Ērān ud Anērān
Webfestschrift Marshak 2003


The Sun God and the Wind Deity at Kizil

Tianshu Zhu*

Kizil1 is a prominent Buddhist cave site in the ancient Kucha kingdom along the Silk Road. Since it is at the intersection of the east and west cultures, the iconography of this Central Asian site has elements from India, Iran, Rome, and China. Among them, the images of the sun god and the wind deity, in particular, reveal the multiplicity of Central Asian culture. In the Kizil caves, the chariot-riding sun god and the scarf-holding wind goddess images are all depicted on the central strip of the ceilings. The iconography of both of the two deities has a complex origin and can be traced back through the Iranian plateau to the Hellenistic world. Although the sun god image has been intensively covered in the previous studies, this paper will focus on be its Buddhist context. Furthermore, the study on the wind deity trace its origin and try to place the Kizil images of the wind deity within the whole picture of the different major forms of the wind deity found between the east and the west.

The Sun God

The sun god is a fairly stable component in the iconographic program at Kizil cave temples. He is generally depicted dressed in armor, riding in a simplified chariot. The image of a solar deity depicted riding on a frontal chariot appeared all across Eurasia, from the Mediterranean through Iranian plateau to India and Central Asia. This motif has been the subject of innumerable studies for almost a century. Quite a number of such sun god images that are found in India and Central Asia are found in the Buddhist context. However, most of them have survived as fragments out of the original iconographic program. This makes the sun god images at Kizil cave temples very unique and of great significance for they reveal the role of the sun god in the original iconographic context at this site.

To give a simple explanation for all of the sun god images in Buddhist art is impossible. The purpose of this paper is to provide a close up study on the sun god images at Kizil and interpret them in a broader context of Buddhist art and Buddhist texts. For the iconography of Kizil sun god images, the frontal chariot and armor will be the focus of this discussion. The images of the solar deity found from Gandhara and the Central Asian regions often seem ambiguous to scholars as to whether he is the Iranian Mithra or Indian Surya. Therefore, this paper will examine how the name of the sun god was conceived and twisted by the Buddhist in the Buddhist texts.

The sun god images at Kizil

Located south of Tian Shan Mountains (now Xinjiang province, China), the Kizil site consists of 334 caves, near 100 with paintings2. According to Chinese scholar Su Bai's typological and chronological study with reference to carbon 14 testing, these Buddhist temples were active from the 4th to the 7th centuries3.

Several different types of caves serving different functions are found at Kizil, such as the central-pillar caves, square caves, monks quarters and so on. Among them, the central-pillar caves intensively decorated with paintings are the place for worship. A median strip of the images of the sun god, wind deity, Naga, moon god and flaming Buddha is usually painted on the vault ceiling in the main hall of these central-pillar caves. Although the fresco on the vault is partially damaged in most of the caves, Caves 8, 34, 38, 87, 98, 126, and 171 display complete depictions.

The following shows the motifs in the medium strip of these cave. The sequence is from the doorway to the back:

Cave 8: moon god, wind deity, standing Buddha, Garuda, standing Buddha, sun god.

Cave 34: sun god, wind deity, Garuda, standing Buddha, Naga, moon god

Cave 38: sun god, wind deity, standing Buddha, Garuda, wind deity, standing Buddha, moon god

Cave 97: sun god, standing Buddha, standing Buddha, standing Buddha, moon god

Cave 98: sun god, standing Buddha, standing Buddha, moon god

Cave 126: sun god, wind deity, standing Buddha, Naga, moon god

Cave 171: sun god, wind deity, standing Buddha, Garuda, standing Buddha, moon god4.

The central strip shown from the above list is marked by a sun deity and a moon god at the two ends in a very consistent manner. A Garuda (mythical bird) is usually shown in the middle. And one or two wind deities, a rain god (Naga) and a standing Buddha or monk with flames often appear in between. The Chinese archaeologist Ma Shichang saw a piece of a Kizil central strip depicting the 12 constellations and the ecliptic in the Hermitage national museum at St. Petersburg in Russia5. The central strip in Cave 196 depicts a row of flying geese. These variations reinforce that the central median strip was meant to symbolize the sky or a celestial realm.

Fig. 1a is a drawing of the central strip in Cave 171. Fig. 1b shows how the it appears in a cave when one facing the main wall and looking up. Such composition of the central strip on the vaulted ceiling is also common in some other Buddhist cave sites in the Kucha region of the same period such as Kumtula and Simsim and so on. The sun god seems meant to be paired with the moon god and has no independent identity in these Buddhist temples.

At Kizil, the sun god is shown almost invariably wearing crown and armor and seated with the cross ankle on a chariot6. In Caves 7 and 17, the presence of his chariot is indicated by the two horses splayed out at the two sides (Fig. 2). But in most of the caves, his chariot is simplified into only two wheels as shown in Fig. 17. His entire body is framed within his mandorla or the solar disk. His hands may is different positions in different caves. Usually, his left hand rests on his leg and his right hand raises up holding some sort of attribute, which is now completely illegible.

As shown in Fig. 1, the central strip from Cave 171, the moon god image is almost identical with the sun god. The major difference is the color of their radiation. The solar disc is shown in red color, while the moon is white. Both of the sun and the moon discs have radiating lines depicted in them. Only the wheel-typed chariot without depicting the horse can be found in the surviving moon god images8.

The sun god in Buddhism

The sun god is one of the primary divinities to many peoples. The cult of the sun god has a long and strong tradition in both the east and west. What related to our study is Surya and Mithra. Surya is the sun god in the Indian tradition. The textual history of Surya can be traced to the Rg Veda9. Surya is a major god in the Vedas as well as in the later Hindu pantheon. In early Iranian (and European) tradition, Mithra is the god of light and truth, later of the sun, also a judge of people's behavior in life after death. The image of the sun god appearing in the Buddhist context leads scholars to seek links between the Buddha and the sun worship10. And, indeed, fragments of allusions have been found from literature sources. A common epithet of the Buddha, "Adicca-bandhu" in Pali, "Aditya-bandhu" in Sanskrit, which can be translated as "kinsman of the sun" occurred in eight texts of the Pali cannon11. It seems there is a kinship between the Buddha and the sun12. In a Sogdian version of the Vessantara Jataka, "Mithra, the Judge of Creation" is mentioned along with other spirits13. This interesting detail is absent from the Pali version of the same text. Then, is the sun god a judge in the Buddhist world system as suggested by some scholars?14. Another group of examples found by scholars were where the sun is mentioned more like a metaphor or symbol, such as "the golden drum shone like the disc of the sun god"15. These examples seem to lack coherence or the significance as a specific Buddhist identity. Then, what kind of role does the sun god play in Buddhism?

The sun god is mentioned in many Buddhist texts, but usually has nothing to do with important Buddhist doctrines. Mainly, he is introduced as a component in the Buddhist cosmos structure in the Hinayana texts. Occasionally, he is brought up as an ancillary figure in the narrative episodes16. In addition, his iconographic description is preserved in the tantric ritual manuals on various mandalas.

As part of the Buddhist cosmology, the sun god and his palace are briefly described mostly in the following Hinayana sutras: the Da Lou Tan Jing (T 23, vol.1) translated into Chinese between 290-306, the Dirghagama (T 1, vol.1) translated in 414, the Abidharmashastra (T 1644, vol.32) translated in 559, the Qi Shi Jing (T 24, vol.1) translated between 581-600, and the Ekottaragama (T 125, vol.2) and so on. The Dirghagama, for example, describes the sun palace as 51 yojanas long on each side and made of various jewels. (There are various explanations concerning the length of a yojana, one says it is about 7 kilometers17). The sun god radiates a thousand rays of light. The life of the sun god is very pleasurable and his life span is 500 years18. The above quality of the world of the sun god appears to be very ordinary among all of the other heavenly realms.

Similar to the other heavens, human beings can be reborn to the sun god's palace by their merits. Three sutras, the Da Lou Tan Jing, the Qi Shi Jing, and the Dirghagama talk about rebirth in the sun palace. If one makes offerings to the shramanas ("Buddhist monks") and Brahman, and almsgiving to the poor with food, cloth, medicine, elephants, horses, chariots, houses, lamps or candles, and can give whatever is asked with no regret, he will be reborn as the sun god after this life19. The interesting point, here, is that it is not worship of the sun god that leads one to be reborn in the sun palace. On the contrary, the sun worship is cleared prohibited in the Da Lou Tan Jing20 and the Dirghagama21. In the Dirghagama, worshiping the sun and the moon are listed along with other superstitions and incantations to be avoided, such as fire worship. Whoever devotes to such wrong practice will be reborn either in a hell or the animal realm.

From the Buddhist point of view, a Buddha is an Enlightened being, released from the three realms of samsara ("transmigration"). He reveals the truth and guides the sentient beings, including various gods, to real salvation. A Buddha is far more superior to any god. A title of "kinsman of the sun" might glorify the Shakya clan, but it is too small for the Buddha. Regarding the large body of Buddhist literature, this is a rare term associated with the Buddha. Theoretically, there is no independent cult of the sun god in Buddhism. His role, like Indra, Brahman, and other divinities, is as a follower of the Buddha. They came to the Buddha's assembly. After the Buddha's nirvana, entrusted by the Buddha, they became the protectors of the Buddhist Dharma and Buddhist practitioners.

In revealing the Buddhist world system, the sun and the moon are usually introduced together in almost equal footing. The moon palace is smaller than the sun palace of 39 yojanas on each side but is made of various precious materials in a similar fashion. The moon god also radiates lights and lives pleasurably for 500 years. Whoever offers almsgiving to the monks and the poor with cloth, food or other things will ascend to the heaven of the moon after death. However, worshiping the moon is wrong and leads one to a bad rebirth22.

In terms of the visual depiction, the sun god images in early Buddhist art correspond to his position in the Buddhist literature. The sun god images found in the Buddhist sites in India and the Gandhara region are generally small sized partial components of the architecture, such as a pillar or a capital, or as fragments of stone or terracotta relief23. It is impossible to examine if there is a parallel iconographic program of the sun and the moon images. However, their location and size at least indicate that they are not in the position that of having their own shrines or altars to receive worship. The following is an examination of several well-known sun god images to show that they failed to support any significant link between the sun worship and the Buddha.

The most renowned sun god image in Indian early Buddhist art is probably the relief from Bodh Gaya (usually dates to the 1st c. AD). The Surya image appears on the northwest pillar of the railing that surrounds the Mahabodhi Temple. Images of Lakshmi, Shiva, and Indra were also carved on the pillars surrounding the temple. The Surya image does not show any more importance than the other Brahmanical deities on the railing at the site.

Another interesting piece with a Surya image appears on a stone lintel from the Huvishka vihara at Jamalpur, a monastery in Mathura (Fig. 3). On the surviving right half of the lintel, a Surya image is situated on the right end. The rest of the relief depicts the post-enlightenment events of the Buddha. Attempting to claim a close relationship between the Buddha and the sun god, Rowland argued that the Surya image at the beginning of the lintel served as a symbol of the Buddha's nativity24. However, as David Efurd has suggested, there might be a moon god on the other end of the lintel25. A similar composition of the sun god and the moon god exists on a Hindu lintel from Gadhwa dated to the Gupta period26.

Rowland provides additional evidence with a Kushan stone sculpture found at Jamalgarhi in Gandhara. The sculpture shows the sun god in a Bodhisattva form (Fig. 4), which is distinguished from most other Surya images of the same period which are dressed like an Indo-Scythian king. Viewing this Bodhisattva-like Surya as a merging of solar and Buddhist iconography, Rowland interpreted this image as Shakyamuni being equal with the sun god27. Actually, the "Bodhisattva" form might be one type of standard representation of Surya iconography according the textual evidence. For example, in the Fantian Qiyao Jing, or"The Sutra on Brahma and the Seven Luminaries," the form of the sun god is portrayed as should be "like a Bodhisattva based on the Indian version. (He) resides in the radiating sun disc, smiling. He has a halo over his head with five colors. He sits on a lotus which is carried by a chariot driven by five horses"28. The original date of the text is unknown. The section on the iconography of the sun god survived in a later iconographic manual called the Cheng Puti Ji. This sutra emphasizes that why the sun god is in form of a Bodhisattva is based on the original Indian images. It is possible that this Jamalgarhi sculpture represents a tradition in the sun god iconography in which he is shown in a Bodhisattva form.

Most of the rest of the Surya images in early Buddhist art from northwest India and Gandhara are found as decorations on pillars. Their position fails to support the argument of any intimate relationship between the Buddha and the sun god or between solar worship and Buddhism. The fresco of the sun god in the 38m colossal Buddha cave at Bamiyan is situated on the ceiling vault (Fig. 19), exactly like these in the Kizil caves. The vaulted ceiling symbolizes the sky; therefore, it is a proper place for the sun god. Three other sun god images at Bamiyan have also been identified in Caves M, K and J29. They are all paired with the lunar deity. In Cave M, the sun and the moon god are on either side above a seated Buddha (Fig. 5a). The sun god winged with helmet and trident is in a chariot (?) driven by horses, while the chariot of the moon god is pulled by birds30. In Caves K and J, the two gods flank the top of a nirvana scene. Although the painting Cave K (Fig. 5b) is partially damaged, we can at least tell that the sun god is in similar chariot and the moon god is in a chariot driven by birds.

What appeared in the three small caves at Bamiyan, in terms of the composition and the vehicles of the solar and lunar deities, represents a standard convention in Buddhist art of Central and East Asia. The depiction of a sun and a moon god at either side of the main figure made its appearance at Dunhuang in China, especially to later Tantra related figures. One such early image is in the Dunhuang Cave 285 (538-539AD)(Fig. 6). A horse-chariot riding sun god and a bird(?)-chariot riding moon god frame the top of the shrines on either side. They strongly recall the Bamiyan painting seen in Fig. 5. In standard Tantric iconography, such as in womb-mandala of Vairocana, the sun god is in the chariot driven by horses and the moon god is in a chariot pulled by geese. Fig. 7 shows a 9th century silk painting of Thousand-armed Avalokiteshvara the found from the library cave (Cave 17) at Dunhuang. A labeled sun god and a moon god enclosed in a red and a white disc respectively were painted on either side of the top of the painting31.

On one hand, the sun and the moon god function as protective deities, while on the other hand, the presence of the sun and the moon together in visual depiction reinforces the cosmological quality of the central figure. This formula, actually, is an old pictorial tradition ubiquitous in the ancient world in both the east and west. In China, the symbolic depiction of the cosmological sun and the moon images, instead of an anthropomorphic sun god and moon god, were used to flank either side of the tops paintings. This was already common during the Han dynasty (202 BC-220 AD) and later became common in Tibetan Tantric paintings. Fig. 8 is a Han dynasty painted banner from Marquis Dai's tomb. The iconographies of the sun indicated by a bird and the moon by a toad in the painting are from the Chinese mythology. The symbols of the sun (star) and the moon (crescent), or anthropomorphic forms of the sun (male bust) and the moon (female bust) gods are commonly in represented accompanying Mithra from Rome to Near East32. In addition, in early Christian art, the sun and the moon can also be found flanking Christ, thereby defining his cosmological quality and distinguishing him from a historical figure33.

This formulary is taken one step further in the iconography of the cosmological Buddha. The term "cosmological Buddha" describes a type of Buddha image where the whole Buddhist world system is depicted on his body. It emphasizes the notion of the Buddha and the totality of all existence being identical. On Cosmological Buddha images, the sun and the moon are usually shown on either side of the Buddha's shoulder. The well-known Cosmological Buddha from Khotan illustrates this idea (Fig. 9)34. The wheel-like sun and crescent moon forms with the cosmological Buddha also appear at Kizil, in Cave 38.

The central strip at Kizil is unique and has not seen in other regions. However, with the whole picture of how the sun god and the moon god have been located and portrayed in Buddhist art, it is not surprising to see the sun god pairing with the moon god demarcates the sky, creating a sense of Buddhist world structure to the cave. The scholars found that there seemed to be a long tradition of the sun and the moon worship in Central Asia. The Huns (Xiongnu) rulers and the Manichean kings of Ulghur ascribed to the sun and the moon gods the power investing kingship with authority35. This may explain the prevalence of the sun god and the moon god images in the region. Just as that in India, the presence of sun god images in early Buddhist art may not go beyond the assimilation of a popular Hindu deity into Buddhism. In Central Asian, the sun and the moon gods had been converted to Buddhism from the independent Local deities36.

The textural source for the sun god image at Kizil

Three elements are particularly significant in the iconography of the sun god and the moon god at Kizil. These three are riding in a chariot, the seated pose with crossed legs and the armor. The Hinayana texts mentioned above present the radiating quality of the sun and the moon gods and the splendor of their palaces, but do not provide the iconography of the two divinities. The Buddhist texts that describe the iconography for the sun and the moon gods are mostly tantric ritual manuals that were translated into Chinese during the Tang dynasty (618-907) which are later than the date of the Kizil caves by Su Bai's chronology. The sun god and the moon god were assimilated into Buddhism as protectors in the Buddhist pantheon. They appear in the outer ring in the Tantric mandalas. Several versions describing the images of the two gods can be found in different texts. The most common version of the sun and the moon god describes them in the forms of a Bodhisattva in a chariot. The textual description which is closet to the non-Bodhisattva typed image in the Kizil paintings is probably the one introduced in the Jialouluo Ji Zhutian Miyan Jing, "The Secret Text on Garuda and Other Celestial Beings". This Buddhist text was translated by a Kashmir monk (Banruoli, Prajnabhava? Meaning "the strength/power of wisdom"), who received at the Tang court in 758 AD37. The following is its instruction on how to paint the sun and moon god:

"The sun god is in form of a Heavenly King (lokapala), wearing armor, sitting in the gold chariot with legs crossed. (His chariot) is driven by four horses, two heading left and two heading right. He has black hair and is wearing a jeweled crown. He has both a halo and mandorla and is enclosed within the sun disc. The sun disc is red with a pattern like a wheel. The appearance of the moon god looks like the sun god -the disc surrounding the moon god is yellow with a wheel-pattern in it. And a crescent moon raised up from the two sides of the chariot surrounds it from the outside. The moon is pale blue. There are four geese flying around the chariots"38.

The images at Kizil agree with the text in many details. Besides the chariots, their major unique characteristics can all find their source from the above description. This includes the armor, the cross-legged posture, as well as the resemblance between the sun and the moon gods and the wheel-pattern in the sun and moon discs. This text also particularly mentions the geese flying around the chariot. This might explain the geese images frequently shown with the sun and the moon images in the central strip at Kizil. The flying geese are absent in the sun god images from other regions.

The Jialouluo Ji Zhutian Miyan Jing is a text on Garuda. The text defines Garuda as "the lord of the Three-Realms and god of the Five-Elements"39. In the central strip of the Kizil caves, Garuda is usually depicted in the center. It is possible that the resemblance between the text and the images at Kizil is more than just coincidence. However, it must be noted that the description of the wind god and water god in this text does not compare favorably with the iconography in the Kizil images.

Although Jialouluo Ji Zhutian Miyan Jing is not the immediate source that is responsible for the whole central strip of the Kizil caves, the close resemblance between the description of the sun god and the moon god images at the site demonstrates a possible existence of a Buddhist literary source for such a depiction. Nevertheless, some of the basic attributes found on the sun god have long been associated with the sun god. For example, the chariot of the sun god can be traced back to the Rg Veda40. In addition, both the Vedas and Mahabharata describe Surya wearing armor. Armor-wearing was not new to early Indian concepts either41. It is possible that the written documents could have come after the visual representation. The sun and the moon god images at Kizil may had a textual source that were very close to the Jialouluo Ji Zhutian Miyan Jing, or the Jialouluo Ji Zhutian Miyan Jing may reflect a certain visual tradition just like those depicted in Kizil.

The iconographic study of the sun god images of Kizil

The frontal chariot

In the tradition of visual representation of the sun god images, the first element that has to be confronted is, of course, the very distinctive frontal chariot. The chariot of the sun god has been intensively studied. Examined by Bernard Goldman, the pictorial convention to depict a frontal chariot with the team of horses and the wheels splayed out in profile in opposite directions, seems to have originated in Western Asia probably within the Iranian context not long before the beginning of the present era. This device became prevalent in Roman art by the 3rd century and is rooted firmly in the Hellenistic tradition. From whence it spread into Europe and was brought back into Sasanid Asia. It was also carried to India by the Saka invaders42.

The cult and the iconography of the sun god in Northern India is linked to Iranian influence. According to the Samba-Purana and Bhavishya-Purana, a group of Iranians of the Maga caste of Shakadvipa migrated to India in order to devote themselves as Brahmans to worship the anthropomorphic sun god who in the Indian sources is named Mihira43. In India, the first representations of the sun god residing in the frontal chariot in Buddhist context can be traced back to the relief at Bodh Gaya. This format, then, became the canonical iconography of the Indian Surya. The origin has been considered by scholars to be derived from Western models probably due to Parthian influence44. A standard image of the Indian Surya can be characterized by the chariot or the horses, the "Kushan" boots, (holding) the lotus, and the halo45.

Regarding the frontal riding chariot, the sun god images at Kizil are in the same type as the Indian Surya. However, with regard to all of the other details, they do not resemble each other at all. The features of the Indian Surya images show their strong connection with Kushan royal portraits, such as the costume and the boots. These are all absent from the Kizil paintings. Therefore, instead of being traced back to the Indian tradition, the source of the image of the sun god at Kizil is usually understood as a direct Iranian influence, and eventually going back to Hellenistic art46.

Mario Bussagli has noticed that the horses of the sun god's chariot depicted in Buddhist caves at Kizil, Kumtura and Simsim only show the fore body or some images only show two wheels with the horses absent. These features are different from the Indian and the divine Roman chariot images. He linked these unique characteristics of the Central Asian sun god images to the Sassanian tradition47. Apparently, such iconography had a much broader background in Central Asia, as seen in Central Asian textiles; Sogdian frescos, and Dunhuang Buddhist cave paintings. Several pieces of textiles of three types with chariot-riding images for the sun god were excavated from Dulan in Qinghai province and Astana in Xinjiang Uighur Autonomous Province (China)48. As seen in fig, 10, the horses are all depicted with just the fore body. In two of the textiles, the central divinity is clearly sitting in the European fashion just like the Kizil sun god. The posture on the other one is unclear. The sun gods do not wear armor, but are dressed like Bodhisattvas. As discussed above, Bodhisattva-like images are one form of representation for the sun god as in the Fantian Qiyao Jing. Two of the three sun god images are stepping on a lotus, an iconographic detail of the Bodhisattva-type of sun god described in the same sutra.

Several pieces of Sogdian frescos depict the chariot-riding deity as well49. These include the 8th century wall painting from Penjikent (Fig. 10) and an 8-9th century fresco from Qal'a-ye kakhkakha at Shakhrestan in Ustrusana (Fig. 11). They both reveal some similar characteristics as the textile found in China. For example, the horses standing on either side only expose the fore body. The central deities also sit with their legs crossed. The Dulan textiles were dated to the 8-9th centuries50. The Dulan textile seems to be of the same period as the Sogdian painting. They are later than the Kizil paintings. The new chronology of the Kizil caves is 4-7th centuries. The Dunhuang sun god image in Cave 285 (Fig. 6) seems to be the farthest eastern example of this type of chariot. It dates to 538-9 by the inscription, which is contemporary to the Kizil caves and is understood as influenced by the Kizil images by some scholars51. However, regarding the position of the sun and the moon gods in the composition of the cave and the chariot of the moon god, the Bamiyan images in the three small caves seem closer to the Dunhuang depiction.

In short, riding on frontal chariot is one of the typical attributes associated with the solar deity. The chariot of the sun god at Kizil belongs within the whole picture of the long and complex tradition for the depiction of the frontal chariot that came from Hellenistic-Iranian influence. The chariot in the Kizil images with only the fore body of the horse shown may have its root in Central Asia as a regional variation.

The armor

The other significant aspect in the iconography of the Kizil sun god is his armor. As indicated in The Secret Text on Garuda and Other Celestial Beings, the sun god should wear armor like the Heavenly King. The four Heavenly Kings were also depicted at Kizil. They are usually shown in the nirvana scene behind the reclining Buddha. It is no surprise that they are wearing the same type of armor as the sun and the moon gods. Unfortunately, most of their bodies are blocked by the Buddha. However, the same type of armor is found on other figures in Kizil paintings such as on Vajrapani in the "Preaching Scene," on Mara in the scene of "The Defeating Mara," on soldiers in the scene of "Dividing the Buddha's Relics (Fig. 13)," as well as on other guardian figures at the Kizil site.

This type of armor has been generally identified as mingguang jia, "bright-brilliant armor", a type of armor recorded in the Chinese textural sources52. This kind of mingguang jia armor is characterized by the two plaques at the chest, which was understood as a representation of the metal reinforcements attached to the body of armor. The term of mingguang jia appeared as early as in the early third century Cao Zhi's petition53. However, from the archaeological evidence found from the tomb figures, the two-plaque armor did not come to be popular until the end of Northern dynasties in the sixth century, and became the major form of armor during the Sui (581-618) and the Tang dynasties54. It has long been recognized that the rapid development of Chinese armor during the fourth to sixth centuries was stimulated by the influence of Iranian prototypes brought by tribesmen into the Chinese central plain55. The armor with the two plaques became the standard costume for the Heavenly Kings56, either as guardians in Buddhist monasteries or in the tombs. In his study of early Chinese armor, Albert Dien finds that the structure of such mingguang jia shown on Chinese tomb figures and stone carvings was not completely clear. And, the mingguang jia on the Heavenly Kings became so elaborate in the Tang dynasty that he began to doubt if it was ever actually worn on a battlefield. Dien further points out that the plaques may not represent visible parts of sheet armor, and there was no evidence that they were used in China57.

If the plaque patterns on the images are not necessarily a real part of the armor, then where did such an iconographic formula come from? In such a confused puzzle, the Kizil images might be helpful to bridge the gap between the conventionalized ornate representation and its counterpart of what could have been as practical armor. Shown in Fig. 2, and 13, the pictorial depictions of the armor on various figures at Kizil all have two scallop shapes at the chest, suggesting shiny metal in one piece. Images in better condition and larger scales reveal some patterning and additional straps crossing the chest plaques. Besides the fresco, some fragments of sculptures also survive from the Kizil site. There was a piece published in Germany in early twentieth century which preserves another view of the structure of this particular form of armor58. Shown in Fig. 14, the armor on the torso is basically the same format of that in the painting with the high collar and the four-part division in the front. However, the two sections on the chest are still lamellas with a large roundel in the middle, rather than one piece of metal. The roundel-like device appears in the painting under the straps (Fig. 13). Interestingly, the breast armor is clearly raised high in shape of the scallop. It is uncertain if they are meant to be supported by the plaques underneath. Fig. 15 is the four heavenly kings painted at Dunhuang Cave 285 of the sixth century, the same cave depicting the sun and the moon gods mentioned above. In this painting, the armor takes the structural format like that seen in the Kizil painting and depicted each lamella like the Kizil sculpture. It is another early image demonstrating that the scallop pattern on the upper armor may not all be meant to be a representation of one sheet of metal. Fig. 16 show the five types of Heavenly King images developed in the Tang dynasty showing a gradual elaboration in chronology. The first type of the early Tang period characterized by the crossing knotted cord closely relates to the armor images found at Kizil, especially on the sculpture. The cord, which looks as if it were on top of the armor, puzzled scholars as to how it would function in relationship with the body of the cuirassier59. In the almost austere depiction of the Kizil armor, the crossing lines are virtually the edge of one piece with the body of the corselet.

The armor shown in the Kizil cave is likely to be a relatively close representation of the real form of armor used in Central Asia. Alexander Strelkoff points out a painted shield showing a mounted Sogdian knight wearing "a coat of mail" similar to that of soldiers in the Buddhist caves at Kizil. The shield was excavated at the Muhg-Qal'a on the Khum River in 193460. The external influence on the development of Chinese armor may have had its impact on China in two ways, the actual armor and its visual representation. Having been copied from copies for centuries, the two may not be exactly identical with each other by the time of the Tang dynasty. When the internal examination within the central plain of China has been exhausted, the Kizil images may be very valuable in reconstructing the missing link.

The identity of the sun god

I have carefully chosen the term of "sun god" for this discussion. However, whether the sun god image at Kizil is Surya or Mithra seems to be a question that cannot be avoided. In concept, Surya and Mithra are both sun gods, but in different religions in different regions and apparently associated with different iconography and sometimes even different media. Solar divinities often have a primary role in religion, each called by a different a name associated with a different mythology. Solar deities have names far more than just Surya and Mithra. For example, he might be called Helios in Greek Bactria, Ohrmazd in Khotan, and Mithra in Iran61, and Surya/Aditya in India. The Sasanian name for the sun god is Mihr, the Middle Persian name is Mithra62.

The Greek Helios rides in a chariot. The standard iconography of Mithra's image is a figure slaying a bull, which has nothing to do with chariot riding. The Mithra images from Iran are sparse. In the meantime, Surya's iconography developed in Northern India depicting the sun god riding in a horse chariot, which is supposed to be influenced from Hellenistic-Iran. Inscribed sun god on Kushan coins is always Miiro/Helios63. The figure of Surya, who was commonly found in stone and terracotta sculpture, never appeared on the Kushan coins64. However, the identification of the sun god is not so simple.

Apparently, many images of chariot riding deities found in Iran and Central Asia have been identified as Mithra, including those images in the Buddhist context. The radiating figures in chariots on several Sasanian seals have been identified as Mithra65. The Sogdian wall painting and the textile discussed above have been identified as Mithra. The well-known sun god standing in the chariot at Bamiyan above the colossal Buddha has also been previously identified as Mithra66. The sun god in a non-Buddhist context could be Mithra. The concern of this paper is the sun deity in Buddhist monasteries. Who is the sun god in the Buddhist temple? Is he a deity of India just as the same source as the Buddhas and Bodhisattvas or is he a local god? It is not only an issue for contemporary scholars, but for the people in ancient time as well. How would a Medieval Buddhist view the sun god?

In searching the various iconography of the sun and the moon gods among the Buddhist texts, I came across an interesting passage in The Sutra Spoken by Bodhisattva Manjushri and the Sages on the Auspicious and Ominous Days and Good and Evil Planets and Lunar Mansions translated by Amogavajra (705-774) in 759. This astrology text served almost as a handbook held by every his disciple and annotated by his Chinese student, Yang Jingfeng, so it would be easy for Chinese to use67. The western week system was introduced in the book. Claimed in the book, in order to enable Chinese to consult with Hu people ("people from Central Asia"), Persian, or Indian, the different names of the seven luminaries used by these foreign peoples68, which the seven days of the week were named after were listed in the chapter II69. The following is the list for the sun and the moon:

"The sun, Taiyang (in Chinese), the Hu name is Mi, or 'Mihr'; Persian name is Yaosenwu, or "yek-sumbad"; Indian name is Adiye, or 'Aditya'.

The moon, Taiyin (in Chinese), the Hu name is Mo, or 'Mah'; the Persian name is Louhuosenwu, or "douh-sumbad"; the Indian name is Sushangmo, or 'Soma'"70.

The same set of the Hu terms were marked into the Chinese almanacs of the Tang dynasty found in Dunhuang. Especially the name for "Sunday," the Day of Mihr has been persistent and prevalent71. Some scholars speculate that the Hu names came from Sogdiana of Samarkand72. The image of the sun god in a chariot seems did go beyond the Tantric Buddhist paintings in later time. However, his name, being associated with Sunday, reached the east coast of the Asian continent and survived in the local dialect of Fujian province at least until the nineteenth century73.

Amogavajra is one of the four most eminent Buddhist text translators in Chinese Buddhist history. His father was a Brahman from north India, and his mother was a Sogdian. He came to China when he was a teenager. What is revealed in the list of names in the above paragraph reflects the awareness of an open mind. Be it Mihr or Aditya, Mithra or Surya, he is the sun god. In essence, they are all the same, as the same as the Chinese Taiyang. In the Chinese translations of Buddhist literature, "ri tian zi" is the word usually chosen for the solar deity. "Ri tian zi" means "the king of the sun" or "the sun god". When he is introduced for the first time, the term Aditya or Surya may be provided, very understandably, not Mithra.

However, the clear ubiquitous concept of the "sun god" avoiding any nationality apparently did not satisfy everyone. The sun god very frequently changed his costume along his journey from India, through Central Asia, to China. That Buddhism can reached to the masses among new groups of people usually takes place when the new culture can somehow identify itself with Buddhism. In China around 518-569AD, there appeared the Xumi Siyu Jing, "Sutra of the Sumeru World in Four Directions", attributes the sun god as Fuxi, the moon god as Nuwa74. Fuxi and Nuwa are the ancestors of Chinese in Chinese mythology. The iconography of Fuxi and Nuwa employed as the sun and the moon gods also appeared in Buddhist art, on the ceiling of Dunhuang Cave 285 (Fig. 17). They are depicted with human heads and serpent bodies dressed in Chinese robe. This is the same cave where the chariot riding sun and moon gods were also painted. The Xumi Siyu Jing, however, did not confuse Chinese Buddhists. It has always been listed no doubtfully into apocryphal sutras75. The depiction of the sun god and the moon god seen in Dunhuang Cave 285 is just a rare case.

There are not enough textual documents surviving from Central Asia to clarify the identity of the sun god at Kizil. Nevertheless, it is probably safe to say that, to the clear mind of a Buddhist, the solar deity was understood as a general sun god, even though his iconography had often been localized. The iconography of various Buddhist figures were very frequently changed into a form that the local people were more familiar with. In this case, the sun god is not alone. From his Kushan costume, to the Central Asian armor, to the Chinese robe of Fuxi, the complexity of sun god images speaks a common phenomenon in the transmission of Buddhism.

The Wind deity

There is another image at Kizil that is closely associated with the sun god in composition and origin besides the moon god, he is the image of the wind deity. The image of the wind deity at this site belongs to a type of scarf-holding wind god. The scarf-holding wind god is derived from Greek-Roman art, and was transmitted to the Iranian plateau, and generated influence to China and even to Japan. The following section will try to study the iconography and origin of the Kizil wind deity, map out its relationship to other forms of wind god in East Asia, and briefly summarize its relevant texts.

The Wind deity at Kizil and its origin

Riding in his chariot, the sun god arrived at Kizil from the art of the classical world. He did not come alone; he brought with him the wind deity. At Kizil, the wind goddess image is frequently shown in the central strip as a celestial being. Fig. 18 is a drawing of a typical representation of the iconography of the wind goddess. Her upper torso merges from clouds, with two hands holding a scarf flowing behind her. She opens her mouth as if to blow wind. Her hair is twisted upward, standing on end. Her gender is clearly indicated by the exposed breasts. The wind goddess is normally depicted twice in the media strip, one in front of the sun god the other in front of the moon god.

This depiction closely resembles the well-known image in the vault ceiling above the colossal Buddha at Bamiyan. Shown in Fig. 19, above the standing sun god, two wind gods were painted at the two top corners, with only half of the torso extended out the clouds, and two arms holding a billowing mantle in the big wind. They look like females. Female wind deities seem only appear in the Bamiyan and Kucha region76.

In the west the taut blowing scarf was the attribute of the wind deities77. Gods or goddesses with wind-blown scarf are the frequently depicted motif in the art of antiquity78. According to Steinmetz, Raff and Neuser, the Greek and Roman wind god images are roughly classified into three types: 79.

  1. Representing the wind god Boreas or Zephyros
  2. Representing the blasting movement of the wind
  3. A cosmological type representing the four or eight cardinal wind

The third type is characterized by only representing the bust of the wind god and is usually associated with the images of Mithra. The Kizil and Bamiyan wind goddess images are shown just with their upper torsos. And the images from both sites are in pairs and are related to the solar deity in a way. Leroy Campbell found out that in the iconography of Mithra reliefs, representations of the two winds are more frequent than all four. In addition, and the wind god images are mainly more a North Iranian motif than a South Iranian80. Studies also show that the wind god was of great importance in northern or northeastern Iran81. Therefore, it is likely that the Kizil type wind goddess image is derived from the third type of Greek-Roman wind god depiction, that is to say, from the bust form of the wind god on Mithra reliefs found in Northern Iran with the background of the worship of the wind god at this region. If we can say that the sun and moon gods at the Kizil caves demarcate the sky, then the original cosmological function of the wind god images seems to properly fit into the iconographic program.

Wind god images in China (and Japan)

The images of the sun god riding in a chariot apparently only appeared in esoteric mandalas and mandala-like Buddhist paintings in later time, whereas image of the wind deity seems to have generated a greater and more complicated impact on East Asia. The traditional wind god in China had a very different iconography. The wind god images first appeared in the Eastern Han dynasty (8-220 AD). Fig 20 shows the various wind god images of the Eastern Han, where the main attribute of the wind god is blowing wind through the mouth. These images were carved in stone found mostly in the Han tombs. The wind deity is represented as a demon type of god in Chinese art. He is portrayed in a full body form, either sitting or standing, and is usually associated with clouds pattern and other natural spirits, especially with the thunder god and the rain god.

By the time of the sixth century, holding a big windbag in his arms became the standardized iconography of the Chinese wind god. Fig 21 shows the various wind god images found either on Buddhist stone steles or in Buddhist cave temples during the sixth century. They were usually depicted on the base with other nature spirits, such as the tree god, thunder god, fire god, and so on82. The type holding windbag is generally thought of as a Chinese indigenous development83. Most of these sixth century wind god images are shown seated and some have their hair standing on end.

In the meantime, the scarf-holding style also appeared in Chinese Buddhist art. This type of image has survived only in a very few number: one is on a four-side stupa, and the other is in the well-known Dunhuang Cave 249. Fig 22 shows the wind god on a votive stupa dated to 501 AD. This piece was found in Chang'an county which is near the modern city Xi'an in the center of China, and is now in the Xi'an Beilin Museum84.Only a section of the original stupa has survived. On one side of the stupa, a wind god and a thunder god are depicted on the lower left and lower right corners. Both of them are in dynamic running pose. The wind god holds a ring-like scarf flowing at his back, while he thunder god plays a ring of linked drums. Pairing the wind god and the thunder god together comes from the Chinese tradition.

Fig 23 is the west ceiling of Dunhuang Cave 249. A thunder god and a wind god are painted on either side of Mount Sumeru, the Buddhist cosmic mountain. Again with their most distinguishing attributes, the thunder god is playing a ring of drums and the wind god is grasping his blowing scarf. Cave 249 is dated to the early Western Wei dynasty (535-556 AD) and later than the stupa in fig 22. Both the style and the iconography of the ceiling paintings in this cave are well known for their Chinese inspiration. Although the scarf is certainly of non-Chinese origin, both of the two scarf-holding wind god images are paired with the thunder god, which is a Chinese convention. It seems that the foreign scarf-holding type of wind god image had already spread to China and became an integrated part of the Chinese style by the sixth century.

The images form the stupa and Dunhuang are different from the bust-type of wind god images of that appeared in Kizil. Particularly the dynamic running pose the image of the two images is interesting. Katsumi Tanabe examined the iconography of the wind god and its transmission from Greece-Roman cult, to Kushan coin, and to East Asian depiction85. Besides holding the scarf, the Greek-Roman wind god images have big wings; however, the Kushan wind god has no wings but he is in a running pose. Most of the Kushan images of running wind gods are found on the reverse sides of Kushan coins (Fig 24). Such depiction may have been derived from the second type of Greece-Roman wind god iconography listed above. This type of wind god image is not intended to represent any particular personified deity, but rather the motion of the wind blasting. As shown in Fig. 24, the up-raised hair, the scarf, and the running posture, all these characteristics of this type are to illustrate speed and movement. According to Katsumi Tanabe, the Kushan wind god images on these coins served as the direct source for the Chinese scarf-holding wind god depiction86.

The wind god images became rare after the sixth century. From the surviving materials, it is difficult to make a conclusion of how long and far the scarf-carrying wind god image had spread. The latest example I can find is a painting of the tenth century (Fig. 25). This painting depicts Shakyamuni defeating Mara, the Lord of Death. At the central top of the painting, the wind god and the thunder god again are painted together. Here they belong to the demons of Mara attaching the Buddha. The wind god, similar to that in Cave 249, is running in the sky in a dynamic pose with a long scarf. He is lifting up a rock (?) in his hand and is about to throw it at Shakyamuni. This painting came from the library cave at Dunhuang.

This Chinese-style wind god went to Japan. Fig. 26 is the wooded sculpture in Sanjusangendo from the thirteenth century87. The wind god grasps a bag of wind slung around his neck. A scarf hangs down over his shoulders. He is with the thunder god who is playing a ring of drums just like the Chinese convention since the sixth century. The notable change is the position of the windbag, which has switched from the front to the back. The curving shape of the windbag closely resembles the blowing scarf in the images of Kizil, Bamiyan, and Dunhuang. The same is true with the way the wind god grasps his attribute. In terms of the visual form, this later type of the wind god is almost a combination of the scarf-holding and the windbag-carrying images, the two types of wind god from two different cultures.

Katsumi Tanabe suggested that this type of wind god iconography is inspired from both the windbag-holding and the scarf-holding images of the sixth century. In Japan, such wind god and thunder god images became a popular motif of the later Rimpa school artists from the seventeenth to the nineteenth centuries. The archaic features such as the bare torso and the running pose kept in the Japanese sculpture and Rimpa school paintings, did not remain in the contemporary Chinese depiction of the wind god. For example, the painting in Fig. 27 is one of a set of Chinese hanging scrolls depicting the complete pantheon for the Buddhist ritual dated to the fifteenth century88. The wind god is in a group with the thunder god, the lightening god and the rain god. All of the four gods seem to have been "civilized." Among them, the wind god is fully dressed, without motion, seemingly to have been cultivated by the Confucians virtue of self-restrain. His windbag is carried on his shoulder. He still bears an animal head, the terrifying monster aspect always associated with his images.

The wind god in texts

In the Buddhist texts, the wind god has no particular significance and seems less important than the sun god. He is either listed as one of four gods of the Four Elements: earth, water, wind, and fire or as a natural spirit in company with the tree god, river god, fire god and so on. There are many wind gods with individual names. He first appears in the context describing the world system and the pantheon of all gods; second, as a walk-on figure in the narrative; third, as attendant in the Buddha's assembly; forth, as protective a deity mostly in the ritual manuals. For example, in the Dirghagama sutra, when the Buddha was explaining our cosmos structure, the wind god is introduced as one of the gods of the Four Elements89. In the Buddhavatamsaka-mahavaipulya-sutra, the wind gods also came to listen to the Buddha's preaching. Eight wind gods with different names were listed along with other nature spirits in this sutra90.

In most of the cases except in the ritual manuals, no iconographic description of the wind god is provided. And almost none of the forms of the wind god in the ritual text appear to be close to the type of his iconography under discussion. In The Secret Text on Garuda and Other Celestial Beings, the wind god has a thousand heads and a thousand arms. He is in a wrathful form, with a serpent as his ornament91.

Interestingly, in the Hindu text Vishnudharmottara, the Indian wind god, Vayu, was described very much like the scarf-holding type of wind god that is found in Central Asia. I don't want to suggest any direct relationship between the two. However, it is simply worth mentioning since they so closely resemble each other. The text says, "He is two-armed, his two hands holding the two ends of the scarf worn by him, his garment being inflated by wind, his mouth being open and his hair disheveled"92. The two-arm, scarf-holding, mouth-opening, and disheveled hair, are all features that match the bust-type images in Kizil and the running-type of wind god in Kushan and China. The above textural description might have been influenced by the Kushan wind god, which might be one of the origin for the Central Asian and the Chinese scarf-holding wind god images93.

To sum up, the images of the Greece-Roman scarf-holding wind god reached Central Asia and China. The cosmological type of bust-like wind god image influenced Bamiyan and Kizil sites in Central Asia. The Kushan running-type of wind god, which derived from the second form of Greece-Roman images, influenced China and lasted at least to the tenth century. In addition, a type of the wind god image found in Japan portrays him grasping the windbag in the same manner as if grasping a scarf and it may have been inspired from both the Chinese and western traditions.

Concluding Remarks

Kizil is located at the northern route in eastern Central Asia. The eastern Central Asian region is now understood mainly occupied by Indo-European speaking Tokharians, including presumably Yuezhi (Yueh-chih) who dominate the northern route and Iranian Shaka in the western part of the southern route94. The Yuezhi was the people who emigrated from northwest of China and formed the Kushan empire in the first to the third centuries AD. Chinese control has reached eastern Central Asia since the Han dynasty; however, these kingdoms of eastern Central Asia enjoyed a more independent status until the Chinese conquest returned to Central Asia in the last Sui (581-617) and early Tang period95. Therefore, it is not surprising that the images of the solar and wind deity from Kizil region have an origin that can be traced back to the art of Iranian world and antiquity. The structure of the caves, the painting style and some of the other iconography of the Kizil frescos also sharply differentiate from the Chinese tradition and bear more resemblance to those in the west.

In the transmission of the iconography of the chariot-riding sun god and the scarf-holding wind god images from the west to the east, the Kizil images are not simply a middle step, but they have their own independent position. The Kizil sun god and wind god and the subsequent Chinese images are different from each other. The Kizil sun god and moon god are identical and dressed in armor, whereas the forms adopted in the China show one on horse-chariot, the other in bird/geese-chariot and most of them are in Bodhisattva form. In addition, the Kizil wind deity is a female and in bust form. The type that influenced China is the male running formed wind god. Ultimately, the origin of the two may be the same, however, the Kizil images do not seem to be the direct source for the Chinese depiction for the sun god and wind god.


Fig. 1a: The central Strip, Kizil Cave 171 (after Ma Shichang, 1996, p.177, Fig. 3).
[Missing Image]
[Click image to enlarge]

Fig. 1b: Kumtula Cave 46 (after Chongguo Shiku, 1992, pl.104).
[Missing Image]
[Click image to enlarge]

Fig. 2: Sun god, Kizil Cave 17(after Ma ShiChang, 1996, p.177, Fig. 4).
[Missing Image]
[Click image to enlarge]

Fig. 3: Lintel with Surya image, Huvishka vihara, Mathura (after Rosenfield, 1967, pl.46).
[Missing Image]
[Click image to enlarge]
Image detail
[Missing Image]
[Click image to enlargeb]

Fig. 4: Surya, Jamalgarhi, (after Banerjea, 1948, p.59).
[Missing Image]
[Click image to enlarge]

Fig. 5a: The sun god and moon god in Cave M, Bamiyan (after K. Maeda and A. Miyaji, 1971, pl.107).
[Missing Image]
[Click image to enlarge]

Fig. 5b: The sun god and moon god in Cave K, Bamiyan (After Higuchi Takayasu,1980, Fig. 20.p.65)
[Missing Image]
[Click image to enlarge]

Fig. 6: The sun and moon gods, Dunhuang Cave 285, (after He Shizhe, 1987, Fig. 4&5)
[Missing Image]
[Click image to enlarge]

Fig. 7: Thousand-armed Avalokiteshvara, painting on silk, 9th c. Dunhuang Cave 17, (after R.Whitfield and A. Farrer, 1990, p.27).
[Missing Image]
[Click image to enlarge]
[Missing Image]

Fig. 8: Painted banner, Marquis Dai's tomb, Changsha, (after Lee, 1993,Fig. 77).
[Missing Image]
[Click image to enlarge]

Fig. 9: Cosmological Buddha, Khotan, (after Bussagli, 1979, p.55).
[Missing Image]
[Click image to enlarge]

Fig. 10: The textiles with the sun god images, Dulan and Astana, (after Zhao Feng, 1995, Fig. 1).
[Missing Image]
[Click image to enlarge]

Fig. 11: Drawing of the Penjikent painting (after Compareti, 2000, Fig. 2).
[Missing Image]
[Click image to enlarge]

Fig. 12: Drawing of the Sakhrestan painting (after Compareti, 2000, Fig. 3).
[Missing Image]
[Click image to enlarge]

Fig. 13: Dividing the Buddha's relics, Kizil, (after Grünwedel, 1912, p.47, Fig. 57).
[Missing Image]
[Click image to enlarge]

Fig. 14: Sculpture, Kizil, (after Alfred von le Coq and Ernst Waldschmidt, 1922-1923, vol.1, pl.28).
[Missing Image]
[Click image to enlarge]

Fig. 15: Heavenly kings, Dunhuang Cave 285, (after He Shizhe, 1987, p.372. Fig. 12-1).
[Missing Image]
[Click image to enlarge]

Fig. 16: Chronology (I-V) of the Tang armor style, (after Yang Hong, 1976, table 3).
[Missing Image]
[Click image to enlarge]

Fig. 17: Fuxi/sun god and Nüwa/moon god, Dunhuang in Cave 285 (after Ji Xianlin, 1998, p.18).
[Missing Image]
[Click image to enlarge]

Fig. 18: Wind goddess, Kizil Cave 38 (After Zhongguo Shiku- Kizil Shiku I, Fig. 113)
[Missing Image]
[Click image to enlarge]

Fig. 19: Bamiyan Sun god and wind goddess image (after T. Maeda and A. Miyaji, 1971, pl.98)
[Missing Image]
[Click image to enlarge]

Fig. 20: Wind god, Han dynasty, (after Gu Sen 1997, p.535)
[Missing Image]
[Click image to enlarge]

Fig. 21:Various windbags held by wind gods (after Jin Shen, 2000, p. 149)
[Missing Image]
[Click image to enlarge]

Fig. 22: Votive stupa (after Li Jian, 2003, Fig. 67, p.146)
[Missing Image]
[Click image to enlarge]

Fig. 23: Ceiling of Dunhuang cave 249 (after Zhang Wenbin, 2000, p.27 p.141)
[Missing Image]
[Click image to enlarge]

Fig. 24: Kushan coins (after Rosenfield, 1967, pl.149,150)
[Missing Image]
[Click image to enlarge]

Fig. 25: Defeating Mara with detail of the wind god and the thunder god (after Zhang Wenbin, 2000, P.141)
[Missing Image]
[Click image to enlarge]
[Missing Image]
[Click image to enlarge]

Fig. 26: Wind god and thunder god, wood, in Sanjusangendo, 13th c. (after Mason,1993, Fig. 186a, & b. p.158.)
[Missing Image]
[Click image to enlarge]

Fig. 27: Wind, thunder, rain, lightening gods (after Shanxi Museum, 1985, Fig. 114.).
[Missing Image]
[Click image to enlarge]


Ph.Ackermann (1938-1939), "Sasanian Seals," A Survey of Persian Art, vol. 1, ed. Arthur Upham Pope, London, pp. 784-815.

G. Azarpay (1981), Sogdian Painting, The Pictorial Epic in Oriental Art, Berkeley.

A. M. Belenizky (1980), Mittelasien, Kunst der Sogen, Leipzig.

Jitendra Nath Banerjea (1948), "Suya: Adityas and the Navagrahas," Journal of the Indian Society of Oriental Art. XVI (1948): 47-100.

Jitendra Nath Banerjea (1956), The Development of Hindu Iconography, New Delhi.

M. Boyce & F. Grenet (1991), A History of Zoroastrianism, vol.III. Zoroastrianism Under Macedonian and Roman Rule, Leiden.

Mario Bussagli (1955), "The 'frontal' representation of the divine chariot," East And West. 6 (1955): 9-25.

Mario Bussagli (1979), Central Asian Painting, New York, 1979 (1st ed. 1963).

Pierfrancesco Callieri (1990), "On the Diffusion of Mithra Images in Sasanian Iran-- New Evidence from a Seal in the British Museum," East And West. Vol. 40 (1990): 79-98.

Leroy Campbell (1968), Mithraic Iconography and Ideology, Leiden, 1968.

M. Carter (1981), "Mithra on the lotus-a study of the imagery of the sun god in Kushano-Sasanian era," Monumentum Georg Morgenstierne, I, Acta Iranica, 21, Leiden, (1981): 74-98.

M. L. Carter (1988), "Revanta, an Indian Cavalier God," Annali dell'Istituto Orientale di Napoli, vol. 48, fasc. 2 (1988): 127-137.

A.Christensen (1944), L'Iran sous les Sassanides. Copenhague, p.144.

G. Combaz (1937), L'Inde et l'orient classique, Paris.

Matteo Compareti (2000), "Iranian divinities in the decoration of some Dullan and Astana Silks," Annali Di Ca' Foscari, Rivista Della Facola Di Lingure Letterature Straniere Dell'universita Di Venezia, Estratto, xxxix, 3, (2000): 331-368.

Ananda Coomaraswamy (1935a), La Sculpture de Bodhgaya, Ars Asiatica XVIII, Paris.

Ananda Coomaraswamy (1935b), Elements of Buddhist Iconography, Cambridge.

Alfred von le Coq and Ernst Waldschmidt (1922-1923), Die Buddhistische Spätantike in Mittelasien, 7 vols, Berlin.

Alfred. von Le Coq(1935), Bilderatlas zur Kunst-und Kulturgeschichte Mittel-Asiens, Berlin.

Th. Raff (1978/79), "Die Ikonographie der mittelalterlichen Windpersonifikationen," Aachener Kunstblatter, 48 (1978/79): 77-218.

T. W. Rhys Davids (1969) trans, Digha-Nikaya, Sacred Books of the Buddhists II, London.

T. W. Rhys Davids and William Stede (1966), eds., The Pali Text Society's Pali-English Dictionary, London.

Albert Dien (1981/1982), "Study of early Chinese armor," Artibus Asiae, v. 43, no 1/2 (1981/1982): 5-66.

Helene Diserens (1997), "Two stone reliefs of Surya from Gum—A study of the sun-chariot and its terms," Silk Road Art and Archaeology: Journal of the Institute of Silk Road Studies, vol.5 (1997.8): 329-351.

David Efurd (1999), Early Iconography of the Indian Sun-god Surya, MA. Thesis, University of Georgia.

A. J. Gail (1978), "Das Sonnenkult im alen Indien--Eigengewachs oder Import?" ZDMG, 128 (1978): 339-41.

Simone Gaulier, Robert Jera-Bezard and Monique Maillard (1976), Buddhism in Afghanistan and Central Asia I, Leiden.

I. Gershevitch ed. (1959), The Avestan Hymn to Mithra, Cambridge.

I. Gershevitch (1975), "Die Sonne das Beste," Mithraic Studies, I, Manchester, ( 1975): 68-89.

H. Goetz (1974), "The Crisis of the Migration Period and Other Key Problems of Indian History," Studies in the History, Religion and Art of Classical and Mediaeval India, ed. By H. Kulke, Wiesbaden, (1974): 64-86.

Bernard Goldman (1988), "The celestial chariot east and west," Bulletin of the Asia Institute, 2 (1988): 87-105.

F. Grenet (1993), "Baniyan and the Mihr Yast," Bulletin of the Asia Institute, 7 (1993): 67-94.

Bernard Goldman (1988), "The celestial chariot east and west," Bulletin of the Asia Institute, 2 (1988): 87-105.

R. Ghirshman (1962), Iran. Parthes et Sassanides, Paris.

Altert Grünwedel (1912), Altbuddhistische Kultstätten in Chinesisch-Turkistan, Bericht über Archäologische Arbeiten von 1906 bis 1907 bei Kuca, Qarashar und in der Oase Turfan, Berlin.

A. Gubaev (1971), "Sasanidskie bully iz zamka Ak-depe," Epigrafika Vostoka, XX (1971): 46-49.

M. Hallade (1965), "The ornamental Veil or scarf," East and West, 15(1965): 36-49.

E. Herzfeld (1920), "Der Thron des Khosro. Quellenkritische und ikonographische Studien über Grenzgebiete des Kunstgeschichte des Morgen- und Abendlandes." Jahrbuch der Preussische Kunstsammlungen, 41, pp. 1-24, 103-47.

A Howard (1991),"In support of a new chronology for the Kizil mural paintings," Archives of Asian Art (1991 XLVI): 68-83.

E. Huber (1906), "Termes Persans dans l'astrologie bouddhique chinoise," Bulletin de l'Ecole Française d'Extrême-Orient, 6 (1906): 39-43.

H. Humbach (1978), "Mithra in India and the Hinduized Magi, Études Mithriaques," Acta Iranica, I, IV, Leiden, (1978): 229-253.

A.V. W. Jackson (1914), "Notes and Allusions to ancient India in Pahlavi Literature," Festschrigt Ernst Windich, Leipzig, (1914): 209-212.

B.A. Jairazbhoy (1963), Foreign Influence in Ancient India, New York, 1963.

Li Jian ed.(2003), The Glory of the Silk Road—Art from Ancient China, Dayton.

D. Klimburg-Salter (1988), "Bamiyan: Recent Research," EW, 38 (1988): 305-12.

D. Klimburg-Salter (1989), The Kingdom of Bamiyan—Buddhist Art and Culture of the Hindu Kush, Rome.

Hans-J Klimkeit (1983), "The sun and moon as gods in Central Asia," South Asian Religious Arts Studies Bulletin, 2 (April, 1983): 11-23.

C. Lacour-Jalouneix (1983), "The origins of the image of the Sun-god in Northern India," South Asian Religious Arts Studies Bulletin, 2 ( April, 1983): 24-40.

Sherman E. Lee (1993), A History of Far Eastern Art, New York, 5th ed. 1993.

V. G. Lukonin (1971), "Po povodu bull iz Ak-depe," Epigrafika Vostoka, XX (1971): 50-52.

K. Maeda and A. Miyaji (1971), Bamiyan: Report of Survey in 1969, Nagoya.

B.I. Marshak (1995/6), "On the iconography of Ossuaries from Biya-Naiman," Silk Road Art Archaeology, 4 (1995/6): 299-321.

Penelope Mason (1993), History of Japanese Art, New York.

Shanti Lal Nagar (1995), Surya and the Sun Cult, New Delhi.

A. K. Narain (1990), "Indo-Europeans in Inner Asia," in Sinor (1990), pp.151-176.

K. Neuser (1982), ANEMOI Studen zur Darstellung der Winde und Windgotsheitenin der Antike, Rome.

Wendy Doniger O'Flaherty (Trans.) (1981), The Rig Veda: An Anthology, New York.

P. Pal (1975), Bronzes of Kashmir, New Delhi.

A. Panaino (1996), "The year of the Maga Brāhmanas", La Perisa e l'Assia Centrale da Alessandro al X secolo, Roma, pp.569-87.

L. P. Pandey (1971), Sun-Worship in Ancient India, Delhi.

D. P. Pandey (1989), Surya, Iconographical Study of the Indian Sun God, Delhi.

Marylin M. Rhie (1999), Early Buddhist Art of China and Central Asia, Boston.

E. Rose (1979), Die Manichaisohe Christologie, Studies in Oriental Religion 5, Wiesbaden.

John Rosenfield (1967), The Dynastic Arts of the Kushans, Berkeley.

M. Rostovtzeef (1922), Iranians and Greeks in South Russia, Oxford.

B. Rowland (1938), "Buddha and the Sun God," Zalmoxis: Revue des Etudes Religieuses I (1938): 69-84.

B. Rowland (1974), Asie Centrale, Paris.

Akira Sadakata (1997), Buddhist Cosmology, Tokyo.

Edward Schafer (1963), The Golden Peaches of Samarkand - A Study of T'ang Exotics, Berkeley.

I. Scheftelowitz (1933), "Die Mithra-Religion der Indoskythen," Acta Orientalia, XI (1933): 293-333.

D. A. Scott (1990), "The Iranian face of Buddhism," East And West, vol. 40, nos. 1-4 (1990): 43-77.

D. Shepherd (1983), "Sasanian Art," The Cambridge History of Iran, III (2), Cambridge, pp.1055-1113.

H. Steinmetz (1910), "Windgötter," Jahrbuch des Deutschen Archäologischen Instituts, 25(1910): 33-35.

H. ven Stietencron (1966), Indische Sonnenpriester, Samba und die Shakadvipiya-Brahmana. Wiesbaden.

Alexander Strelkoff (1938-39), "Iran and the pre-Islamic art of west Turkistan," A Survey of Persian Art, vol. 1, ed. Arthur Upham Pope, London, 1938-39, pp. 449-458.

Katsumi Tanabe (1990), "The Kushan representation of Anemos/Oada and its relevance to the Central Asian and Far Eastern wind gods," Silk road art and archaeology, 1 (1990): 51-80.

G. Verardi (1997), "Surya," Enciclopedia dell'Arte Antica Classica ed Orientale, Secondo supplemento 1971-94, vol.V, Roma, 1997, pp. 495-97.

Roderick Whitfield and Anne Farrer (1990), Caves of the Thousand Buddhas—Chinese art from the Silk Route, London.

Stig Wikander (1941), Vayu : Texte und Untersuchungen zur indo-iranischen Religionsgeschichte, Uppsala.

Xu Xinguo & Zhao Feng (1996), "A preliminary study of the textiles excavated at Dulan," China Archaeology and Art Digest, vol.1, n.4 (1996): 13-34.

Zhang Wenbin, ed. (2000), Dunhuang—a Centennial Commemoration of the Discovery of the Cave Library, Beijing.

Reference in Chinese and Japanese:

Chuang Shen (1960), "Mi ri kao," Bulletin of the Institute of History and Philology (Academia Sinica), 30 (1960): 271-301.

He Shizhe (1987), "Dunhuang Mogaoku di 285 ku xibi neirong kaoshi," Duhuang Shiku Yanjiu Guoji Taolunhui Wenji, Shiku Kaogo bian, Shenyang, 1987, pp.350-382.

He Shizhe (1999), "Shi shi za ji," Dunhuang Yanjiu, vol.62, 4 (1999): 50-55.

Higuchi Takayasu (1980): Bamiyan no Sekkutsu, Kyoto.

Deng Wenkuan (1998), "Qi yao ri," Dunhuang Xue Da Cidian ed. By Ji Xianlin, Shanghai, 1998, p.612.

Gu Sen (1997), Zhongguo Han hua tu dian, Hangzhou.

Ji Xianlin (1998), ed. Dunhuang Xue Da Cidian, Shanghai, 1998.

Jin Shen (2000), "Guanyu shenwang de tantao," 1994 Dunhuang Xue Guoji Yantaohui Wenji, Lanzhou, 2000, p.227-249.

Ma Shichang (1996), "Kizil zhongxinzhu ku zhushi quanding yu houshi de bihua," Zhongguo Shiku-- Kizil Shiku II, Beijing, 1996, vol.2, pp. 174-226.

Shanxi Museum (1985), Baoning Si Mingdai Shuilu Hua, Beijing, 1985.

Sato Masahiko (1965), Chugoku no dogu, Tokyo, 1965.

Soma Takashi (1970), "Koki-ko-Anosoku no kihei ni tsuite," Kōkogakku zasshi, 56.2 (1970):83-97.

Su Bai (1989), "Kizil bufen dongku jieduan huafen yu niandai deng wenti de chubu tansuo,"Zhongguo Shiku--Kizil Shiku I, Beijing, 1989, vol.1, pp. 10-23.

Yang Hong (1976), "Zhongguo gudai de jiazhou II", Kaogu Xuebao, 1976.2: 59-96.

Zhao Feng (1995), "Wei Tang Zhijin Yiyu Shendi", Kaogu, 2, (1955): 179-83.

Xingjiang Kuche Kizil Shiku Yanjiusuo (2000),Kizil Shiku Neirong Zonglu, Xinjiang, 2000.

Chongguo Shiku-Kumutula Shiku, Beijing, 1992>.

Reference in: T. TaishoShinshu Daizokyo

Da Lou Tan Jing, T. 23, vol.1.

Dirghagama, T.1, vol.1.

Abidharmashastra, T 1644, vol.32.

Qi Shi Jing, T. 24, vol.1.

Ekottaragama, T.125, vol.2.

Fantian Qiyao Jing, T.(images) vol.8.

Guang Hong Ming Ji T. 2103, Vol.52.

Song Gao Zeng Zhuan T.2061 Vol.50.

The Secret Text on Garuda and Other Celestial Beings, T.1278, vol.21.

The Sutra Spoken by Bodhisattva Manjushri and the Sages on the Auspicious and Ominous Days and Good and Evil Planets and Lunar Mansions, T. 1299, vol.21.

Fahua Yi Shu, T.1721,vol.34.

Miaofa Lianhua Jing Xuanzan, T. 1723, vol.34.

Buddhavatamsaka-mahavaipulya-sutra, T.278, vol.9


* The Ohio State University, Columbus

1 Some smaller cave sites in the adjunction area of Kizil in Kucha region also share the same features in terms of the depiction of the sun god and the wind god. To be convenient, Kizil will be the focus in the discussion in this paper.

2 Xingjiang Kuche Kizil Shiku Yanjiusuo, 2000, p.1.

3 Su Bai (1989), pp. 10-23. A Howard (1991), pp. 68-83.

4 Ma Shichang (1996), p.177.

5 Ma Shichang (1996), p.176. note.

6 There are a few variations showing the sun in symbolic form in Cave 38 or as a monk-like figure in Cave 34.

7 Ma Shichang (1996), p.177.

8 Ma Shichang (1996), p.177.

9 Jitendra Nath Banerjea (1948), p.48.

10 Benjamin Rowland (1938), pp.69-84; Ananda Coomaraswamy (1935a), p.44, n.2; Ananda Coomaraswamy (1935b), pp.25-28.

11 For the texts containing this epithet see the entry for adicca in T. W. Rhys Davids and William Stede (1966), p. 99.

12 David Efurd (1999), p.18.

13 I. Gershevitch ed. (1959), p.34.

14 Matteo Compareti (2000), p.341; B. Rowland (1938), pp.69-84; D. A. Scott (1990), pp.56-60; F. Grenet (1993), p.91.

15 For this study and more examples, see Hans-J Klimkeit (1983), pp. 14-5.

16 For some of the narratives that involved the sun god, see the Chapter II of David Efurd's thesis, pp.20-26.

17 Akira Sadakata (1997), p.25.

18 T. 1, vol.1, pp.145bc.

19 Dirghagama, T. 1, vol.1, p. 145c; Da Lou Tan Jing, T. 23, vol.1, p.306a; Qi Shi Jing, T.24,vol.1, p.359a.

20 T. 23, vol.1, p.289a.

21 T .1, vol.1, p.128a. Similar passage is also shown in the Pali canon. Of the Digha-Nikaya, see T. W. Rhys Davids (1969) p. 24.

22 T1, vol.1, p.147b c, 128a. T 23, vol.1, 307a

23 Helene Diserens (1997), p.335.

24 B. Rowland (1938), pp.77-78.

25 David Efurd (1999), p.64.

26 John Rosenfield (1967), pl.46; Shanti Lal Nagar (1995), pl.6.

27 B Rowland (1938), pp.76-77.

28 This sutra was cited into the volume four of Cheng Puti Ji in T(images) vol.8, p.122.

29 Deborah Klimburg-Salter (1989), p.161.

30 Deborah Klimburg-Salter (1989), p.158.

31 The sun god and the moon gods have been miss labeled from each other in this paining.

32 Leroy Campbell (1968), pp.134-138.

33 E. Rose (1979), p.1154ff.

34 Simone Gaulier, Robert Jera-Bezard and Monique Maillard (1976), p.23, Fig. 39.

35 Hans-J. Klimkeit (1983), pp.11-12.

36 Hans-J. Klimkeit (1983), pp.21.

37 For the information on the translator, see Song Gao Seng Zhuan vol.3T.2061 Vol.50, p720c.

38 T.1278, vol.21, p334c.

39T 1278, vol.21, p331a.

40 "Seven bay mares carry you in the chariot, O sun god with hair of flame, gazing from afar. The sun has yoked the seven splendid daughters of the chariot; he goes with them, who yoke themselves" The Rig Veda: An Anthology, trans. Wendy Doniger O'Flaherty (1981), p. 189-190.

41 M. L. Carter (1981) pp.77-78, note 16.

42 Bernard Goldman (1988), pp.88-90. For the reference on the Saka invaders, also see: M. Bussagli (1955), pp.16-17; P. Pal (1975), p.42; H. Humbach (1978), p.239; Pierfrancesco Callieri (1990), p.90; G. Verardi (1997), pp.495-96; Matteo Compareti (2000), p.336.

43 A.V.W. Jackson (1914):210; I. Scheftelowitz (1933), pp.293-333; B.A. Jairazbhoy (1963), p.153; H. ven Stietencron (1966); J. M. Rosenfield (1967), pp.192-195; Goetz (1974),p.68; I. Gershevitch (1975),pp.68-89; A. J. Gail (1978): 339 -41; H. Humbach (1978): pp.229-231; M. Carter (1981), pp.80-88; C. Lacour-Jalouneix (1983), p.27; M. Carter (1988), p. 132,137; D. P. Pandey (1989), pp.21-23, 65; Pierfrancesco Callieri (1990), pp.89-90; M. Boyce & F. Grenet (1991), p.485; A. Panaino (1996), pp.569-87; G. Verardi (1997), p.495; Matteo Compareti (2000), p.336.

44 Jitendra Nath Banerjea (1956), p.438.

45 Pandey 1971:66-88; Pandey 1989: 81-92, 100-108

46 Mario Bussagli (1955), pp.20, 25 n.46. Bernard Goldman (1988), p.99.

47 Mario Bussagli (1955), pp.19-20.

48 Zhao Feng (1995), pp.179-83; Matteo Campareti (2000), pp.331-368.

49 A. M. Belenizky (1980), pp. 146, 189-93; G. Azarpay (1981), pp.141-3; F. Grenet (1993), p.87; B.I. Marshak, (1995/6), p. 304, note 5.

50 Xu Xinguo & Zhao Feng (1996), pp.13-34.

51 He Shizhe (1987), p.359;

52 Yang Hong (1976), pp.68-80.

53 For the translation of the Cao Zhi's petition, see Albert Dien (1981/1982), p.16. It is still questionable if the term mingguang jia referring to the same format of armor when it was used in the third century and when it was used in the Tang dynasty several centuries later; and if mingguang jia is the two-plaque armor attributed by the contemporary scholars. However, to be consistent with the convention, I will use mingguang jia in this paper.

54 Yang Hong (1976), pp.68-80.

55 For this subject see Sōma Takashi (1970), pp.83-97. It was also mentioned by Albert Dien (1981/1982), p.41; M. Rostovtzeef (1922), p. 203.

56 Albert Dien (1981/1982), p.31.

57 Albert Dien (1981/1982), p.31.

58 Alfred von le Coq and Ernst Waldschmidt (1922-1923), vol.1, pl. 28.

59 For the different explanations of the cord, see Satō Masahiko (1965), p.102; Albert Dien (1981/1982), p.33-34.

60 Alexander Strelkoff (1938-39), p. 457.

61 M. Carter (1981), p.88.

62 A. Christensen (1944), p.144.

63 Pierfrancesco Callieri (1990), p. 91.

64 Combaz (1937), pp.131-91; L.P. Pandey (1971); D. P. Pandey (1989); Pierfrancesco Callieri (1990), p. 91.

65 One of such seal in Berlin, has very well been published, see E. Herzfeld (1920), pl 108, Fig. 14; Ph. Ackermann (1938-1939), pp. 784-815; R. Ghirshman (1962), p.243, Fig. 289; D. Shepherd (1983), p.1100; Pierfrancesco Callieri (1990), p. 87, Fig. 6. Several other seals from Ak-Depe, Turkmenistan, the illustrations have not been published. See A. Gubaev (1971), pp. 48-49; Lukonin (1971), p. 50; Pierfrancesco Callieri (1990), p. 86.

66 B. Rowland 1938:69-84; 1974: 86-87; Kilmburg-Salter (1988), pp.305-12.

67 T.1299, vol.21, p.387a.

68 The seven luminaries are the sun, the moon and the five planets.

69 T.1299, vol.21, p398b.

70 T.1299, vol.21, p.398b. Zhuang Shen (1960), p.287.

71 Chuang Shen (1960), pp.271-301; Deng Wenkuan (1998), p.612.

72 E. Huber (1906), pp. 39-43; Edouard Chavannes and Paul Pelliot Deng's opinion is cited in Dong Wenkuan (1998), p.612. Edward Schafer (1963) thinks that they are Iranian not actually Sogdian, pp.343, n.67.

73 Chuang Shen (1960), pp.280.

74 Guang Hong Ming J, v.8, T2103, Vol.52, p.47. There is an apocryphal text further equates the sun god and the moon god with Avalokite§vara and Mahasthamaprata who are the primary Bodhisattvas attending Amitabha (the Buddha of infinite light). Therefore, Fuxi would be the sun god and Avalokite§vara; Nuwa would be the moon god and Mahasthamaprata. This has been cited in two commentaries of the Sad-dharma-pundarika (The Lotus Sutra). Fahua Yi Shu, "The Commentary of the Lotus Sutra," T.1721, vol.34, p.464b; and Miaofa Lianhua Jing Xuanzan T. 1723, vol.34, p.675b.

75 For the study of the sutra, see He Shizhe (1999), pp.52-54.

76 Katsumi Tanabe (1990), p.62.

77 See A. von Le Coq (1935), p.87.

78 For the study of the depiction and transmission of the scarf, see Hallade (1965), pp.36-49.

79 Steinmetz (1910), pp.33-35; Raff (1978/79), pp.77-218; and Neuser (1982).

80 Leroy Campbell (1968), p.166.

81 Stig Wikander (1941), p.202ss

82 Chang Qing, p.1127-1141, Jin Shen, p.227-247.

83 For the review of the controversial discussion on the origin on the windbag, see Katsumi Tanabe (1990), p. 64-66.

84 Li Jian ed. (2003), p.144

85 Katsumi Tanabe (1990), p.51-80.

86 A unique standing wind god is found in Gandhara sculpture. He is with other figures and presumably attending the Buddha. See Rosenfield (1967), pl.76. This might be a variation of the running-type of wind god images to fit the context.

87 Penelope Mason (1993), Fig. 186a, & b. p.158.

88 Shanxi Museum (1985), p.217.

89 T. 1, vol. 1, p.136.

90 T.278, vol.9, p396a

91 T.1278, vol.21, p.334c.

92 Banerjea (1956) p.527.

93 Katsumi Tanabe (1990), p.60-61.

94 A. K. Narain (1990), p.173; Marylin M. Rhie (1999), pp.240-241.

95 Marylin M. Rhie (1999), pp.241-244.

© The Author/s

Actualizado el 24/07/2004

Valid HTML 4.01!
Any Browser!
Creative Commons License
Como lo hacemos?