity is represented surrounded by an aureole, sitting on a cushion with a

chessboard pattern, resting in turn on two bulls (Williams 1973: 142). This depiction is identical to

the wall fragment found at Balawaste [Plate 17]. In fact, both deities are three-headed with the main

21

body depicted in blue, the feminine face in white and the ferocious head with red hair. The central

face bears a third eye on the forehead, a moustache, heavy round earrings such as the Kizil

examples, and a crown (D VII 6) or a skull (Balawaste 0200). Moreover, the deity is ithyphallic,

wears jewels, a yellow scarf on the shoulders and a tiger dhoti. With his upper hands he holds the

sun and the moon, with the left a tridaṇḍī or triple stick, and a pomegranate (Balawaste) or a

ḍamaru, drum typical of Śiva (D VII 6) with the right hand (Chandra 2009b: 168; Williams 1973:

142). The only difference between the two representations is that the disposition of the celestial

attributes and the two lateral faces is reversed (Andrews 1948: 22-23).

Two Buddhist texts - the Karoragyūshotenmitsugonkyō and the Jialouluo Wang ji zhutian

miyan Jing (“The Sutra of the king Garuda and of the secret mantras of each divinity”) - portray

Maheśvara with the main body depicted in blue and the lateral faces in red and white. He has three

heads: the central face of a deva, a celestial being, the right face of a ferocious yakṣa and the left

face with feminine aspects (Granoff 1979: 79; Matsumoto 1937: 732).

Plate 17
Plate 16
22

The most important site with regards to the quantity of recoveries is the Shrine D 13, where

archaeologists found rather well-preserved paintings on the four walls. Among these paintings two

represents a divine triad of Hindu origin (Baumer 2000: 85) [Plates 19-20]. Furthermore, the

wooden tablet D 10.8 was probably unearthed from this site, as Baumer hypothesizes that Stein’s

temple X corresponds to the actual shrine D 13 (Baumer 2009: 175-177). The composition of the

two divine triads is very similar. On the two sides there are Hindu gods with their attributes, while

in the centre, there is a feminine figure with babies on her lap. Moreover, all deities bear a nimbus

and look to the right side, which is a typical characteristic of Hindu divinities found at Dandan-

Oilik (Baumer 2009: 176). The triad discovered on the western inner side of the northern wall

depicts on the left three-headed Maheśvara adorned with a diadem (keyūra), and equipped with its

classical attributes, such as the bull Nandi, the lingam, a fruit (mātluṇga?), and the sun and moon

(Lo Muzio 2008: 186-187; Baumer 2009: 176). The deity on the right side was identified by

Baumer with Brahma, but may represent instead Mahākāla due to his terrific aspect underlined by

the spiders on his chest and the snakes twisted around his arms (Lo Muzio 2008: 188-190). In the

middle is represented a woman with a child, probably Hārītī, which demonstrates powers of

promoting fecundity and protecting children (Chandra 2009b: 166). The southern end of the western

wall triad is quite similar; Hārītī is in the middle, while there is Kārttikeya-Skanda on the left and

Varaha (third avatar of Vishnu) or a graha on the right. Baumer confuses Kārttikeya’s vehicle the

peacock with the goose Hamsa, therefore identifying the deity with Brahma (Baumer 2000: 90).

However, as seen at Yungang and Dunhuang, the cockerel, the peacock, the bow and the arrow are

typical of Kārttikeya’s attributes. Another image portraying the deity was found in 2002 in temple

CD 4 at Dandan-Oilik. Kārttikeya is depicted with the same attributes as in D 13, holding the

cockerel with the left hand and riding the peacock. However, instead of the bow and arrow, he holds

the sun and the moon (Guorui, Tao, and Yuzhong 2008: 157-158) [Plate 18].

23

It is important to highlight that shrine D 13 was probably dedicated to children or to the

worship of people that wanted descendants (Baumer 2000: 90). This theory is corroborated by the

fact that Maheśvara, Kārttikeya, Hārītī and the graha have the role of infants’ protectors. As far as

Maheśvara is concerned, he was seen as a popular deity, the defender of families. Two types of

invocation were dedicated to him, one in order to obtain an heir, and the other to ask for protection

in dangerous circumstances. For example, the Mahātyāgavat and the Koṭikarṇa were two Jātakas

dedicated to Maheśvara, concerning the obtaining of children (Iyanaga 1983: 733-34). Moreover,

on the Lala Bhagat pillar, Kārttikeya was called with the epithet Bālapyāditya, “one who is the

protector of children (…)” (Mukherjee 1987: 252). While, according to Baumer, graha was a

Khotanese spirit guaranteeing the health of offspring (Baumer 2009: 179).

Plate 18

24

Plate 19

Plate 20

25

Another votive tablet (D 10.8) was found in shrine D 13 representing Maheśvara accompanied with

Pārvatī holding a cup, recalling Kizil Umāmaheśvaramūrti’s iconography. Unfortunately the tablet

is quite damaged and the painting is obliterated. However, some of Maheśvara’s characteristics are

still recognizable. He is three-headed and four-armed; the left face has a ferocious appearance,

while the right one is not identifiable anymore but probably had feminine traits. Some features are

similar to the DVII 6 image, particularly the vajra hold in the left hand, the necklace made of bow,

the red tiger dhotī and the lingam (Williams1973: 112) [Plate 21].

Following to Williams, there are three other representations of Maheśvara found in Khotan;

however, their state of conservation does not really allow for identification of the deities

represented. Among these there is the Wooden Panel No. 1925,0619.36 [Plate 22], which is housed

in the British Museum and was probably acquired by Clarmont Skrine in the 1920’s at an antiquity

shop in Khotan (Waugh– Sims-Williams 2010: 96); but also a mural fragment from Kuduk-kol 014

and another votive tablet named by Williams as Kha. i. 51 (Williams 1973: 143).

Plate 21 Plate 22

26

1.4 Kizil and the Umāmaheśvaramūrti iconography

The images of Maheśvara found at Kizil are of a particular iconographical type, called

Umāmaheśvaramūrti, namely the representation of Śiva accompanied by his wife Pārvatī (Umā), no

depiction of Maheśvara alone is found in the grottoes. In this site both Śiva and his consort are

always depicted in añjalimudrā (the attitude of prayer) around the Buddha figure. There are four

examples of Umāmaheśvara in Kizil, of which one is almost obliterated, with only the deities’ busts

visible (Cave 189) [Plate 25]. The others come from the Devil Cave (600 ca.) [Plate 24], Māyā

Cave (416-525) [Plate 23] and Gorge Grotto (600 ca.) [Plate 26-27] (Lo Muzio 2002: 56, 58). It is

significant to point out that the figures of Umāmaheśvaramūrti found at Kizil are characterized by

uniformity in the iconography. Maheśvara is represented next to Pārvatī, sitting in the European

position with crossed ankles on Nandi. He wears a

tiger skin around his chest, is three-faced with

ferocious and feminine lateral heads, and in three

examples, the faces are surrounded by a halo and

bear round earrings. Moreover, according to Lo

Muzio, all the figures hold the sun and the moon

with their upper hands (Lo Muzio 2002: 52). The

only differences are in the fact that in two of the

paintings, Śiva embraces Pārvatī (Gorge, Māyā Plate 23

Caves), while in just one case, the deity is ithyphallic (Gorge


Cave). It is to be observed that all the

Umāmaheśvara figures differ from the images of Maheśvara found at Yungan, Dunhuang and

Khotan, which represent icons used for worship and meditation, while in this case they are included

in a narrative context and are found around the Buddha figure. In the Devil Cave, the couple is

depicted in a parivāra scene, while in both the Gorge and Māyā Grottoes the “conversion of the

yakṣa Āṭavika” (Lo Muzio 2002: 52) is included in the scene.

27

Plate 24

Plate 25

28

Plate 26

Plate 27

29

Chapter 2

2.1 The influence of the Maheśamurti image on the Śiva figures of China and Xinjiang

The figures of Maheśvara uncovered in Buddhist sites in China and Xinjiang seem to be a reflection

of what scholars call “Maheśamurti images.” The idea that the image of Maheśvara came directly

from the three-headed Indian sculptures of Śiva cannot be attested with certainty but their

iconographic influence is highly probable. In order to demonstrate this thesis the Maheśamūrti

image will be explained.

Three-headed Śiva sculptures were found in India from the beginning of the late Kuṣāna

dynasty until the 10th century A.D. During the medieval period they probably were linked to the

Lakulīśa-Pāśupata sect through which they had strong popularity (Kumar 1975: 105). These figures

have been the subject of scholar debates for years who attributed them the significance of a

Trimūrti, namely the Brahmanical Triad composed by Brahmā, Viṣṇu and Śiva. However, in 1916,

Rao in his volume Elements pointed out that this ascription was not correct, as ichnographically, it

did not matched. The three-headed image was instead explained as a Maheśamurti figure, “the fully

manifested Supreme Śiva endowed with the powers of creation, protection and destruction” as it is

found in the Śaivāgama (s) (Banerjea 1955: 122). Banerjea partially agreed with this theory saying

that instead it seems to represent both the calm and ferocious Śiva together with his consort Umā

(Pereira 1968, note 12: 86). However, his opinion contradicts the estimation of several scholars

such as Barrett, Collins, Kumar, and Pereira who support Rao’s theory.

Kumar in his article written in 1975, suggests a clearer iconographic distinction between the

five-headed and three-headed Śiva’s sculptures, pointing out that the former can be called Sadāśiva,

while the latter Maheśamurti. He classifies the Indian Maheśamurti’s sculptures in three categories -

seated, standing and the bust - all examples attributing to a specific Indian area. The characteristic

of the Maheśamurti images is that they possess three distinct heads, each one typified by a specific

symbolism. Sadyojāta represents the meditative aspect of Śiva, and together with Tapuruṣa is
30

always portrayed with a meditating expression, Vāmadeva symbolizes the concept of creation and

acquires a feminine appearance, and is sometimes adorned with jewels, while Aghora represents the

concept of destruction and is usually depicted in the form of a Yakṣa, with a ferocious aspect

(Kumar 1975: 106). The disposition of the faces change according to the sect, or varies depending

on the ritual. It could be either Tatpuruṣa (centre), Vāmadeva (left) and Aghora (right), or Sadyojāta

(centre), Aghora (left) and Vāmadeva (right). Identifying the central face is quite difficult due to

the lack of specific features but it can be recognized thanks to the disposition of the lateral heads.

The figure can have either four or six arms holding objects such as a skull, a snake, a pomegranate,

a rosary, a lotus and a mirror (Pereira 1968: 85). Maheśamurti images symbolize the union between

Prakṛti and Puruṣa, namely the sources of Creation. It can be attested that the “Maheśamurti image

is a Śivaite version of the Brahmanical Trinity” indicating the three rasas: peace, love and terror

(Kumar 1975: 106). It is also interesting to point out that Indian Maheśamurti images were never

found in paintings but only represented in sculptures or high reliefs, unlike in the Central Asian and

Chinese examples. These statues mainly came from the north of India, underlining the possible

influence they had on Central Asian iconography. However, Central Asian and Chinese

representations of Maheśvara were only depicted seating, contrary to the Indian examples, which

were mainly represented standing.

31

Plate 28 Plate 29

According to Collins, the first examples of multi-headed Śiva were found during the Kuṣāṇa

Dynasty on the coins of Vima Kadphises (90-100 A.D.), Kaniṣka (127-151 A.D.) and Huviṣka

(140-180 A.D. ca.) (Collins 1988: 112-113) [Plates 28-29]. However, at the same time Kuṣāṇa

coins depicted Zoroastrian, Hindu, Greek and Buddhist deities, reflecting a multi-religious society

(Samad 2010: 32). Perhaps these divinities were used “as a vehicle of royal propaganda, a direct

expression of the ideology of the ruling house” (Collins 1988: 114). For these reasons, it is very

hard to attribute an identity to the god represented on the Kuṣāṇa coins, designated as “OHϷO”

(pronounced Wēš) from Kaṇiṣka onwards. He is usually depicted three-headed with a jaṭā-mukuta

(typical Indian headdress), with four arms holding a triṣūla, a vajra, a water vase and a club, and

sometimes he is ithyphallic and stands in front of a bull. Some of the attributes came from an Indian

background, such as the water vessel, and the multiplied heads and arms; but others were clearly of

a western origin such as the club, the trident and the diadem (Samad 2010: 45).

32

Such a syncretistic divinity was not only found

on coins but also on a terracotta panel

discovered in Bactria together with other tablets

belonging approximately to 150-350 A.D.

(Carter 1995: 585). These represent rare

examples of Kuṣāṇa paintings, and all illustrate

the same scene, namely a worshipper standing in

front of a deity. In this case unfortunately the

panel is broken and just the right part – depicting

the divinity – is left [Plate 30]. The god is

recognizable as Wēš due to its similarity with

the god found on Vima Kadphsises coins. He

has three heads - surrounded by a nimbus - and Plate 30

four arms. However, just one attribute is visible

on the upper right hand, the trident. The central head is bigger, with a third horizontal eye on the

forehead and a moustache, while the lateral faces are smaller and probably represent a child on the

right and a young man on the left (Carter 1995: 581). A similar representation is found on another

panel depicting a three-headed deity with a trident, a water vessel and a drape leaning on the left

forearm (Metropolitan Museum of Art, No. 2000.42.3). However, among the lateral faces just the

left one is visible, showing an old, bearded man. According to Carter, both terracotta tablets

represent the “variant of a single deity” (1995: 581), underlining the syncretistic nature of Wēš. The

faces’ differentiation is quite hard to notice on the Kuṣāṇa coins, which can be due to the fact that

these are nowadays quite damaged, or simply that there was not a will to distinguish the heads by

the side of the Kuṣāṇas. However, Carter points out that in the Huviṣka coin [plate27], the lateral

faces are different, one has a tuft of hair on the head while the other does not (Carter1995: 581).

The terracotta panel most likely had a votive role, as it bore four holes on the corners probably, used
33

to fasten it to the walls of a shrine, which is reminiscent of the Khotan wooden tablets. Moreover, it

also demonstrates that Wēš was a strongly worshipped god during the Kuṣāṇa Empire.

Wēš’ identity is controversial among scholars. On one hand there were those who thought

that Wēš was the Iranian wind god Vāyu (Humbach 1975, Grenet 1994), while on the other hand,

there were those who thought it was the Hindu god

Śiva (Bernard 1981, Cribb 1997, Tandon 2011).

There is not a clear answer, but we agree with

Samad, who points out that Wēš is neither Vāyu

nor Śiva, but a composite image, a result of the

syncretistic religious background typical of the

Kuṣāṇa Empire. However, even if the

identification of Wēš with Śiva cannot be proved,

it is certain that the former figure was essential for

the development of early anthropomorphic Śiva’s Plate 31

iconography in the Northern Indian subcontinent

and Gandhāra (Samad 2010: 47-48). In fact, several scholars underlined the influence of Kuṣāṇa

coins on Maheśamurti images (Pal 1973/1974: 39), as they even saw them as their predecessors

(Collins 1988: 116). The first example of Maheśamurti figure comes from Akhun Ḍhērī, Chārsada

(Gandhāra) and belongs to the 3rd century A.D. [Plate 31]; it is comparable to the depiction of Śiva-

Wēš on the Vāsudeva’s coins (Coomaraswamy 1972: 55). Here, Śiva is represented standing next to

Nandi, with three heads, three eyes and six arms, holding the ḍamaru, the triśūla and the

kamaṇḍalu (the water pot). The left face is bearded and probably corresponds to the ferocious

Aghora, while the central face to Sadyojāta and the right face to the feminine Vāmadeva (Collins

1988: 113).

34

The three statues representing the

couple of Siva and Parvati, belonging to

around the 4-6th centuries A.D., are important

in explaining the development of the

Maheśvara iconography in China and

Xinjiang. The first one was found in

Raṅgamahal (Rajasthan), a clear example of

the “late Kuṣāṇa or early Gupta terracotta

plaque” belonging to the 4th century A.D.

(Agrawala 1956: 61) [Plate 32]. Here, Śiva is

ithyphallic, portrayed seating on his bull


Plate 32
Nandi in the meditation position; he has

three heads and two hands. With the left

hand, he holds the water vessel, while the right is placed next to his chest. On the right side there is

a worshipping devotee and on his left sits Parvati. According to Kumar, his middle head with the

third eye shows his Tatpuruṣa aspect; therefore, the right head should show the Aghora aspect and

the left the Vāmadeva (1975: 109). However, according to Williams the faces are not differentiated,

explaining in this way why the Yungang examples do not bear distinguished expressions, unlike the

Dunhuang, Khotan and Kizil images that mirrored the later Maheśamurti sculptures (1973: 144).

The most important characteristic of this sculpture is the fact that it precedes the later statues of

Śiva holding the sun and the moon with the upper hands. In this case, however, it is not Śiva that

holds these attributes but a gana, an attendant, appearing from Śiva’s headdress and surmounted by

Gaṅgā “descending from heaven at the request of king Bhagīratha” (Kumar 1975: 109). The sun

and the moon in this sculpture were confused by Barrett with a cakra and a kapāla, a skull cup

(Barrett 1957: 59). Through this analysis, it seems that the sun and moon as emblems of Śiva-

which later on became constant attributes of Maheśvara in China and Xinjiang - originated in the
35

Indian subcontinent and were not elements added by Chinese and Central Asian Buddhist

craftsmen. In fact, according to Agrawala, the adoption of celestial attributes as Śiva’s symbols

belongs to the Musanagar Kuṣāna tradition (1987: 18). Additionally, the Nīlamatapurāṇa, a 6th – 8th

century Kaśmīri text, explains the Indian provenance of these elements, as it confers to Śiva the sun

and moon as attributes. When the world was dived into darkness by the demon Jalodbhava, Śiva

took the sun and the moon with his two hands in order to help Vishnu to find and annihilate the

devil spirit (Pal 1973/1974: 42; Banerjee-Agrawala 1970: 222). However, the representation of

figures holding the sun and moon with the upper hands probably originated in the Iranian-Central

Asian context where these elements were connected to regality, they had the “power of investing

kingship with authority”, and were later on adopted by India (Klimkeit 1983: 11). In fact, in the

Kuṣāṇa epoch the sun and the moon were represented on the helmets of princes and kings (Klimkeit

1983: 15, Chandra 2009b: 168). These elements were also used by the Hepthalites as it is

demonstrated by a seal, belonging to the 6th century A.D. and portraying a composite image of Śiva,

Viṣṇu and Sūrya holding the sun and moon (Banerjee 1970: 282) [Plate 33].

Plate 33

36

Plate 34 Plate 35

A second sculpture representing Śiva in the form of Maheśamurti with his consort was

found at Bāramūla (Kashmir) and belongs to the 6th century A.D. [Plate 34]. The deity is ithyphallic

and represented with three heads, three eyes, and six arms of which two were destroyed (Pal

1973/1974: 39). With his left upper arm he holds the sun/moon, while the middle hand is placed on

his wife and the lower holds a water pot. The only remaining right hand holds a rosary, the two

others probably used to hold a triṣula and the sun/moon (Pal 1973/1974: 39). The most interesting

characteristic of this sculpture is the way the heads are depicted, which make it look much older

than the Raṅgamahal statue. These are sculpted frontally and not in profile like the other

Maheśamurti examples and the female head (Vāmadeva) is on the left while the angry-bearded man

(Aghora) is on the right. The way the hair and the feature of the faces are represented is highly

reminiscent of the Gandharān examples (Pal 1973/1974: 40-41). There is another sculpture from

Kashmir which ichnographically is comparable with the Śiva from Bāramūla [Plate 35]. According

to Lo Muzio, the statue probably belongs to the 6th century A.D. rather than to the late 7th and early

8th century A.D. as Barrett pointed out (Lo Muzio2002: 26). The deity is represented three-headed

37

with the Tapuruṣa aspect in the centre, while the female and male aspects are placed respectively on

the left and right. He has six arms, the right ones holding a sun/moon, a trident and a lotus flower,

while the left ones supporting a sun/moon, a rosary and a club. Śiva is ithyphallic and sits on Nandi

together with his consort Pārvatī (Barrett 1957: 59).

Plate 36

Unlike the Śiva sculptures analyzed above, which belong to a Hindu context, a very

interesting and unique statue representing Śiva-Maheśamurti was found in Fattehgargh (near

Bāramūla, Kashmir) and was thought to belong to a Buddhist background due to a little Buddha

placed in its headdress [Plate 36]. The schist sculpture was life-size but is nowadays broken from

below the knees. It represents a case of addorsed images as the Bāramūla sculpture, with a three-

headed figure on the front identifiable as Śiva, and a demon on the reverse identifiable as Mahākāla.

Śiva has three heads as the usual Kashmiri disposition, with the Tatpuruṣa in the middle, and the

female and ferocious aspects on the left and right sides (Granoff 1979: 64). The deity is depicted

standing next to the bull Nandi, carved on the right, and appearing from behind. It is clearly

comparable to the contemporary example from Bāramūla. Śiva is adorned with a necklace,

38

bracelets, earrings and a tiger dhotī, leaving the upper torso naked. The deity was four-armed, but

nowadays just the right arm remains intact and holds a rosary, while another one should have held a

club (Granoff 1979: 64-65). Moreover, both the three faces have a third eye on the forehead; the

central one has a meditative appearance, large earrings and a high jaṭāmukuṭa. The lateral heads are

smaller; the left face has a gentle smile and a typical feminine headdress, while the right has a

ferocious aspect underlined by both the open mouth and the burning hair crowned by a skull

(Granoff 1979: 64). This figure stylistically resembles both the 6th century sculptures from

Gandhāra and Kashmir (Granoff 1979: 76). The most important element of this statue is the seated

Buddha placed in its jaṭāmukuṭa, which attribute a Buddhist identity to this Śiva. This unique

iconography is probably the result of the religious syncretism typical of Kashmir (Klimburg-Salter

1982: 87). In fact, Fattehgargh was situated between Bāramūla, a Brahmanical city dedicated to

Śiva worship, and Huṣkapūra a very famous Buddhist town (Granoff 1979: 77). Moreover,

Buddhism in Kashmir and North India started to be subjected to transformations due to the

influence of Tantric Hinduism from the 4th century A.D. onwards (Hambis 1977: 100).

The sculptures analyzed until now are essential in order to understand the evolution of the

Maheśamurti’s iconography and its influence outside India. It is clear how the Kuṣaṇa coins and

paintings influenced earlier Gandhāran and Northern Indian Śiva iconic art. The first statues

representing Śiva under the form of Maheśamurti were coined in the North of the Indian

Subcontinent, namely in Rajāsthan, Gandhāra, and Kaśmir regions from where they must have

reached Central Asia and eventually China. In fact, several legends narrate the connection between

Kashmir and Khotan (Klimburg-Salter 1982: 89). The similarity between the Maheśamurti figures

and the Maheśvara images found in Xinjiang and China is evident, suggesting that there was a

“prescribed iconographic formula” in Hinduism that Buddhist art adopted in order to depict a new

indigenous divinity (Granoff 1979: 79). Moreover, the Fattehgargh statue represents the first Indian

example of Maheśvara’s adoption into the Buddhist Pantheon, testified by the insertion of a Buddha

into Śiva’s headdress.


39

2.2 Wešparkar: Maheśvara’s Sogdian counterpart

The three headed Śiva iconography seem to have also influenced the depiction of the Sogdian

divinity Wešparkar. In fact, Sogdians adopted the Hindu deities’ iconography in order to depict

their own gods (Grenet 1994: 44; Baumer 2009: 176; Zhang 2009: 36-37). Apparently Sogdiana

encountered Hinduism and Buddhism in two events, both during the Kidarite dynasty (5th century

A.D.) and the Hephtalite Empire (6th century A.D.) (Baumer 2009: 176). Two texts sustain this

hypothesis: a Sogdian translation of the Vessantara Jātaka and another Buddhist text (P 8, 41-42).

The former describes Zurvān, Adhvagh and Wēšparkar in the form of the Hindu deities Brahmā,

Indra and Śiva, attributing three heads to Wēšparkar. While the latter gives the three deities a

double name: Brahmā-Zurvān, Indra-Adhvagh and Mahādeva-Wēšparkar (Grenet 1994: 44).

According to Grenet, the assimilation of Śiva with the Iranian god of the atmosphere, Vayu,

started during the Kushan epoch. This is possible just if it is assumed that Wēš, the god depicted on

Kushan coins, is in reality the divinity Vayu that later on will be worshipped in Sogdiana as

Wešparkar (1994: 43). Three representations of Wešparkar, namely “the God of the Winds”

Plate 37 Plate 38

40

(Baumer 2009: 176), were found in Sogdiana, all dating to the 8th century ca. The most important

one was recovered in room 22 of the Living Quarter I (Pendjikent), where the divinity is identified

by a name written on the right leg (Grenet 1994: 44) [Plate 38]. The deity, like his Hindu model is

three-headed: the right face is feminine, blowing on a horn, the central face has the aspect of a

young man, while the left face looks like a ferocious lion. As such, demonstrating the will of

Sogdian artists to adjust the Indian iconography of Śiva to the Iranian concept of Vayu/Wešparkar

(Grenet 1994: 44), who was also characterized, like the Hindu divinity, by a double nature both

creative and destructive (Carter 1995: 579; Baumer 2009: 176). All three heads bear a third eye on

the forehead and are surrounded by a halo. Moreover, the deity is represented with an armour and

holds a trident in the right hand (Antonini 2005: 313). The other two depictions of Wešparkar

portray him with a bow and an arrow riding a horse [Plate 37], which allows Mode to identify the

deity on the right of the Khotanese wooden panel D X.3 as Wešparkar [Plate 39].

Plate 39

Due to the analogous iconography between the Hindu and Sogdian deities, some scholars identified

Maheśvara images uncovered in Khotan as Wešparkar. According to Mode, there are two different

representations of Śiva, both in Khotan and Sogdiana, one of “Indian” and the other of “Iranian”

origin (1991/1992: 187). In fact, the deity depicted on the right of the votive panel DX.3 clearly

41

differs from the other portraits of Maheśvara that we have already analyzed (D 13, D 7.6 etc.). He

therefore hypothesize that Śiva had a double iconography; he could appear under both the traits of

Wešparkar and Maheśvara. Mode and Baumer conclude that the votive tablets reflected an Indian

iconography absorbed by tantric Buddhism and were eventually used to visualize Sogdian divinities

(Baumer 2009: 181).

It is interesting to underline that also Zhang and Trombert hypothesized a Sogdian

background for both the cave 285 at Dunhuang and cave 8 at Yungang. Zhang points out that the

equation of Maheśvara with the Sogdian god Wešparkar was also attested by a text belonging to the

8th century called Tongdian (2009: 36). However, he focuses on the fact that the Maheśvara image

discovered in cave 285 had a syncretic origin. In fact, it not only was grounded on a Hindu

iconographic model but also on a Central Asian one. This is testified by the presence of the wind

god figure in Śiva’s headdress, visibly resembling the Kushan coins’ depictions of the god OADO

(Zhang 2009: 38). This syncretic image could only be depicted by an artisan with a Zoroastrian,

Hindu and Buddhist background; according to Zhang a Sogdian. Moreover, Kārttikeya seems to

hold in his right hand a bunch of grapes, just as the image of Maheśvara found in cave 8 at

Yungang, which according to Trombert, is the most ancient illustration of grapes in Western China,

and probably is the only contribution of Sogdian artisans to the art of Northern Wei (Trombert

2005: 271). Therefore, this is an additional proof of the probable Sogdian presence in these

Buddhist sites.

42

Chapter 3

3.1 The process of assimilation of Hindu deities into Buddhism

The development and formation of Buddhist iconography in China occurred throughout the 5th and

6th centuries A.D. (Compareti 2008: 19). It is exactly during this period that the Hindu deities

started to be assimilated into Buddhism, and it is during this epoch that the first representations of

Maheśvara were carved at Yungang. The initial figures reflected this period of formation and

evolution through an iconography, which was not yet established. Even if there is uniformity in the

early Maheśvara images, they change slightly from one cave to another. The process of assimilation

of Hindu deities inside the Buddhist pantheon was part of a wider process of the formation of

Esoteric Buddhist art (Sørensen 2011a: 115). However, it must be noted that there are no

recognizable Esoteric images in China before the first half of the 7th century A.D. (Sørensen 2011b:

256).

Hindu deities came to China through Indian Buddhism and were part of a more general

group of spirits with tutelary rules; several texts describe them as divinities protector of the dharma.

The most important is the Tuoluoni zaji (Miscellaneous collection of Dhāraṇīs), written during the

Liang dynasty (502-552 A.D.) (Sørensen 2011a: 115). However, they eventually acquired a more

autonomic place inside the Buddhist pantheon, even obtaining cultic independence (Sørensen

2011b: 257). This phenomenon is observable through the regularity with which the Hindu deities

started to appear in Chinese sites from the Nánběicháo epoch, but also through the fact that they

were represented in larger scale. Moreover, as later demonstrated, the position that the deities

adopted within the grottoes was indicative of the importance they acquired during the 5th and 6th

centuries.

As far as Maheśvara is concerned, he is described by Sørensen as one of the most important

Hindu gods adopted by Buddhism. Several ritual texts were dedicated to him, such as the

Moxishouluo tian fayao (“Essential Methods of the God Maheśvara”) (Sørensen 2011a: 116). His
43

tutelary character with regards to the Buddhist law is emphasized by several scholars (Iyanaga

1983: 733-734; Soymié 1987: 24; Sørensen 1991/1992: 287; Baumer 2009: 178), and acquired a

political meaning both in Yungang and Khotan. In order to support this thesis the deity’s spatial

location at both Yungang and Dunhuang caves will be analyzed.

3.2 Spatial location of the Dharma protectors at Yungang and Dunhuang

In order to understand the role that

Maheśvara had within the Buddhist pantheon,

both the literary and visual sources are required.

In fact, the study of this deity’s images in their

own context corroborate his protective function,

being mostly represented – together with

Kārttikeya - on both sides of the doors and

window reveals as if they were appointed to

control and guard the sacred place [Plate 40]. The

apotropaic function of the doorframe is a

characteristic that probably originated in India

where, since an early period, the dvāra-pālas,

namely the gate-keepers, were placed on the right

and left doorjambs and on the lintel of a temple’s

threshold [Plate 41]. They protected the shrine Plate 40

from the dangers and contaminations coming

from the external world. The act of entering a temple was like an initiation that led to the

purification of the worshipper (Kramrisch 1946: 313-316). As Kramrisch mentioned the “door,

44

niche and window are the architectural symbols of the passage of man to God (…) they are places

of initiation and manifestation” (1946: 322).

Plate 41

Returning to Yungang , the “Dharma protectors” were carved in large dimensions, which

shows that they were more visible to the worshippers and maybe used during the meditation ritual

as visualization aids. The Yungang grottoes were destined to the Chan guan meditation, a practice

based on the mental visualization of the divinities. Therefore, Maheśvara not only had the role of

tutelary deity but also the function of “visual image” (Yi 2010: 283). This is attested by the

examples found in cave 8, where Maheśvara and Karttikeya are depicted on the east and west

reveals of the entrance arch of the main shrine; they are also found in cave 10 where they are placed

on the lintel of the door that lead to the main grotto. Finally, in the later cave 35, the pair of

Maheśvara-Karttikeya is found on both sides of the window reveals placed above the entrance.

However, it is true that Maheśvara’s position within the structure of the grottoes change from cave

to cave. In some cases, as in the grotto 12, he is represented on the ceiling in a smaller scale and

45

seems to acquire more of a decorative role, even though, his collocation in the middle of the ceiling

of the main room, surrounded by dancing devas seems to suggest the artisan’s will to give more

visibility to the deity.

As far as the Dunhuang caves are concerned, there is only one image of Maheśvara on the

left side of the main central Buddha niche of grotto 285. However, its place next to the middle

Buddha, likely suggests a protective role. Moreover, the fact that he is represented together with

Karttikeya and Ganesha his two sons, and two of the four guardian kings, strengthen this hypothesis

(Søresen 1991/1992: 287).

3.3 Some historical and social elements linked to Maheśvara’s representation

Yi, gives a new interpretation of the Yungang caves analyzing the reason why the dharma

protectors were represented in almost every grotto during the second phase (Yi 2010: 52).

Concerning this hypothesis, it is fundamental to explain that Yi places under the same category of

“dharma protectors” all the multi-headed and multi-armed deities. A similar kind of examination is

conducted by Caswell who also creates a separate group for the multi-headed and multi-armed

divinities, underlining their exotic traits and stylistic differences compared to the other gods. He

characterises them as Hindu deities adopted by the Buddhist pantheon, as the multiplication of

limbs has its origins in the Indian iconography. In fact, the depiction of multi-armed and multi-

headed gods was extraneous to the Chinese culture of this epoch, when the artistic representations

were the reflection of the human life (Caswell 1988: 72; Watson 1995: 122). The Buddhas,

bodhisattvas and other deities depicted in the Yungang grottoes were represented in the form of

human beings, without any preternatural characteristics. Among the dharma protectors, the only

recognizable ones, through their attributes are Maheśvara and Karttikeya.

According to Yi, the construction of the Yungang caves was closely connected to politics, as

Northern Buddhism was usually employed by the court in order to express its own ideology.

46

Therefore, Yi sees the figures in the grottoes belonging to the first phase, namely the Tan Yao

caves, as indicating the five Emperors. On the other hand, the twin grottoes of the second phase

represented “the joint rule” of two Emperors: Xiaowen and Dowager Wenming (Knauer 1983: 33).

The same idea is seen in the multi-headed and multi-armed deities which were represented very

frequently during the second phase grottoes, reflecting the necessity to protect the dharma during

this period (Yi 2010: 60). In fact, from 446 to 452 the first persecution of Buddhism by the Emperor

Tài Wǔ, took place throughout which statues and monasteries were destroyed, and monks were

oppressed (Knauer 1983: 29). The “spiritus rector” of the grottoes, as Knauer calls Tan Yao, the

chief of the monks who directed the work of the Yungang caves from 460 to 485, experienced this

persecution (Knauer 1983: 30). He, therefore, transformed his fear into a strong will in protecting

the dharma which is reflected in the Yungang iconography. In fact, the caves were executed by

Emperor Wen Cheng as an act of expiation after his grandfather’s action (Knauer 1983: 30).

Moreover, the iconographic program of the caves relied on Tan Yao’s translation of some sūtras

and on some of its scripts (Knauer 1983: 32). The most important text for the images of Maheśvara

was the Dajiyi shenzhou jing, its completion coincided with the conclusion of the second phase

caves. This text underlined the role of the Dharma protectors in defending the sovereigns and in

releasing the reign from adversities (Yi 2010: 144). Therefore, it can be concluded that it is not so

unusual to find in this historical moment and with such occurrence the depiction of Dharma

protectors.

Khotanese scriptures and visual material also testified the role of Maheśvara as a tutelary

deity. Chandra corroborates this thesis pointing out that the murals and votive panels found at

Dandan-Oilik depicted the deities protectors of Khotan, giving so a double function to the

divinities, both religious and political. In fact, Dandan-Oilik, literally “the houses of ivory”, was

called Li-hsieh in the Han Annals and was part of the six fortified cities (Chandra 2009b: 163). It

had a strategic location and was, therefore, victim of numerous assaults by the Tibetans. Due to the

need to defend its proper city, Dandan-Oilik bestowed symbols of power and protection to the
47

images themselves. In effect, in both the Prophecy of Khotan and the Prophecy of Gośṛṅga the

Buddha is described designating bodhisattvas and Devas as guardians of various cities of Khotan.

The worship of these deities’ icons guaranteed the eternal defence of the kingdom (Chandra 2009b:

164). Maheśvara was part of these tutelary divinities; he protected Khotan from pestilence,

enemies’ attack and general danger against its people. According to literary sources, together with

Vaiśravaṇa, Sañjaya and Hārītī, he gave a dhāraṇī - a speech - in order to defend the prince of

Khotan, Vajrasena, from all dangers (Chandra 2009b: 165). Moreover, it is interesting to point out

that the sun and moon were seen as icons guaranteeing stability and also as the representations of

the Mahāsannipata and the Saddharma-puṇḍarīka sūtras, belonging to the Mahāsannipāta class.

The recitation of these texts was essential for the protection of the region, which is why Maheśvara

symbolically holds them in his upper hands (Chandra 2009b: 164).

48

Conclusion

Despite some differences in the attributes of early Maheśvara images uncovered in the Tarim basin

and Shanxi areas, it is possible to argue that they formed a rather homogeneous group. It is believed

that with Maheśvara, not only we participate to the acquisition of an Hindu deity into the Buddhist

Pantheon, but precisely to the adoption of a particular form of Śiva - Maheśamurti - into Buddhism.

Therefore, through a clear process of appropriation, that culminates in the new role of the protector

bestowed to Maheśvara, the Buddhist iconographers transformed the Hindu deity into one of their

gods.

The wide diffusion of the Paśupata sect, and therefore of the Maheśamurti concept, could

explain the popularity that acquired the Maheśvara image during the 5th – 8th century A.D. in the

Tarim Basin and Shanxi areas. Apparently, traces of this sect were also found outside India, for

example in Cave 198 (Devil Cave) at Kizil, where a portrait of a man was identified with Lakulīśa,

the Paśupata sect’s founder (Ghose 2008: 57). The Maheśamurti sculptures originated in the

Northern regions of the Indian subcontinent, Kaśmir, Rājasthān and Gujarāt, and in Gandhāra,

during the 3rd century A.D. (Collins 1988: 120) and started to acquire popularity from the 6th

century A.D., as it is testified by the large concentration of such figures at Elephanta and Ellora

grottoes. Therefore, it may be that the iconography of Maheśamurti propagated from the Northern

Indian regions in two directions: both to South of India, and to Khotan, finally reaching eastern

China. However, as it was previously highlighted, the first representations of Maheśvara were found

at Yungang (5th century A.D.), followed by Khotan (6th-8th century A.D.), raising difficulty in

tracing a clear geographic linearity in the diffusion of this image. However, the constant presence of

the sun and moon symbols in Chinese representations of Maheśvara testify that the Maheśamurti

concept may have travelled from the Indian subcontinent border to China through Central Asia, as

the celestial attributes originated in the Iranian regions and eventually were adopted by Indian art

49

(Mukherjee 1987: 254; Coomaraswamy 1972: 149; Kumar 1975: 117). Moreover, the adoption of

Śiva iconography by both Buddhists and Sogdians in order to represent their own deities reflect the

flexibility of Hindu faith and the rich cultural syncretism typical of the Silk Road.

50

You might also like