The Image of the Buddha: Buddha Icons and Aniconic Traditions in India and China

Claudia Wenzel, Heidelberg Academy of Sciences and Humanities

I. Discourses on images: Byzantium and the East

This paper, based on my talk in the series “Byzantium beyond its Eastern Borders,” hopes to contribute to the study of aniconic Buddhist art in India and China by referring back to the iconoclasm (Bilderstreit) of the eighth and ninth centuries in the Byzantine Empire, and the subsequent development of an image theory, to justify the already well-established image cult. By deliberately adopting methodological approaches and terms that have been used for some time in Byzantine art history, relevant visual and textual evidence about the Buddhist tradition will be restructured and evaluated, and similarities and dissimilarities will be indicated.

Sometime between 726 and 730 CE, Emperor Leo III (717-741 CE) supposedly had an icon of Christ removed from the Chalke Gate of his palace and replaced it with a cross, the symbol of the former power and magnificence of the empire. This act, which at first may not have intended to be iconoclastic, is seen as the inauguration of the first iconoclastic period (730-787 CE) in Byzantine (Krannich et al. 2002, 4-5). Under Leo’s son, Constantine V (741-775 CE), the theological debate between iconodules (those who venerate images) and iconoclasts (those who destroy images) took shape. The iconoclasts won their first victory with the Synod of Hiereia, presided over by Constantine V in 754. Their position was manifested in the horos of the Synod, which addressed issues of “false” and “true” images, and the fundamental problem of how the divine or the nature of Christ can be represented at all. However, after Constantine’s death, the resolutions of the Seventh Council of Nicaea in 787 under Empress Irene ended the first iconoclastic period and reinstated the cult of images. During the second iconoclastic phase (813-842), the debate heated up one more time and the conflict escalated. Even after the restoration of icons by Empress Theodora in 843 and the Triumph of Orthodoxy, the rivalries between the emperor and the patriarchs continued.

Not only text sources, but also polemical miniatures in book illuminations like that of the Chludov Psalter reflected the theological controversies. Right from the beginning, images were at the centre of the debate, which developed with increasing aggressiveness. Empress Irene had the cross above Chalke Gate again replaced by an icon of Christ that was later stoned by soldiers of Leo V in 814 when the second iconoclastic phase began. Another cross took its place. Finally, a mosaic of Christ supplanted the symbol (Belting 1994, 159-160). Likewise, figural representations in the church of Hagia Sophia had been destroyed and replaced with crosses (Speck 1998, 63, Abb. 47).

The image wars of the Byzantine iconoclasts had a seminal impact, not only on the later history of the religious icon, but also on the theoretical foundations of image theory in general. The Christian theologians finally had to take a stand on what an image is or can be, or what it should not be taken for. In this respect, Byzantine iconoclasm has turned out to be immensely fruitful for the modern discipline of art history in the West.

However, in the field of East Asian art history, not much is yet known about a comparable discourse on religious images. One reason for the lack of a coherent study on Chinese Buddhist image discourse is certainly the “disconnected discourse on images in textual sources,” as Shinohara Koichi has rightly pointed out (Shinohara 2004, 207). At first glance, it may even seem that such a discourse on image theory is not to be expected in Buddhism because, unlike the monotheistic religions of Judaism, Christianity and Islam, Buddhism never bothered about prohibiting the making of images of the Enlightened One. This is true for early Indian Buddhism, which had not yet developed representations of the Buddha in human form, as well as for Chinese Buddhism, which learned about Buddha images when the new religion spread to its territory during the first two centuries CE. At that time, image making was already promoted as an adequate means of accumulating religious merit. Still, the Chinese were well aware of the question of how to represent the ultimate truth of Buddha’s enlightenment.

Buddha Amitabha

Fig. 1: Buddha Amitābha with inscribed pedestal, dated 746. Grey limestone, height including pedestal 94 cm, width of pedestal 49,5 cm, photograph by Jürgen Liepe, courtesy of Staatliche Museen zu Berlin, Stiftung Preußischer Kulturbesitz, Museum für Asiatische Kunst, Ostasiatische Kunstsammlung.

Berlin’s Museum für Asiatische Kunst holds an Amitābha statue (figure 1) that was created around the time when the first iconoclastic phase had reached its climax in Byzantium. The statue illustrates the contemporaneous Chinese discourse on image theory. It is dated to the fifth year of the era Tianbao 天寶(746) of the Great Tang Dynasty that constituted the image friendly, golden age of Buddhism in China. The pedestal of the statue bears a votive inscription that provides information about the donors and their motivation for sponsoring “one stone image of Amitābha” 石弥陁像一區, a description of the statue’s original location in terms of a sacred geography (“In the north [the statue] connects to the high hills that are Mount Gṛdhrakūṭa; in the south it faces the great river that is the eightfold meritorious water [of the Pure Land]” 北連高阜,乃耆闍崛山;南對大川,即八功德水。), and the standard wish for a good reincarnation. The opening lines read as follows:

As a matter of general principle, while highest truth is devoid of any image, without images there would be nothing to make visible its [being the] truth; and while highest principle is devoid of all words, how, without words, would its [being the] principle be made known.[2]

The inscription’s tone of apology for image making is not at all an isolated case. There are many more examples of votive inscriptions on statues or image stelae[3] that justify the creation of the work, or at least reflect upon the need for such holy representations. What is remarkable is that these inscriptions discuss the problem of representing the highest truth not only for the medium of images, but also for the medium of words or language. Yet, whatever medium is chosen, the central question that remains is the truthfulness of representation. The inscription on the Amitābha pedestal from 746 names such truthfulness “highest truth” (zhizhen 至真) and “highest principle” (zhili 至理).

Truthfulness of representation was likewise the point of departure for the Byzantine theologians who debated on the refusal or affirmation of images. Basically, the Christian Godhead is invisible and can therefore not be represented in material matter. Of God’s three hypostases (Father, Son, and Holy Spirit) only the Son, the incarnate Christ, became visible. As Christ in his incarnation has shown, the divine can touch material matter and is able to imbue it, which in turn explains the efficaciousness of relics, both primary and secondary. Charles Barber has shown how the Christian icon was already well established in the late seventh century and prior to the outbreak of iconoclasm, by tying it firmly to the cult of relics, because “both are matter transformed into a holy state—the relic having become holy by contact and the icon by having a specific form, a likeness, impressed into its material nature” (Barber 2002, 36). In the course of the following two centuries, iconophile patriarchs developed a sophisticated construct to explain how man-made icons can be truthful images because of their participation in the divine, and they continually refined their analysis of the relation between the archetype—Christ in his incarnation—and his image. Leading thinkers of the iconoclasts’ camp, on the other hand, insisted that the only admissible icon is the Eucharist, sanctified by the declaration “This is my body”; apart from it, only the cross could be used to represent the Godhead, yet not as an icon, but as a figure or sign thereof (Barber 2002, 103-104). In addition, these iconoclasts saw the figure of the cross as equivalent to verbal testimony, emphasizing the antithesis of word and image (Barber 2002, 93).

As John Strong (Strong 2004, 3) pointed out in quoting Robert Sharf, contemporary buddhologists have noticed that “Buddhism…..has come to bear an uncanny resemblance to medieval Christianity… [with] its saints, relics, and miraculous images” (Sharf 1999, 79). While a systematic comparison between Christian and Buddhist image theories is missing, efforts have been made in recent years to get a clearer picture of the nature of religious images in East Asia. The two most remarkable volumes in this respect are the books on Images, Miracles, and Authority in Asian Religious Traditions, edited by Richard H. Davis in 1998, and Images in Asian Religions: Texts and Contexts, edited by Phyllis Granoff and Koichi Shinohara in 2004. The conclusions reached by the contributors of the first volume are at times quite contradictory: In examining miracles performed by Buddha images, Robert L. Brown states that the relationship between the Buddha and his image is not a direct one, and other factors like the objective nature of the image or its personality are more likely agents for the miraculous powers of images (Brown 1998: 50-51). For Phyllis Granoff, the painted image of the Buddha is, like his mantra, capable of having an effect in this world because of the Buddha’s inherent presence (Granoff 1998: 76). And by considering icons of the Buddha at three different levels of decreasing significance from original to copy and replications of the copy, Donald F. McCallum seems to indicate a decrease in the Buddha’s presence from original to various replications (McCallum 1998: 218-221). Although these are valuable contributions to an understanding of the Buddha’s image as a wonder-working icon, not much has been said about representations of the Enlightened One in non-pictorial form. Consequently, in their second volume, Granoff and Shinohara stressed the need for more scholarly works on apologetics for and arguments against image worship in the Asian context (Granoff and Shinohara 2004: 1).

For this study I will rely more strictly than before on the methodological construct provided by studies in Byzantine image discourse to understand claims of truthfulness in representations of the Buddha. To this end, I differentiate between modes of representations that I like to term “iconic” and “aniconic.” For a long time both terms have been in usage in art-historical discussions of early Indian Buddhist art. They pertain to representations of the Buddha in narrative reliefs depicting him either in human form (“iconic”), or lacking an anthropomorphic image (“an-iconic”). However, for the wider scope of this study, that also includes medieval Chinese Buddhism, these terms need further clarification. The “iconic mode” still relates to anthropomorphic depictions of the Buddha that appeared in India at the end of the first century CE at the latest and prevailed thereafter. It is reasonable to further divide iconic representations into more general depictions of the Buddha in human form and the Buddha’s true icon in particular. For the latter group, a discussion about Byzantine icon theories may prove especially rewarding. The “aniconic mode” of representation is found in the Indian cultural sphere as well as in China, but it produced distinct modes of expression according to the different historical settings. Early Indian Buddhism developed various aniconic types within the scope of narrative pictures that never show the Buddha in human form. Instead, his presence was indicated by means of symbols or by emblematic representations as well as by his “non-image.”[4] In contrast, the Chinese aniconic phase refrained from any pictorial representation whatsoever and only carved the words of the Buddha in Chinese calligraphy. The restriction to words on the Chinese side has the advantage that it provides written texts that can be analyzed, although the texts themselves never fail to suggest the limitations of language in the same way that suggest the inadequateness of images, as we shall see. Early Indian Buddhism, it seems, knew other ways to avoid the limitations of language: without leaving the confines of the pictorial, it found impressive ways to give form to the invisible.

II. Aniconism in India: symbols, emblematic representation, and the non-image

From the very beginning, Buddhist practice and devotion in India was aniconic in the sense that anthropomorphic images of the Buddha were not made. But this first aniconic phase was in no way iconoclastic. Early Indian Buddhists did not have an image of the Enlightened One. Just like Hinduism and Jainism, Buddhism did not know of any icons to be venerated during the four centuries before CE. For that reason, there was also no need for an outspoken prohibition of image making. The Buddha was simply not rendered in human form, but that does not at all mean that he was not represented.

With regard to India, I would propose that three different kinds of aniconic representations can be distinguished: Firstly, by alluding to the Enlightened One with a certain set of symbols; secondly, by emblematic representations of the Enlightened One’s body; and, thirdly, by representing him as a non-image.


An example for the first kind of aniconic representation is the Veneration of the new-born bodhisattva by gods, a relief on the Ajātaśatru Pillar of the stupa from Bārhut in India from the second century BC (Snellgrove 1978, 30, plate 8). The new-born bodhisattva, Buddha Śākyamuni, in his last rebirth as Gautama Siddharta, occupies the center of the relief but remains invisible. Instead, his presence is symbolized by a richly decorated seat, surmounted by a canopy, and a pair of footprints, that are touched by the god Maheśvara in veneration. An inscription gives evidence for this iconographic theme (Snellgrove 1978, 30). Since the inscription explicitly mentions that these are the gods venerating the new-born bodhisattva, there is no doubt that the future Buddha is meant to be present in this relief.

The same might be assumed for another relief on the same stupa pillar from Bārhut, the Descent of the Buddha from the Heaven of the Thirty-three Gods (Snellgrove 1978, 30, plate 9). After his enlightenment, the Buddha ascended to the heaven of the thirty-three gods to preach to his deceased mother who had been reborn into this heavenly place. Waiting for the Buddha’s return, the god Indra built a triple staircase for him to descend from this heaven. In the center of the relief we see this staircase and the prints of Buddha’s feet on it; one foot at the topmost step and the second on the bottommost. Worshippers are lined up to the right and in front; to the left, a decorated throne in front of a tree illustrates the following visit of the Buddha to a nearby town. 

A condensed version of the Buddha’s Descent from the Heaven of Thirty-three Gods from Butkara was recently on display in the Gandhara exhibition in Bonn, Berlin, and Zürich (Gandhara 2008, 106, Kat. Nr. 45). It may be interesting to note that the Butkara relief depicts the Buddha’s footprints in a traditional way, parallel to each other on the bottommost step of the staircase, while the earlier Bārhut relief separates them, indicating perhaps the actual movements of the Buddha as he descends, thus strengthening the Buddha’s actual presence in the image.

In some cases the problem of the Buddha’s presence in the image is much more complex and diversified. The Enlightenment at the Tree Sanctuary on the Prasenajit pillar of the Bārhut Stupa (Snellgrove 1978, 42, plate 22) is a well known example of this complexity of meaning. The tree of enlightenment with its characteristically pointed leaves in the topmost panel of the relief is decorated with garlands and parasols and venerated by the gods standing next to it, indicating that the enlightenment has just taken place. But the architectural structure underneath the tree indicates that a sanctuary has already been build around the holy place where worshippers venerate the Buddha’s richly decorated seat. Behind the seat, two examples of the so-called omega symbol appear, each consisting of a lotus wheel and a trident (trīsūla) that represents the three jewels: Buddha, dharma (doctrine) and sangha (community). Furthermore, the veneration scene is separated by a fence from more worshippers shown below, and a tall column crowned by an elephant that was erected outside the fence (Snellgrove 1978, 42).

While it might be argued that the relief depicts the sanctuary of the Buddha’s enlightenment, rather than the actual enlightenment itself, the inscription incised at the roof of the sanctuary (bhagavato Sakamunio bodho) translates as “The Enlightenment of the Lord Śākyamuni,” thus pointing to the actual event. Furthermore, one figure seated to the left of the middle panel can be identified as Māra (Ghandara 2008, 74, Abb. 3), who is depressed because he had been defeated by the Buddha, who he could not distract from attaining enlightenment. The relief’s inscription and the figure of Māra indicate the Buddha’s actual presence in the image, while the architectural structure does not (Luczanits 2008, 73). This seeming contradiction might not have bothered the Buddhist believers at Bhārhut in ancient India because they had already conceived of the Buddha as a divine being whose enlightenment was timeless, or beyond time, and for them the relief was showing “a scene of a cosmic and divine event” (Brown 1998, 44).

Emblematic representations

The second mode of aniconism is the emblematic representation of the Enlightened One’s body. One striking example can be found at the Northern gate of “stūpa one” in Sāñcī from the first century CE (Gandhara 2008, 76, Abb. 6, and Snellgrove 1978, 41, plate 21). Here, the Buddha’s body is represented from top to bottom by the omega symbol crowning a lotus rosette, a column like structure consisting of two parallel borders framing a sequence of floral motifs, and the Buddha’s footprints with the Wheel of the Law (dharmacakra). These symbols are joined to represent the head, body axis, and feet of the Buddha.

Emblematic representations are continued in the south of India (Luczanits 2008, 75), for example, at the great stupa at Amarāvatī and in a well-known relief from the stupa’s railing (Klimburg-Salter 1995, 103, Kat. Nr. 61). Rāhula, the natural son of the Buddha, is brought to his father. His mother had sent him to ask for his inheritance, and the Buddha thereupon ordained him as a monk. In the background of the relief two houses, a curtain, and a door indicate a public space, separate from the interior, where a crowd has assembled around the Buddha’s throne. To the left, several palace women watch the little boy Rāhula, who is gently pushed towards his father’s seat by a nobleman. Two more women kneel in front of the Buddha’s footprints, while the space to the right is occupied by a crowd of ordained monks. The Buddha’s seat is a high throne on which a round cushion has been placed. Behind is a flaming column with a Wheel of the Law and the omega symbol. While the throne occupies the center of the relief, the flaming pillar runs along the central axis, thus positioning the unseen Buddha in the very center of the composition. The beholder is even able to imagine the more than life-sized figure of the Buddha, who is said to have been extraordinarily tall.

The non-image

The third mode of aniconic representation might be called “the pictorial non-image.” It uses compositional means already seen in the emblematic representation of the Buddha’s body, but in this case without depicting anything at all. Another scene from Amarāvatī offers a fine example. It is part of a relief illustrating four scenes: the conception of the Buddha by the dreaming Māyā, the interpretation of the dream to the king and queen at court, the birth of the Buddha, and the presentation of the newborn to the wise Śākyavardhana, who foresees his Buddhahood (Knox 1992, 121, plate 61).

In the birth scene, queen Māyā is standing to the right, next to a smaller female servant. Her outstretched left arm reaches up to the branches of a sāl tree, under which she is about to give birth. While pushing her right hip forward, the baby Buddha emanates from her right side without causing her any pain. The queen’s pose resembles that of a dancer, emphasizing the ease and painlessness of the birth. To the left, four male figures holding a long cloth are arrayed in a semicircle. These are the four lokapālas, or heavenly kings, who are prepared to receive the newborn child.

In the pictorial composition of the birth scene, all the figures are positioned along a line encircling the new-born Buddha, The last lokapāla and the female servant are, therefore, seen from behind, while it is to be imagined that the beholder of the scene steps forward to complete the circle around the baby Buddha. The stool next to Māyā functions not only as repoussoir, but also points explicitly to the center of the composition where the holy birth takes place. This center is empty, or rather, occupied by the Buddha’s non-image. To ensure that the beholder partakes in viewing the Buddha’s non-image, two tiny footprints have been added on the section of the cloth that the first lokapāla holds, right in the center of the composition as a whole. It can indeed be said that the Buddha’s absence in the very center of the picture gives weight to his actual presence.

This last example dates to the second century CE, when representations of the Buddha in human form had already begun to be created. Seckel proposed that during a transitional period in the second and third centuries CE, aniconic and iconic representations of the Buddha coexisted (Seckel 1976, 33-36 und 2004, 49-54). Still, his attempt to identify samples that unite human and symbolic representations of the Buddha in one image was only partially successful. The Amarāvatī relief of three anthropomorphic Buddhas alternating with three stupas (Bachhofer 1929, II, plate 131) that he mentions (Seckel 1976, 33 and 2004, 49) dates to the eleventh rather than to the third century. For the relief of the worship scene (Bachhofer 1929, II, plate 129) showing the Enlightened One in human form below, and the adoration of his emblematic representation in the register above (Seckel 2004, 49-50, plate 52), Bachhofer has given a more convincing interpretation, suggesting that the representations refer to the Three Jewels and stand for the Buddha and the dharma, while a third representation of a stupa, symbolizing the sangha, is lost.

Nevertheless, there are rare examples that combine the Buddha in iconic and aniconic mode, such as the wall painting from the central Asian cave temples at Kyzil of the first quarter of the seventh century (Seckel 2004, 19, figure 9, and Karow 1989, 147), formerly in the Museum für Völkerkunde in Berlin, until it was destroyed in World War II. Part of the wall painting depicting the Ajātaśatru legend is a bodhisattva who presents a piece of cloth illustrating the four great events in the Buddha’s life: his birth, enlightenment, first sermon and death. In all scenes, the Buddha is portrayed in human form; only in his birth scene does he remain unseen and is symbolized by a halo.

Looking at the three different kinds of aniconic representations of the Buddha shown above, it might be said in conclusion that the early ‘aniconic’ phase of Indian Buddhism seems to share with Christianity the notion that the highest truth—Buddhahood or enlightenment, and the godhead respectively—was ultimately invisible. In fact, invisibility of the Buddha secured the truthfulness of the depiction. Different modes of invisibility were deployed for pictorial representation. First of all, symbols may represent the presence of the Enlightened One. Among these, the footprints of the Buddha deserve special attention because of their close affinity to relics. As already noticed by Quagliotti 1998, 126-127, in quoting Brown 1990, 95-96, the Buddha touching the ground with his feet has to be understood as an “act of grace,” and the footprints left behind are thus relics of touch.

In emblematic representations like the one from the Northern gate of stupa one in Sāñcī, the Buddha’s footprints are complemented with more symbols that roughly delineate the position of the Buddha’s entire body within the pictorial space. This development is taken even further in the Buddha’s non-image, where the space for the human body of the Buddha is already outlined but remains invisible. An additional emphasis is provided by the symbolic footprints that reassure the observer of his appearance in the human world.

III. The icon of the Buddha in India and China

Representations of the Buddha in human form can be attested in India at the end of the first century CE, at latest. They emerged around the same time in Māthura, south of New Delhi, and in Gandhāra in the Northwest of Pakistan. One of the earliest dateable sculptures of the Buddha is dated by inscription to the thirty-second year of the Kaniska era, which probably corresponds to the year 110 CE (Klimburg-Salter 1995, 115-116, and Sharma 1984, 190). It was carved from the white spotted sandstone of the Māthura region (Klimburg-Salter 1995, 115-116 and 251-252; Sharma 1984, 190). Sculptures like this show the Buddha in human form but not in a form that any ordinary human being might be able to possess. The Buddha had to be portrayed as the fully Enlightened One in human form. This was achieved by endowing his image with the special bodily attributes that all enlightened ones are said to possess, not only the historical Buddha, born as Siddharta Gautama, but all Buddhas of the past and future, in our, or any other world of the ten directions. The importance of the bodily attributes for any Buddha image in India and China, no matter if painted, sculpted, or even mental, can hardly be overestimated. Descriptions of these are listed in the Buddhist canon, specifying thirty two primary and eighty secondary attributes (xianghao 相好) of a superhuman being. While most of the former appear in all lists, some attributes appear in only some lists. For the most part, however, these lists are homogenous. Some of the thirty-two primary attributes found in the text sources are reflected in Buddha images, such as the early example from Māthura: The Buddha’s well proportioned, extremely tall, and dignified body with its slender and long limbs is likened to that of a lion king. Other attributes mentioned in the texts pertain in particular to his head, hands, and feet, thereby giving shape to the Buddha’s superhuman uşņīşa and ūrņā as well as to the dharma wheels on his hands and feet. His skin was said to be of a golden hue, emitting light from every single pore. In short, all parts of the Buddha’s body, even those that are not easily depicted like his teeth or tongue, or even features that cannot be represented, like his voice, are determined by his being enlightened. His physical shape and condition are a necessary effect of his enlightenment experience.

A remarkable parallel can be drawn with the icon of Christ. The portrayal of his physical features follows the notion of a “delineation,” which places every human being into a coded system of particular features that establish a distinctive identity. In Letter of the Three Patriarchs to the Byzantine Emperor Theophilos (829-42), the patriarchs of Alexandria, Antioch, and Jerusalem stated that “The icon of a man is not inscribed according to nature, but according to position.” (Barber 2002, 107-110). In other words, there was no doubt about what the historical, incarnated Christ “looked like,” because there was only one way the son of the tripartite Godhead could have manifested physically in the world. Taking this thought even further, iconophile theologians such as Theodore of Stoudios (759–826) underlined in their writings the fact that an icon has to be seen as a living eyewitness of actual events (Barber 2002, 131).

In this manner, the thirty two major bodily attributes of a Buddha are witness to his completion of awakening and are, therefore, the only adequate way to represent him in human form. Even though the Buddha is shown in his human shape as a result of his last incarnation in the human world, his particular bodily marks are witness to his transcendence of the human realm.

This is why the bodily attributes displayed in Buddhist sculpture varied only slightly over the course of time. Generally speaking, differences in style aside, the iconography of the Buddha’s human form remained remarkably stable. During the Chinese Northern Qi Dynasty (550-577), new stylistic influences from India and Central Asia had been absorbed. A beautifully sculpted and gilded Buddha from the spectacular 1996 Qingzhou findings (Nickel 2001, 165-167, Katalog Nr. 18) presents the typical style of this period. The youthful and elegant figure displays a low uşņīşa, while the ūrņā is either missing or simply painted. The face and the uncovered parts of the body were gilded, representing the light emitting from the Enlightened One’s body. The Qingzhou sculpture is only one of countless examples of Chinese Buddha icons that were made of stone, wood, or bronze. They were portable or set up in wooden temple architecture or hewn directly from the walls of cave temples. They all testify to a firmly established image cult in China that is further evidence of numerous written and painted sources.

The tradition of King Udyāna’s first image

In the Christian tradition, the legend of St. Luke the evangelist, who portrayed the virgin mother and the son, was developed roughly from the sixth to the eighth centuries, and propagated the idea that Christ’s image, but also that of his mother, the virgin, can be regarded as an historical portrait (Belting 2004, 70-72). In this way, images of the mother and child not only claimed authenticity, but also lent credence to the historicity of the divine event of the incarnation. The Buddhist tradition knows of a comparable legend about a first image of the Enlightened One; but, in contrast to the painted portrayal by St. Luke, the Buddha’s first image was three dimensional and carved of sandalwood.

Earliest references to the creation of a first image by the youthful king Udyāna (優填 or 優陀延王), King of Vasta, are found in two canonical scriptures on image making—the Sutra Spoken by the Buddha on the Making of Buddha Images (佛說作佛形象經, T# 692, 16:788a-c) translated in the Later Han-dynasty (25-220), and the Sutra Spoken by the Buddha on the Retribution of Merits [Attained by] Making and Installing Buddha Images (佛說造立形象福 報 經, T# 693, 16:788c-790a) translated in the Eastern Jin dynasty (317-420). The legend is further developed in the Ekottarāgama (增一阿含經, T# 125, 2:706a2-26) by Gautama Sanghadeva from around 385; it has been treated extensively by Soper (Soper 1959, 260b-261b, 259-265, 46-49, 70-71, 88-89) and, more recently, by Carter (Carter 1990, 26-27) and Shinohara (Shinohara 1998, 153-154 and 169). In the late fifth century, the Udyāna legend was linked with the story of the dream of Emperor Ming (58-76 CE) of the Han dynasty, the official tale of how Buddhism was introduced to China. The now lost scripture Mingxiang ji 冥祥記by Wang Yan王琰 claims that the Buddha’s painting that the emperor received was in fact the same first image produced at the behest of king Udyāna in the Buddha’s lifetime, thus giving authority to Chinese icons of the Buddha (Carter 1990, 1-2). However, the older version of the story in the Ekottarāgama does not mention a painting but refers to an image sculpted from sandalwood:

King Udyāna was distressed at not being able to behold the Enlightened One’s appearance when the Buddha dwelt in the heaven of the thirty-three gods to preach to his late mother Maya. Therefore Udyāna decided to have a sandalwood image made, five feet in height. When the Buddha descended from the heaven of the thirty-three gods, he accepted it and pointed out the various merits achieved by image making (T#125, 2:706a2-26 and 708a27-b14). This scene is found illustrated in a stone relief from Gandhāra (Karow 1989, 89), where we see the Buddha holding the image up in his hand and Udyāna kneeling in front of him.

The Udyāna legend became embellished over time with details relating how one or many artisans were miraculously transported to the heavens to create the likeness of the living Buddha. Competing with the Udyāna story, a second legend about a golden image, ordered by king Prasenajit, soon emerged. Udyāna’s image seems to have enjoyed greater popularity in China because it was the sandalwood image that appeared in the dream of Emperor Wu (r. 502-549) of the Liang dynasty, who sent envoys to Sravasti to fetch it (Shinohara 1998, 153-155). Apart from the seminal role the Udyāna legend played in establishing an image cult in China, it also set the stage for an upcoming image discourse, as we shall soon see.

Even today, there is an icon named the Udyāna Shaka in the Seiryoji in Kyōtō that is believed to be a direct copy of the legendary Udyāna image that was brought from China to Japan by the monk Chonen in 985. One Japanese version of the story of how the icon came to Japan even claims that the original found in China was secretly transposed with the copy that had been made for the Japanese delegation. Seiryoji’s Shaka icon was always believed to embody miraculous powers that could be transferred to a replica, if the copy was done with appropriate accuracy (McCullum 1998, 219-221). Of the roughly one hundred surviving replicas, the most famous is the Saidaiji Shaka that was created in 1249 in front of the original icon in Seiryoji. Upon its completion, the Saidaiji Shaka miraculously produced bead-like relics as a sign of its sacred power (McCullum 1998, 214-215).

Divine and miraculous images (ruixiang 瑞像)

The Udyāna story does not only establish the tradition of a first image in the likeness of the Enlightened One, but also makes a claim for the anthropomorphic Buddha image to be a divine image (lingxiang靈像 ) that is religiously efficacious, even to the point that it is equal to the extraordinary salvific powers of the Buddha’s relics. The salvific efficacy of a divine image is affirmed by its ability to work miracles, which, once it happens, turns out to be not at all surprising, but rather expected and matter-of-fact. Furthermore, the type of miracles produced by images is foreseeable and consists primarily of various appearances of light and the self-induced movement of the image (Brown 1998, 26-31).

Stories about miraculous images (ruixiang 瑞像) were recorded in various text sources compiled during the sixth and seventh centuries and range from the Biographies of Eminent Monks (Gaoseng zhuan 高僧傳, T#2059, 50: 322-424) by Huijiao 慧皎, from around 531, to the Assembled Records of Response of the Three Jewels in the Spiritual Realm (i.e. China) (Ji shenzhou sanbao gantong lu集神州三寶感通錄, T#2106, 52: 404-435), compiled in 664 by Daoxuan 道宣(596-667). Within these miracle stories, a group of marvelous icons termed Aśoka images can be identified. Historically, King Aśoka (reigned ca. 273-232 BCE) of the Maurya Dynasty left his famous stone edicts to testify to his widespread propagation of Buddhism, but at his time still no anthropomorphic Buddha images were known. In China, his fame for having distributed 84, 000 relic stupas all over the world probably suggested that he had also sent sacred images (Shinohara 1998, 141-148). Even though stories of Aśoka images had become popular in the south of China, the notion of images possessing miraculous powers probably first arrived in the northern parts of China in the fifth century, and furthermore seems to have come from India proper (Shinohara 1998, 159). These Aśoka images are characterized not only by their wonder-working abilities, but also by their miraculous origin that is usually narrated in stories about wondrous discoveries.

One typical story about such images is the finding of two stone images floating on the Song River淞江 in Wu 吴during the Western Jin period (265-316) that were brought later to the Tongxuan temple 通玄寺. The account is preserved in several versions in a number of texts that were listed by Shinohara (Shinohara 1998, 176). One identifies them as Aśoka images, and was illustrated during the Tang dynasty in a mural on the west side of the south wall in cave 323 at Dunhuang with narrating text in accompanying cartouches (Fojiao dongzhuan gushi huajuan 2000, 147, plate 122).

To the upper right, two large Buddha statues on lotus pedestals with halos are depicted in the middle of a river. Cartouches identify them as “Buddha Vipaśyin” 維衛佛 and “Buddha Kāśyapa” 迦叶佛. Some people have already gathered at the shore, among them monks who pay respect to the miraculous apparition. The cartouche beside it tells the whole story:

Here are the two stone Buddhas from the Western Jin dynasty that came floating along the Song River in Wu. As the billows gathered strength, [the statues] floated against the current and approached the boatmen, who took a hold of them. On the Buddhas’ robes there were inscriptions; the first read: Buddha Vipaśyin; the second read: Buddha Kāśyapa. These statues are seen in Wu Prefecture for worship.

In the middle ground, the statues have already been taken aboard a boat by some laymen and a monk, who seems to point to the direction of their final destination. A small cartouche above the boat scene gives the location as “River Song in Wu” 吴淞江. Nevertheless, the endeavor of securing the images had not been an easy one. To the right, a ceremonial space has been delineated by raising fan-banners 幡 (Wu Hung 1986, 302), and at least two Daoist figures are shown paying respect. The text in the cartouche beneath them explains:

When the statues floated along the river, the world hoped for an auspicious omen. Invited […]  calling […] the Way sent [them] down, [Daoist priests] welcomed the statues with seals and sacrificial ceremonies. [Since they] could not obtain [the statues] for several tens of days, they returned.”

In contrast, Buddhist monks and laymen were more successful. The cartouche directly beneath the boat tells the end of the story:

The occurrence of a numinous response does not depend on the works of people. When believers in the Buddhist Law held that [the statues] had been sent by the Buddha, the wind and waves calmed down, and [the statues] were welcomed and taken to Tongxuan monastery where they are worshipped until this very day.

As this example shows, the origin of images was considered highly significant. Apart from Buddha icons that are associated with Emperor Aśoka, the miracle stories also account for images that were discovered buried in the ground, in relic-like manner, usually with some supernatural appearance of light indicating the place where the pious finder was supposed to dig for them (Shinohara 1998, 148-151). In this way, the miracle stories reflect the authorization of an image cult being established in China. One of the early propagators of such a cult seems to have been Emperor Wu (r. 502-549) of the Liang dynasty, who often appears and plays a central role in such miracle stories, and whose figure apparently was modeled after the prototypical emperor Aśoka (Shinohara 1998, 152-156).

The true countenance (zhenrong 真容): The Buddha’s shadow

Apart from the Udyāna (or Prasenajit) image that represents the legend of the Buddha’s first image, and the notion of wonder-working divine images exemplified by the Aśokan type of images, the Buddhist tradition also knows of another true image of the Enlightened One, the so-called Buddha’s shadow (foying 佛影) at Nagarahāra (Hadda, Afghanistan). It might be assigned to the category of acheiropoietos images, those “not made by hand.”

In the Byzantine context, the acheiropoietos images can be seen as an early solution to the problem of manufactured items being thought of as unsuitable for the accommodation of the divine. In the later sixth century, two particular important acheiropoietos icons that claim a miraculous origin became known: The sacred Mandylion, a cloth bearing Christ’s portrait kept in Constantinople, was understood to be his bodily imprint and, therefore, worshiped as a relic of touch (Belting 2004, 64-70). John of Damascus narrated in his writings how Christ “took a cloth, and having pressed it against his face, impressed its portrait upon the same, which it has kept until now.” The so-called Camouliana icon, also on cloth, was discovered by a pious Christian woman hidden in a fountain in her garden. It had the wondrous ability of making copies of itself when it came in contact with other materials, and even transferred its protective and talismanic powers to the copy (Barber 2002, 24-25).

The Buddha’s shadow at Nagarahāra is the image that comes closest to the idea of an image that was not made by human hands, but by the Buddha himself by means of bodily contact. According to legend, the Buddha penetrated the rock of the cave at Nagarahāra with all of his body and manifested his shadow on the wall as a kind of reflection for the sake of the nāgas that had been converted by the Buddha. Various accounts by traveling monks on the Buddha’s shadow cave have been transmitted, among them the famous report by Faxian who traveled in India from 399 to 414. They all tell of a particularity of the shadow, namely that it appears clearly only when seen from a distance, but becomes dull as one approaches (Soper 1959, 265-268). The apocryphal Sutra on the Ocean-like samādhi of Buddha Contemplation (Guan fo sanmei hai jing觀佛三昧海經, T#643, 15:645c-697a) traditionally said to be translated by Buddhabhadra from sometime in 412, contains the best known description of the phenomena:

釋迦文佛踊身入石,猶如明鏡人見面像,諸龍皆見佛在石內映現於外。[...] 爾時世尊結加趺坐在石壁內。眾生見時,遠望則見,近則不現。諸天百千供養佛影,影亦說法。(T#643, 15:681a27-b4)
[Buddha Śākyamuni’s] whole body penetrated the rock; and just as in a bright mirror a man can see the image of his face, so the Nāgas all saw the Buddha within the rock while radiantly manifesting Himself on its exterior... The Tathāgata sat cross-legged within the rock wall, while everyone watched; although only those who looked from afar could see Him, for close by He was invisible. The various gods in their hundreds and thousands all adored the Buddha’s ‘shadow,’ and the ‘shadow’ also preached the Law. (translated by Soper 1959, 266)
During the fifth century, the cult of the Buddha’s shadow thrived in China; it is known that the influential monk Huiyuan 慧遠 (334-416) built a cave resembling the famous site at Nagarahāra:
遠乃背山臨流營築龕室。妙算畫工淡彩圖寫。色疑積空望似烟霧。暉相炳噯若隱而顯。(T#2059, 50:358b12-14)
[Hui]yuan then built a grotto, backing against the mountain and overlooking the stream, and worked out a wonderful scheme by which a painter drew [the form] in pale pigments. The colors one might have taken for layers of air; seen from a distance they were like mist, [from which] the glorious body-signs gleamed forth as if they were now hidden and now revealed. (translated by Soper 1959, 32-33)

By the middle of the seventh century, when Xuanzang玄奘 (602-664) visited the western regions, the Buddha’s shadow cave was in decline, and it was hard to behold the Enlightened One’s countenance any longer. Only those with ardent belief were able to observe it, as Xuanzang reports:

昔有佛影煥若真容。相好具足儼然如在。近代已來人不遍覩。縱有所見髣髴而已。至誠祈請有冥感者。乃暫明視尚不能久。(T#2087, 51:879a2-5)
In the old days there was a Buddha’s ‘shadow’ here, as luminous as if it had been the True Countenance. The major and minor attributes were complete, and as awe-inspiring as if He had been really present. In recent times people have not seen it so fully; at best what was visible was only a summary likeness. For those who pray with complete faith there is a mysterious manifestation, which may be glimpsed clearly for a while, but does not last long. (translated by Soper 1959, 267)

Xuanzang distinguishes clearly between the Buddha’s true countenance (zhenrong 真容) that is characterized by completeness of the major and minor bodily attributes and a mere likeness (fangfu 髣髴 or 仿彿) of that countenance that is simply not able to benefit the believers as the true countenance once did. The decline of the shadow’s salvific powers, undoubtedly, reflects the belief in the decline of dharma (mofa 末法) that was widespread in Xuanzang’s time. The theory of the decline of dharma speaks of three phases: At first there is the most salvivic era of the true dharma (zhenfa 真法), which is followed by an era of mere semblance of the dharma (xiangfa 像法), before the final period of decline, when even the semblance of the Buddha’s teaching is lost.  In the seventh century, it was generally held that the period of the decline of dharma had already set in; therefore, Xuanzang narrates that the true countenance of the Buddha was virtually lost, only manifesting itself incompletely and temporarily as a response to the most ardent prayers. If it was to appear at all, it would merely resemble the true countenance, but most of the time it remained indistinguishable or invisible.

The tree varieties of Buddha icons discussed so far—the first image of Udāyana, the divine images from the time of Aśoka, and the true countenance of the Buddha’s shadow—all show a remarkable relationship to Buddha relics and their worship. While the divine images are able to produce relics themselves, the other two are outstanding examples of Buddha images that join the ranks of contact relics, or paribhoga relics (Strong 2004, 20, note 51). This is obvious for the Buddha’s shadow at Nagarahāra, as we have seen; but also for the Udyāna image, there is a transmission from the Buddha who actually touches the image and pats its head after his descent from the heavens (Soper 1959, 261, and Carter 1990, 7). Relics of the Buddha are generally classified into bodily (śārīrika) relics and relics of use or touch (paribhoga), which had some kind of direct physical connection with the Buddha, and a third category of commemorative (uddesika) relics, into which fall Buddha images more generally (Strong 2004, 20-21).

Mental images

We have to keep this classification in mind if we want to understand the most astonishing of all true images of the Buddha, the ones that are only created mentally by a process of ongoing visualizations (guanxiang觀像). The Sutra on the Ocean-like Samādhi of Buddha Contemplation, which strongly propagated the cult of the Buddha’s true countenance at Nagarahāra, is probably the oldest scripture in a group of six visualization (guan 觀) sutras that came to be known in China during the fifth century and exerted considerable influence thereafter.

Based on references in the Chu sanzang jiji (出三藏記集 T#2145, 55:11c11-24) and the Biographies of Eminent Monks (T#2059, 50:335c11), the translation of the Sutra on the Ocean-like Samādhi of Buddha Contemplation is traditionally attributed to Buddhabhadra 佛陀跋陀羅 (359-429), an Indian monk allegedly born in Kapilavastu in North India who came to China via the southern sea route. He is believed to have been highly knowledgeable about Central Asia, since he had studied in Kashmir or Gandhāra for several years. Upon arriving in Chang’an, he enjoyed close association with Kumārajīva and translated the Larger Sukhāvatīvyūhasūtra with Baoyun. In ca. 411 he was banished from Chang’an, but was subsequently invited to Lushan by Huiyuan. At Huiyuan’s urgent request, he translated the Dharmatrātadhyānasūtra 達摩多羅禪經 for which Huiyuan wrote a preface. Later, possibly after leaving Lushan sometime in 412, Buddhabhadra translated the Guanfo sanmeihai jing.

The attribution to Buddhabhadra was recently contested by Yamabe Nobuyoshi, who holds that the author or authors of the scripture are as yet unknown (Yamabe 1999, 296-297). A number of later works that Yamabe also lists (Yamabe 1999, 34-37) quote extensively from the Sutra on the Ocean-like samādhi of Buddha Contemplation, thus bearing testimony to the scripture’s growing influence. A long passage from chapter nine, “On the Visualization of the Image,” is found in the seventh-century work Essential Teachings of Scriptures (Zhujing yaoji 諸經要集 , T#2123, 54:1-194) by Daoshi 道世 (dates unknown). These quotations frame the discussion of paying respect to the Buddha, his image and superior monks, but not to secular rulers (Shinohara 2004, 184-189). Daoshi also compiled the encyclopedic anthology The Jade Forest in the Dharma Garden (Fayuan zhulin 法苑珠林, T #2122, 53:269-end). This work has a comparable discussion of how to pay respects to the Buddha, but in contrast to the Essential Teachings of Scriptures that quotes the visualization practices described in the Sutra on the Ocean-like Samādhi of Buddha Contemplation, The Jade Forest’s section on paying respects to the Buddha is dominated by miracle stories (Shinohara 2004, 197) which focus on divine images like those of the Aśoka type. While image worship is an important topic in Daoshi’s works, a vital distinction is made between the worship of material images and the contemplation of mental images.

The Sutra on the Ocean-like Samādhi of Buddha Contemplation promotes the visualization of mental images. These are created by means of contemplation in the practitioner’s mind, after he has observed, as a first step, a material image of the Buddha, until he is able to reproduce all the Buddha’s major and minor bodily attributes correctly and completely in his mind with his eyes closed. The mental image created is then examined step by step—visualizing the distinguishing bodily attributes of the Buddha from the uşņīşa at the top of his crown to the dharma wheel on the soles of his feet, downwards and upwards, as many times as possible. The contemplation proceeds from one image to thousands that fill all mental space in the ten directions—from sitting images to images that can stand up and walk around freely and, furthermore, talk to and instruct the practitioner. Such visualizations are normally hampered by the practitioner’s karmic obstructions and afflictions, which must be eliminated from time to time with purification and repentance rites. Near the end of the visualizations, all images suddenly disappear, and the practitioner’s insight into the emptiness of all phenomena causes the Buddha’s shadow, namely the real body of the Buddha, to appear. This means that despite all limits of space and time, the practitioner is actually transferred into the very presence of the Buddha himself, and receives a prophecy of his own future Buddhahood, with Buddha Śākyamuni reaching out with his right hand to touch the practitioner’s head.

As soon as this contemplation is successful, the practitioner encounters the Buddha in his real body. Although his image, be it material or mental, is not identical with the Buddha, the Buddha is at last found in his image, because his real body shares certain characteristics—the thirty-two bodily attributes of enlightenment—with the image. While the Sutra on the Ocean-like Samādhi of Buddha Contemplation does not elucidate any further on the detailed mechanism of this process, a comparative look at the Christian icon might clarify: In defending the use of icons, Patriarch Nikephoros (758-828) referred back to Aristotelian philosophy of constructing a formal relation between the archetype (Christ) and its image (Christ’s icon), which was seen as equal to the relationship between cause and effect. Even though archetype and image are different entities, they share a partial quality, namely likeness, which is given by the archetype to representation, and is represented in the icon. While the Aristotelian background is probably completely irrelevant for the mechanism described in the Sutra on the Ocean-like Samādhi of Buddha Contemplation, it is striking that the likeness that mediates between the icon and its archetype results in “a full visualization of that which has given itself to vision” (Barber 2002, 117). Furthermore, Nikephoros elucidated on the fact that “there is nothing of presence in the icon; it is a showing without representation,” but “it becomes the point of departure for the contemplation” (Barber 2002, 121). Likewise, material and mental images of the Buddha serve as a point of departure for his contemplation; but prior to the encounter with the Buddha’s real body, all the images that have been created mentally are abandoned and disappear because the Buddha is not present in these images. In his thirty-two bodily attributes, the Buddha has given himself to representation, rewarding the practitioner with his presence in a vision (jianfo 見佛), as soon as the process of correct visualization (guanfo 觀佛) is completed.

Versions of the Udyāna legend: iconodules and iconoclasts

The story about Udyāna manufacturing and presenting a first image to the Buddha is a key to the reconstruction of arguments that favor and refute images in the Buddhist context. The Sutra on the Ocean-like Samādhi of Buddha Contemplation, which strongly promoted the cult of the Buddha’s shadow and visualizations of mental images leading to a true vision of the Buddha, narrates a version of the Udyāna story that unmistakably advocates an iconophile viewpoint. This version by far exceeds the image-friendly stance taken by the Ekottarāgama, and suggests that the compiler(s) of the Ocean-like Samādhi Sutra must be ranked among the most fervent Buddhist iconodules. The following events are narrated when King Udyāna brought his golden image on the back of an elephant to the place of the Buddha’s descent:

爾時金像。從象上下猶如生佛。足步虛空足下雨華。亦放光明來迎世尊。時鑄金像。合掌叉手為佛作禮。爾時世尊。亦復長跪合掌向像。時虛空中百千化佛。亦皆合掌長跪向像。(T#643, 15: 678, b10-14)
At that time, the golden image dismounted the elephant like a living Buddha, and, walking through the air and raining flowers from under its feet, it welcomed the World-Honored One by emitting light. Then the golden image joined the palms together and raised the hands to pay obeisance to the Buddha. At that time, the World-Honored One also knelt in front of the image with palms joined. At the same time, the hundreds and thousands of manifested Buddhas in the air also joined their palms and knelt in front of the image.

Soper (Soper 1959, 260) had already noticed the audacity of having the Buddha himself kneeling in front of his own image. This rather outrageous gesture befits a scripture that fervently advocates images as a practical means for practitioners to get not only a glimpse of the Buddha’s real body, but also to get as close to one’s own enlightenment as possible by attaining a prophecy from the Buddha himself.

Surely, such a biased favoring of images did not remain uncontested. The voices of those who highlighted the limitations of images are found in another image-friendly scripture, the aforementioned Sutra Spoken by the Buddha on the Merits of Image Making in the Mahāyāna 佛說大乘造像功德經 (T#694, 16:790-796), translated by Khotanese Devaprajñā between 689 and 691. It is the latest sutra in the group of those that deal with the merits of image making, and it clearly reflects reservations about the flourishing image cult. Here, the story about king Udāyana presenting his sandalwood image to the Buddha takes an interesting turn:

爾時閻浮提內國王、大臣、并四部眾,皆以所持種種供具,供養於佛。時優陀延王頂戴佛像,并諸上供珍異之物,至如來所而以奉獻。佛身相好具足端嚴,在諸天中殊特明顯,譬如滿月離眾雲曀。所造之像而對於佛,猶如堆阜比須彌山不可為喻。但有螺髻及以玉毫少似於佛,而令四眾知是佛像。爾時優陀延王白佛言。世尊,如來過去於生死中為求菩提,行無量無邊難行苦行,獲是最上微妙之身無與等者。我所造像不似於佛。竊自思惟深為過咎。(T#694, 16:793a11-21)
At that time, all the kings, great ministers, and the four kinds of people of the countries in Jambudvipa all gave the various gifts they were holding in worship of the Buddha. Then King Udāyana, carrying the Buddha statue on [the top of] his head, and with rare things as offerings, approached the tathāgata seat and presented all this respectfully. The major and minor attributes of the Buddha’s body ever endowed with splendor were particularly evident among the gods, like the full moon leaving the obscuring multitude of clouds. When the man-made image was compared with the Buddha, it resembled a small hill that cannot be turned into Mount Sumeru. Although the spiral headdress and the urna were a little similar to that of the Buddha, the four kinds of people [present] were still induced to know that it is a [mere] image of the Buddha. At that time king Udāyana addressed the Buddha and said: “World-Honored One, when in the past the tathāgata sought  bodhi in the [endless circle of] life and death, he practiced immeasurable and limitless hardships and austerities, and obtained this unsurpassed, wondrous body that nothing can equal. The image that I made is not similar to the Buddha.” And he secretly thought to himself that he had made a grave mistake.

In this version of the story, Udāyana realizes that, despite all his efforts, the image that he made is only an image and “not similar to the Buddha.” Even though the thirty-two bodily attributes—here represented by the “spiral headdress and the ūrņā”—are correctly depicted, the gap between the image and genuine body of the Buddha in all his glory seems unbridgeable. The story ends with the Buddha comforting Udāyana, pointing out that he nevertheless attained considerable merit because he had already “made immeasurable beings achieve the benefit of deep faith.” Udāyana is assured that in the future believers will obtain great blessings from the image that he made.

By the end of the seventh century at the latest, the image cult in China had obviously already faced resistance and was in need of encouragement. The opposing argument found in the Sutra Spoken by the Buddha on the Merits of Image Making in the Mahāyāna is the same as the most widespread contention of the Byzantine iconoclasts: it is a man-made or manufactured image (suozao zhi xiang 所造之像), utterly unfit for comparison to the splendor (duanyan 端嚴) of the Buddha’s real body, and inappropriate to the point of becoming an example of “faults and errors” (guojiu 過咎). The vocabulary used by the Buddhist critique of image making is similar to that of the Christian iconoclasts who held that “an image is deemed to render an insufficient, if not deceptive, account of its subject” and might be defined as “false image,” since the truthfulness of any pictorial representation can be seriously questioned (Barber 2002, 56).

That said, what was the counterargument of the Buddhist “iconoclasts”? One answer is found in the writings of another defender of images, the seventh-century Chinese vinaya specialist Daoxuan 道宣(596-667). In 662, Daoxuan had, along with other leading clerics of his time, presented a memorial to the Tang throne to ward off the immediate danger of losing part of their autonomy after emperor Gaozong issued an order to reconsider the matter of monks and nuns paying respect to, and bowing in front of, rulers and parents. In the course of this debate on how to pay respect, Daoxuan had also attempted to discuss image worship, claiming that images as representations of the Buddha served as monastic objects of worship for the purpose of “paying respect to the Buddha” (Shinohara 2004, 202). Daoxuan had taken up the matter before in his earlier Vinaya Commentary, finally revised in 636, in which he included a chapter on “paying respect to monks and images” (sengxiang zhijing僧像致敬, T#1804, 40.131b-135a). In this chapter, he argued for the importance of upholding the monastic hierarchy, which is given expression in paying respect to elders, and to the Buddha himself, in the form of his image. Nevertheless, he had to admit that there are forms of worship that are still superior to image worship (Shinohara 2004, 199-201).

What are these superior forms of worship? In his Vinaya Commentary, Daoxuan quotes a story from the Commentary on the Great Perfection of Wisdom (Da zhidu lun大智度論, T#1509, 25:137a1-21). The story relates a kind of contest between the Buddha’s disciple Subhūti and the nun Utpalavarnā about how to best venerate the Buddha. When the Buddha was about to descend from the heaven of the thirty-three gods, Subhūti dwelt in contemplation in a rock cave, pondering:

佛常說,若人以智慧眼觀佛法身,則為見佛中最。(T#1509, 25:137a4-6)
The Buddha has always preached that contemplating the Buddha’s dharma body with the eye of wisdom is the ultimate among [all kinds of] Buddha visualizations.

Thereupon he decided not to go to the place where the Buddha was about to descend, which was awaited by the multitude of gods and men alike. In contrast, the nun Utpalavarnā had made all the necessary efforts and even used magic to get to the place of the Buddha’s descent and to be the first to see the Buddha’s body and pay homage to him. When she finally succeeded in doing so, the Buddha addressed her saying:

非汝初禮。須菩提最初禮我。所以者何。須菩提觀諸法空,是為見佛法身,得真供養,供養中最。非以致敬生身為供養也。以是故言:須菩提常行空三昧,與般若波羅蜜空相相應。以是故佛命令說般若波羅蜜。(T#1509, 25:137a16-21)
You are not the first to pay homage; Subhūti was the first to pay homage to me. How is that? Subhūti contemplated the emptiness of all phenomena, which is seeing the Buddha’s dharma body, and achieved the true worship, the utmost among [all kinds of] worship. He did not consider paying respect to the living body as veneration; therefore I say: In persistently practicing the samādhi of emptiness, Subhūti responded to the attributes of emptiness of prajñāpāramitā. For that reason the Buddha orders the teaching of the prajñāpāramitā.

As this story shows, even though venerating the Buddha’s body and his image is advisable, the highest form of veneration is the contemplation of emptiness of all phenomena (guan zhufa kong 觀諸法空), which is equivalent to a vision of the Buddha’s real body, his dharma body (jian fo fashen 見佛法身). By contemplating emptiness, Subhūti experiences prajñāpāramitā, the perfection of wisdom, which is highly recommended by the Buddha. Could it be then that advocates of the doctrine of emptiness, or even a cult of prajñāpāramitā, favored a more iconoclastic point of view?

IV.  Aniconism in China

An eponymous body of scriptures, known as the Perfection of Wisdom Sutras, canonize the teachings on emptiness and prajñāparāmitā that played a significant role in the intellectual life of sixth century China. As will be shown, the impact of these scriptures is clearly seen in the second aniconic phase in China when selected passages of these sutras appear as rock carvings in the mountainous landscape of Shandong province. Teachings on emptiness also played a significant role in theoretical discussions about iconoclasm.

The threat of iconoclasm

In China, iconoclastic movements and persecutions of the sangha threatened Buddhist belief three times. The last and most severe of these persecutions in 845[5] ended its golden age irretrievably. Imperial attacks on the religion included the demolition of images, the closing or destruction of monasteries, the confiscation of monastic land, and the forceful return of monks and nuns into lay life. The economic reasons behind this are well known: Monasteries had accumulated incredible wealth and thereby gained political influence; clerics did not pay any taxes and were not subject to secular law, and the sheer material value of Buddhist statues provided an incentive to melt them down and make them into coins or weapons. But apart from political and practical considerations, educated circles also discussed iconoclasm in theory; for example, immediately before the second Buddhist persecution was decreed at the court of Emperor Wu of the Northern Zhou dynasty (556-581).

Emperor Wu had, for several consecutive years, held debates on the issue of the superiority of the three teachings, namely Confucianism, Taoism, and Buddhism. In 577, when he had just defeated the northern Qi-Dynasty, he summoned 500 eminent monks to his court and declared that monasteries, scriptures, and images were about to be destroyed. Buddhist sources such as the tenth fascicle of Daoxuan’s Guang hongming ji 廣弘明集 (T# 2103.52:153a28-154a9) and the Biographies of Eminent Monks report that, in one last debate, only the brave monk Huiyuan [淨影]慧遠 (523-592) defied the emperor who was trying to beat the Buddhists at their own game. In arguing for the uselessness of Buddha images, the emperor referred to the Buddhist Mahāyāna teaching of emptiness of all phenomena, and said:

且自真佛無像,則在太虛,遙敬表心。 (T#2060, 50: 490a28-29)
The true Buddha is beyond representation, for He resides in the Great Void; the distant reverence [that we feel for Him should be] revealed in our hearts. (translated by Soper 1959, 138)

Thereupon Huiyuan explained:

詔云。真佛無像。信如誠旨。但耳目生靈,賴經聞佛籍像表真。若使廢之,無以興敬。帝曰虛空真佛。咸自知之。未假經像。(T#2060, 50:490b10-13)
His Majesty has proclaimed that the true Buddha is beyond representation; and truly these are like the words of a god. But the ear and eye create the spirit; and it is by relying on the scriptures, or by listening to a Buddha, or with the aid of images, that the truth is made manifest. If they are now to be done away with, there will be no way to arouse devotion.” The Emperor replied: “The true Buddha of the Void is known naturally by all men, with no borrowing from scriptures or images. (translated by Soper 1959, 119)

Even though Huiyuan managed to rebut the last argument, his plea was of no avail, and the persecution was initiated. Only when the emperor suddenly died one year later was the Buddhist teaching rehabilitated.

With regard to the status of the images, Huiyuan holds that they manifest the truth just like the scriptures or words of the Buddha; nevertheless, he also admits that “the true Buddha is beyond representation.” The crucial point here is that images and words are manifestations of the Buddha’s truth in this world, while the Buddha himself is identical to the great void or emptiness of all phenomena. This seeming contradiction is resolved in the doctrine of the twofold truth (erdi 二諦), according to which all phenomena simultaneously participate in an ultimate truth and a conventional truth. While the latter is reality as experienced by the unenlightened, the ultimate truth equates emptiness and the transcendence of all language and reasoning (Muller 2009, article “twofold truth” in DDB). When the emperor used the doctrine of emptiness of the ultimate truth to validate his annihilation of Buddhist images and scriptures, he did not understand that ultimate and conventional truths are mutually dependant on each other: without affirming a “conventional truth,” there is no “ultimate truth” that can be claimed as superior. In his defiance of Buddha’s images and scriptures, Huiyuan reminds him of this fact.

The discourse on true attributes: Rejecting the bodily attributes of the Buddha and their visualization

Mount Tie

Fig. 2: Giant rock scripture at Mount Tie in Zoucheng, Shandong province, China, with carved text passage expounding the path of the bodhisattva, dated 579 CE, stone, height 51.70 m, width 14 m, photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2008

When the Chinese turned to the aniconic in the second half of the sixth century, it was not to avoid representations of the Buddha in human form, as early Indian Buddhism did, while using his non-image to enhance and intensify the pictorial representation; instead, the Chinese refrained from any kind of pictorial representation and focused on the Buddha’s written words. Selected passages taken from the Buddha’s “golden words” were carved into the natural rock bed of hills and mountains in Shandong Province. In contrast to writing sutras on paper or cloth, carving them into cliffs beneath the sky, or into the rough walls of cave temples, was a monumental undertaking. The inscription of Mount Tie in Zoucheng, for example, covers 725 square meters, and its sheer size makes it impossible to read the whole text from a single standpoint, or even to see it as a whole (figure 2). The text of Mount Tie expounds on the path of the bodhisattva, the enlightened being who, according to the Mahāyāna ideal, practices compassion and wisdom and finally achieves Supreme Enlightenment in order to save all living beings. We are informed in a subsidiary inscription called “Stone Hymn” that the text was carved in stone to preserve it for eternity against the expected “inferno at the end of the eon,” namely the end of the Buddhist teaching and the world. Mount Tie’s carving was finished on September 23rd 579, according to the western calendar; about one year after the second persecution of Buddhism had come to an end with the sudden death of Emperor Zhou Wudi, as mentioned before. It may well be seen as a triumphant return to the spreading of Buddhist teachings.

Mount Tao

Fig. 3: Rubbing of the rock inscription at Mount Tao in Tengzhou, Shandong province, China, with invocations of prajñāpāramitā, Buddha Avalokiteśvara, and Buddha Amitābha, and remains of a colophon, rock inscription undated, probably second half of the sixth century, ink on paper, height 1.82 m, width 1.65 m, photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2007

To a remarkable extent, the stone-carved passages deal with definitions and praises of the perfection of wisdom, prajñā-pāramitā. An excerpt in ninety-eight characters from the Sutra on the Great Perfection of Wisdom spoken by Mañjuśrī (文殊師利所說摩訶般若波羅蜜; T #232, 8:726a-732c), on the nature of the prajñā-pāramitā was carved in six places, and another excerpt in fifty-four characters from the same sutra on the nature of Buddha contemplation in two more places. Sometimes, the prajñā-pāramitā is likened to a talisman, and its invocation becomes spell-like in character. The perfection of wisdom is all the more equated to salvific Buddha figures. At Mount Tao 陶山 in Tengzhou 滕州, prajñā-pāramitā was carved next to the names of the Buddhas Avalokiteśvara (Guanshiyin 觀世音佛) and Amitābha (Omituo 阿弥陁佛), not only giving it the same status enjoyed by these Buddhas, but also suggesting that it  was invoked orally in the same way these Buddhas usually are (figure 3).


Fig. 4: Giant Diamondsutra carved in Stone Sutra Valley at Mount Tai, Shandong Province, China, undated, probably second half of the sixth century, stone, height of carved surface 32 m, width 62,1 m, photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2006

In other carved sutra passages, prajñā-pāramitā is linked to the discussion of the real attributes (shixiang 實相) of the Buddha. At the largest inscription site in Shandong province, in Stone Sutra Valley at Mount Tai 泰山, one third of the Diamondsutra (Vajracchedika prajñāpāramitāsūtra ) was carved in columns of irregular length, covering a surface of about 1800 square meters (figure 4). As one of the most important Buddhist texts in all of East Asia, this sutra from the group of Prajñāpāramitā scriptures has been essential for establishing the teachings of the Mahāyāna in China.

The text itself is in the form of a dialogue between the Buddha and his disciple Subhūti, whom we have met before in the veneration contest with the nun Utpalavarnā. At one point, Subhūti affirms that whoever hears this sutra being preached by someone with a pure heart will bring forth the real attributes, and he explains what they are:

世尊! 是實相者,則是非相,是故如來說名實相。
World-Honored One! These real attributes are non-attributes; for that reason the Tathāgatha says they are called Real Attributes.

Since the real attributes of the Buddha are non-attributes, thus no different from emptiness, certain conclusions about the Buddha’s body and its perception are unavoidable. The first concerns the visibility and representability of the Buddha in his physical body:

須菩提! 於意云何? 可以三十二相見如來不? 不也,世尊! 不可以三十二相得見如來。何以故?如來說三十二相,即是非相,是名三十二相。(T#235, 8:750a20-23)
The Buddha asked: Subhūti, what do you think? Can the Tathāgatha be seen in his 32 attributes? [Subhūti answered:] No, World Honored One! The Tathāgatha cannot be seen in his 32 attributes. Why is that? Because the 32 attributes that were taught by the Tathāgatha are actually non-attributes. This is called ‘32 attributes’.

The second treats the possibility of contemplation of the Buddha:

須菩提! 於意云何? 可以三十二相觀如來不? [。。。] 須菩提白佛言:世尊! 如我解佛所說義,不應以三十二相觀如來。爾時,世尊而說偈言:若以色見我,以音聲求我,是人行邪道,不能見如來。 (T#235, 8:752a11-18)
[The Buddha furthermore asked:] Subhūti, what do you think? Can the Tathāgatha be contemplated in his 32 attributes? [And Subhūti answered:] World Honored One! According to my understanding of the meaning expounded by the Buddha, the Tathāgatha should not be contemplated in his 32 attributes.At that time the World Honored One spoke the following verse: If someone saw me in form, or sought for me in sounds, such a person would walk the wrong way and could not see the Tathāgatha.

The Diamondsutra thus mentions the issue of the Buddha’s bodily attributes two times. The first time, the visibility (jian 見) of the Buddha in his thirty-two attributes is rejected; this rejection includes the Buddha’s anthropomorphic image. The second time, even the contemplation (guan 觀) on the thirty-two bodily attributes is denied or dismissed, indicating a rejection of the Buddha’s mental images as well. As the verse sums up, either path would not lead to a vision of the Buddha or into the Buddha’s presence. At Mount Tai, the selection of the Diamondsutra seems to suggest a refusal of the Buddha’s image that is then given form in the giant rock carving.

Mount Ge

Fig. 5: Rock cut passage from the Vimalakīrtinirdeśasūtra at Mount Ge in Zoucheng, Shandong province, China, dated 580, stone, height 20.88 m, width  8.36 m, photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2006

Apart from the scriptures of the prajñāpāramitā group, other Mahāyāna works expound the teachings of emptiness as well. From the body of texts most popular in China, a passage from the Vimalakīrtinirdeśasūtra, or the Sutra Spoken by Vimalakīrti (維摩詰所說經 T#475, 14:554c28 – 555a24) was chosen to be carved at Mount Ge 葛山in Zoucheng (figure 5). The inclining western slope of this low mountain bears an inscription in ten vertical columns of about forty characters each, on a surface of 175 square meters (figure 6). The text is still too large to be taken in as a whole, just like those at Mount Tie and Mount Tai. The passage itself is located near the end of the sutra, when the wise layman Vimalakīrti finally encounters the living Buddha and discusses how to contemplate the body of the Enlightened One.

Mount Ge reconstruction

Fig. 6: Virtual reconstruction of the sutra passage carved at Mount Ge, ink on paper rubbings of single characters pasted on virtual 3D model of the rock, damaged characters added from the printed Taishō edition of the Buddhist canon, processed by Ke Peng in 2008, courtesy of Heidelberger Akademie der Wissenschaften

In the carved text, the noble Vimalakīrti says the following about the contemplation of the Buddha’s body and its real attributes:

如自觀身實相。觀佛亦然。。。。不觀色不觀色如。不觀色性。。。。 非四大起。同於虛空。。。。不可以智知。不可以識識。無晦無明無名無相。無強無弱非淨非穢。不在方不離方。非有為非無為。無示無說。。。。非有相非無相。同[4]真際等法性。。。。 非大非小。非見非聞非覺非知。。。。無已有無當有無今有。不可以一切言說分別顯示。(T# 475, 14:554c29-555a23)
As if contemplating the real characteristics of my own body—so do I view the Buddha. ... I neither view him as form, nor view him as the suchness of form, nor view him as the nature of form. ... He does not arise from the four great elements and is identical to space. ... He cannot be understood with wisdom, nor can he be known by consciousness. He is without darkness (i.e., ignorance), without brightness (i.e., understanding), without name, and without characteristic. He is without strength, without weakness, and neither pure nor defiled. He does not occupy a region, nor does he transcend the regions. He is neither conditioned nor unconditioned. He is without manifesting and without explaining. ...he neither has characteristics nor is without characteristics. He is identical to the true limit and equivalent to the Dharma-nature. ... He is neither great nor small. He is neither vision, nor hearing, nor perceiving, nor knowing; ... without past, without future, and without present. He cannot be discriminated or manifested using any verbal explanations at all. (translated by McRae 2004, 185-186)

And, as one is tempted to add, “he cannot be made manifest by any image at all.” The real attributes of the Buddha, promoted in the Vimalakīrtinirdeśasūtra, are definitely different from the thirty-two bodily attributes of the Enlightened One that characterize his image, but also different from anything “made manifest by any discursive analyses.” In this respect, the rock carvings at Shandong are to be understood at least as aniconic, maybe as iconoclastic. After all, even language and discriminative thinking is rejected; denying the message of the rock-carved words as well, leaving only the mere efficacy of the Buddha’s golden words.

Buddha names: The Buddha’s aniconic presence

Mount Hongding

Fig. 7: Rubbing of “Buddha King of Great Emptiness” carving at Mount Hongding in Dongping county, Shandong province, China, rock carving around 564 CE, ink on paper, height 9.20 m, width 3.40 m, photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2006

Apart from selecting passages from the Buddhist canon, the aniconic rock carvings in Shandong province portray the Buddha’s presence in another way: Where anthropomorphic images of the Buddha such as those on the outer walls of many cave-temples would be expected, only Buddha names are carved on the cliffs. Among the giant Buddha names carved into the opposing slopes of Mount Hongding洪頂山in Dongping county東平縣, the one that reads “Buddha King of Great Emptiness” is the largest one, measuring approximately nine meters in height (figure 7). This particular Buddha, whose name is invoked not only on the cliff of Mount Hongding, but in eight more places on the former territory of the Northern Qi, is mentioned nowhere in the Buddhist canon. The “Buddha King of Great Emptiness” is a creation of the Buddhist circles responsible for the art of rock carvings in the sixth century, and is to be understood as an exaltation of the concept of great emptiness. In the commentary literature of this time, great emptiness is seen equaling the perfection of wisdom, which in turn is called mother of all Buddhas, or Buddhahood itself. Consequently, only prajñāpāramitā, the highly abstract, deep and mysterious concept of the perfection of wisdom, is able to compensate for the insufficiency of images and words.

V. Conclusions

While a comprehensive comparative study of Christian and Buddhist icons is still lacking, in recent years awareness of the affinity of Buddhist and Christian medieval culture has grown. While in the field of Byzantine studies research in image discourse is already longstanding, the topic was touched upon only occasionally in Chinese studies. This paper assembles pictorial and textual sources relevant to image theory in Chinese Buddhism, and is an attempt to restructure and reread them along lines set by analyses of Byzantine image discourse.

A number of parallels become apparent when looking at the practice of an image cult in both cultural spheres: The key questions for any kind of representation in a religious context are the fundamental invisibility of the divine and the problem of truthfulness of the medium chosen to represent it. Anthropomorphic representations of Christ and Buddha may or may not be adequate embodiments of the divine. In the case that they are held to be so, the Byzantine theologians, as well as the Buddhist clerics, follow the same legitimating strategies: The need for a first image is recognized, and a legend based on the idea of a portrayal of the historical Christ/Buddha is developed around it. Another possibility is offered in images of the true countenance, which are factually seen as relics of contact that recommend themselves for worship. As for the kinds of images that are not produced via bodily contact with Christ/Buddha, the problem of their manufactured nature has to be solved. One successful strategy consists of concealing their man-made nature and declaring them to be acheiropoietos, images“not made by hand.” This claim is usually substantiated by stories of unexpected discoveries found in the Chinese sphere in numerous examples of so-called Aśoka images. While theories of how exactly the relation between the archetype and the image is to be understood differs in Byzantine and China, a common feature is the belief that even though the images and the divine are essentially different, they basically posses the potential to evoke a vision of the real body of Christ/Buddha.

Arguments against images as adequate embodiments of the divine also seem to be similar. Their manufactured nature is the main critique in the version of the Udyāna legend presented in the Sutra Spoken by the Buddha on the Merits of Image Making in the Mahāyāna. There are probably many more arguments to be discovered in the Buddhist scriptures, and a lot of work still needs to be done to give a comprehensive account of image defenders and image critics in Chinese Buddhism, whom we better not call iconoclasts too hastily. While iconoclasm and persecution was a severe threat to Buddhist teachings in China, it is always portrayed as state-controlled demolition. The image discourse within the Buddhist community seems to have always been rather moderate, cultivating the view that images, as well as language, will always be expedient means to enlightenment, but should never be taken for the ultimate truth of Buddhahood itself.

Finally, another point of contact for image discourses in Byzantium and China is what Charles Barber describes as the “strong antithesis of word and image,” which in Byzantium was “typical of the rhetoric of second iconoclasm.” In his analysis of the five iconoclastic poems that were written close to the cross that was placed on the Chalke Gate in 815, Barber notices that the iconoclastic theologians who wrote the poems repeatedly connect the symbolic figure of the cross to the prophets’ verbal testimony in the Old Testament (Barber 2002, 93). Likewise, the golden words of the Buddha seem to have been the Chinese counteraction to Buddhist image veneration.

It seems as if a discussion of the thirty-two bodily attributes designating the Buddha as the Enlightened One according to the Mahāyānan teaching on emptiness initiated a second aniconic phase in Buddhism. In contrast to the first aniconic phase of early Indian Buddhism that realized the aniconic representation of the Buddha in a pictorial context, the second aniconic phase in China relied only on writing the Buddha’s golden words on natural rock surfaces. The sutra texts selected for carving clearly reflect the idea that the thirty-two bodily attributes of the Buddha are as empty as all phenomena, and therefore inappropriate for attaining a vision of the real Buddha. Instead, the subtle and wonderful perfection of wisdom is praised and invocated as the mother of all Buddhas. By simply carving various Buddha names, prajñāpāramitā, and the often evoked “Buddha King of Great Emptiness,” aniconic representations of the Buddha in China thus gained a new quality: Instead of shaping the rock into a human-shaped Buddha figure, the Buddha’s name is carved onto it in carefully crafted calligraphy. In the same way that anthropomorphic Buddha images have been used as an aid for contemplation and the generation of mental Buddha images, sutra passages and Buddha names were suitable for recitation and invocation, another kind of mental contemplation that would take the believer closer to the Enlightened One, not via the image, but via the word or language. Nevertheless, as the Chinese were well aware, seeing the Buddha in form, or seeking him in sounds, would mean taking the wrong path in one’s quest for awakening, as the rock cut Diamondsutra at Mount Tai lets the believers know. In this respect, and contrary to the above mentioned “strong antithesis of word and image” in Byzantium, the iconic and the aniconic have in China always been seen as two sides of the same coin.


List of Illustrations:

Fig. 1:  Buddha Amitābha with inscribed pedestal, dated 746. Grey limestone, height including pedestal 94 cm, width of pedestal 49,5 cm. Photograph by Jürgen Liepe. Courtesy of Staatliche Museen zu Berlin, Stiftung Preußischer Kulturbesitz, Museum für Asiatische Kunst, Ostasiatische Kunstsammlung.  

Fig. 2: Photograph of giant rock scripture at Mount Tie in Zoucheng, Shandong province, China, with carved text passage expounding the path of the bodhisattva, dated 579 CE. Stone, height 51.70 m, width 14 m. Photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2008

Fig. 3:  Rubbing of the rock inscription at Mount Tao in Tengzhou, Shandong province, China, with invocations of prajñāpāramitā, Buddha  Avalokiteśvara, and Buddha Amitābha, and remains of a colophon. Rock inscription undated, probably second half of sixth century. Ink on paper. Height 1.82 m, width 1.65 m. Photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2007.

Fig. 4:  Photograph of giant Diamondsutra carved in Stone Sutra Valley at Mount Tai, Shandong Province, China. Undated, probably second half of the sixth century. Stone, height of carved surface 32 m, width 62,1 m. Photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2006.

Fig. 5:  Photograph of rock cut passage from the Vimalakīrtinirdeśasūtra at Mount Ge in Zoucheng, Shandong province, China, dated 580. Stone, height 20.88 m, width  8.36 m. Photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften, 2006.

Fig. 6:  Virtual reconstruction of the sutra passage carved at Mount Ge: Ink on paper rubbings of single characters pasted on virtual 3D model of the rock; damaged characters added from the printed Taishō edition of the Buddhist canon. Processed by Ke Peng in 2008. Courtesy of Heidelberger Akademie der Wissenschaften.

Fig. 7:  Rubbing of “Buddha King of Great Emptiness” carving at Mount Hongding in Dongping county, Shandong province, China. Rock carving around 564 CE. Ink on paper, height 9.20 m, width 3.40 m. Photograph taken by and courtesy of Heidelberger Akademie der Wissenschaften. 2006.


Bachhofer, Ludwig. 1929. Die frühindische Plastik. 2 Vols. Pantheon, Firenze: Casa Editrice.

Charles Barber 2002. Figure and Likeness:  On the Limits of Representation in Byzantine Iconoclasm. Princeton: Princeton University Press.

Belting, Hans. 2004. Bild und Kult. Eine Geschichte des Bildes vor dem Zeitalter der Kunst. München: Verlag C.H. Beck.

Brinker, Helmut. 2001. “Auf Buddhas frühen Spuren in China” In Die Rückkehr des Buddha. Chinesische Skulpturen des 6. Jahrhunderts. Der Tempelfund von Qingzhou, edited by Lukas Nickel, 19-41. Zürich: Museum Rietberg.

Brown, Robert L. 1990. “God on earth. The Walking Buddha in the Art of South and Southeast Asia.” Artibus Asiae 50: 73-107.

——— 1998a. “Expected Miracles: The Unsurprisingly Miraculous Nature of Buddhist Images and Relics.” In Images, Miracles, and Authority in Asianedited by Richard H. Davis, 23–35. Boulder: Westview.

——— 1998b. “The Miraculous Buddha Image: Portrait, God, or Object?” In Images, Miracles, and Authority in Asian Religious Traditions, edited by Richard H. Davis, 37-54. Boulder: Westview

Carter, Martha L. 1990. The Mystery of the Udayana Buddha. (Supplemento n. 64 agli Annali – vol. 50 , fasc. 3) Napoli: Istituto Universitario Orientale

Coomaraswamy, Anda K. 1972. Elements of Buddhist Iconography. New Delhi: Munshiram Manoharlal. First published 1935 by Harvard University Press

Davis, Richard H.,ed. 1998.  Images, Miracles, and Authority in Asian Religious Traditions. Boulder: WestviewPress.

Dunhuang wenwu yanjiusuo 敦煌文物研究所 [Research Institute for Cultural Relics in Dunhuang], ed. 1981-87. Dunhuang Mogaoku:敦煌莫高窟 [Mogao Cave Temples of Dunhuang]. 5 vols. (Zhongguo shiku 中国石窟 [Cave temples in China]). Beijing: Wenwu chubanshe, .

Effenberger, Arne. 1998. “Vom Zeichen zum Abbild – Frühzeit christlicher Kunst.” In Byzanz. Die Macht der Bilder, edited by Michael Brandt and Arne Effenberger,14-39.  Hildesheim: Dom-Museum Hildesheim.

Fojiao dongzhuan gushi huajuan 佛教东传故事画卷 [Illustrated volume of stories about the transmission of the Buddhist teaching to the east]. 2000, edited by Sun Xiushen 孙修身. (Dunhuang shiku quanji 12 敦煌石窟全集 12, Dunhuang yanjiuyuan zhubian 敦煌研究院主编). Shanghai 上海: Shanghai renmin chubanshe 上海人民出版社, Shanghai shiji chuban jituan 上海世纪出版集团.

Granoff, Phyllis. 1998. “Divine Delicacies: Monks, Images and Miracles in the Contest between Jainism and Buddhism.” In Images, Miracles and Authority in Asian Religious Traditions, edited by Richard Davis, 55-97. Boulder: Westview Press..

——— and Shinohara Koichi, eds. 2004. Images in Asian Religions: Texts and Contexts. Asian Religions and Society Series. Toronto: UBC Press.

Huntington, Susan L. 1992. “Aniconism and the multivalence of emblems: Another look”. Ars Orientalis 22: 111-156.

Huntington, Susan L. 1990. “Early Buddhist Art and the Theory of Aniconism.” Art Journal 49, no. (4): 401-408.

Karow, Otto. 1989. Symbolik des Buddhismus. Tafelband. Symbolik der Religionen 22. Stuttgart: Anton Hiersemann.

Klimburg-Salter, Deborah E. 1995. Buddha in Indien. Die frühindische Skulptur von König Aśoka bis zur Guptazeit. Wien: Ausstellungskatalog des Kunsthistorischen Museums.

Knox, Robert. 1992. Amaravati. Buddhist Sculpture from the Great Stūpa. London: British Museum Press.

Krannich, Thorsten, Christof Schubert, and Claudia Sode, eds. 2002.  Die ikonoklastische Synode von Hiereia 754. Studien und Texte zu Antike und Christentum 15. Tübingen: Mohr Siebeck.

Kunst- und Ausstellungshalle der Bundesrepublik Deutschland GmbH, ed. 2008. Gandhara – Das buddhistische Erbe Pakistans. Legenden, Klöster und Paradiese.  Mainz: Verlag Philipp von Zabern.

Luczanits, Christian. 2008. “Der frühe Buddhismus und Gandhara.” In Gandhara – Das buddhistische Erbe Pakistans. Legenden, Klöster und Paradiese, edited by Kunst- und Ausstellungshalle der Bundesrepublik Deutschland GmbH 2008, 72-77. Mainz: Verlag Philipp von Zabern.

Muller, A. Charles, ed. Digital Dictionary of Buddhism. Accessed June 22, 2011.

Museum für Ostasiatische Kunst Berlin-Dahlem, ed. 1970. Ausgewählte Werke ostasiatischer Kunst. Berlin: Wissenschaftliche Bearbeitung Beatrix von Ragué.

McCallum, Donald F. 1998. “The Replication of Miraculous Images: The Zenkoji Amida and the Seiryoji Shaka.”  In Images, Miracles, and Authority in Asian Religious Traditions, edited by Richard H. Davis, 207-226. Boulder: Westview.

McRae, John R. 2004. “The Vimalakīrti Sutra. Translated from the Chinese (Taishō Volume 14, Number 475).” In The Sutra of Queen Śrīmālā of the Lion’s Roar and the  Vimalakīrti Sutra, 65-201.  Berkeley: Numata Center for Buddhist Translation and Research.

Nickel, Lukas, ed. 2001. Die Rückkehr des Buddha. Chinesische Skulpturen des 6. Jahrhunderts. Der Tempelfund von Qingzhou. Museum Rietberg Zürich. Ausstellungskatalog Altes Museum, Staatliche Museen zu Berlin, Museum Rietberg Zürich, Royal Academy of Arts. Zürich: Museum Rietberg.

Quagliotti, Anna Maria. 1998. Buddhapadas. An essay on the representations of the footprints of the Buddha with a descriptive catalogue of the Indian specimens from the 2nd century B.C. to the 4th century A.D. Kamakura: Institute of the Silk Road Studies.

Seckel, Dietrich. 1976.  Jenseits des Bildes – Anikonische Symbolik in der buddhistischen Kunst. Abhandlungen der Heidelberger Akademie der Wissenschaften: Philosophisch-historische Klasse, Jahrgang 1976, 2. Heidelberg: Abhandlung.

——— 2004. Before and Beyond the Image. Aniconic Symbolism in Buddhist Art, translated by Andreas Leisinger and edited by Helmut Brinker and John Rosenfield.  Supplementum 45. Zürich:  Artibus Asiae Publishers.

Sharma, R.C. 1984. Buddhist Art of Mathurā. New Delhi: Agam Kala Prakashan.

——— 1992.  Bharhut Sculptures. New Delhi: Shakti Malik Abhinav Publications.

Shinohara Koichi. 1998. “Changing Roles for Miraculous Images in Medieval Chinese Buddhism: A study of the Miracle Image Section in Daoxuan’s ‘Collected Records’.”  In Images, Miracles, and Authority in Asian Religious Traditions, edited by Richard H. Davis, 141-188. Boulder: Westview.

——— 2000. “The ‘Iconic’ and ‘Aniconic’ Buddha visualization in medieval Chinese Buddhism.” In Representation in religion: studies in honour of Moshe Barasch, edited by  Jan Assmann and Albert I. Baumgarten, 133-148. Leiden: Brill.

——— 2004. “Stories of Miraculous images and Paying Respect to the Three Jewels: A Discourse on Image Worship in Seventh-Century-China.” In Images in Asian Religions: Texts and Contexts. Asian Religions and Society Series, edited by Granoff, Phyllis and Shinohara Koichi,180-222. Toronto: UBC Press.

Snellgrove, David L. 1978. The Image of the Buddha. Paris: Unesco.

Soper, Alexander Coburn. 1959. Literary Evidence for Early Buddhist Art in China. Supplement 19. Zurich: Artibus Asiae.

Speck, Paul. 1998. “Bilder und Bilderstreit.” In Byzanz. Die Macht der Bilder Katalog zur Ausstellung im Dom-Museum Hildesheim, edited by Michael Brandt and Arne Effenberger, 55-67. Hildesheim: Dom-Museum Hildesheim.

Strong, John S. 2004. Relics of the Buddha. Buddhisms: A Princeton University Press series, edited by Stephen F. Teiser. Princeton: Princeton University Press.

Taddei, Maurizio. 1995. “Was bedeutete der Buddhismus für die frühe indische Kunst?” In Buddha in Indien. Die frühindische Skulptur von König Aśoka bis zur Guptazeit, edited by  Deborah E. Klimburg-Salter, 41-9. Wien: Ausstellungskatalog des Kunsthistorischen Museums.

Wenzel, Claudia. 2007. “Anikonik im chinesischen Mahāyāna-Buddhismus: Die Wahren Merkmale des Buddha.” In Weltbild – Bildwelt Ergebnisse und Beiträge des Internationalen Symposiums der Hermann und Marianne Straniak-Stiftung, Weingarten, edited by Walter Schweidler, 271-292. Sankt Augustin: Academia Verlag.

Wong, Dorothy C. 2004. Chinese Steles. Pre-Buddhist and Buddhist use of a symbolic form. Honolulu: University of Hawai’i Press.

Wu Hung. 1986. “Buddhist Elements in Early Chinese Art (2nd and 3rd Centuries A.D.).” Artibus Asiae XLVII 3 (4): 263-352.

Yamabe Nobuyoshi. 1999. The Sūtra on the Ocean-Like Samādhi of the Visualization of the Buddha: The interfusion of the Chinese and Indian Cultures in Central Asia as Reflected in a Fifth Century Apocryphal Sūtra. UMI Microform 9930977, Dissertation, Yale University.

[1] The rigid parallelism in the inscription notwithstanding, the last two characters are as they are rendered here. One may have expected to read 施其 instead.

[2]  Unless otherwise indicated, all translations are by the author.

[3] Like the four-sided stele dedicated by an yi  society led by Li Sengzhi and Wang A’quan (Li Sengzhi, Wang A’quan heyi zao simian xiangbei李僧智王阿全何邑造四面像碑), dated 520 in the Victoria and Albert Museum in London, sandstone, height 1.73 m, that has: “Since the divine doctrine is of subtle transcendence, it cannot truly be manifested but that words and images are the only means to comprehend this doctrine. Since perfect knowledge is profoundly deep, it cannot be truly fathomed but that representational figures are the only means of displaying the glorious signs [of buddhahood].” Published in Wong 2004, 77-82, inscription translated by Perceval W. Yetts. The George Eumorfopoulos Collection: Catalogue of the Chinese and Corean Bronzes, Sculpture, Jades, Jewellery and Miscellaneous Objects. Vol. 3, Buddhist Sculpture, pp. 43-50. pls. 8-13. London: Benn Ltd., 1932.

[4]  Or, as Brown 1998, 52, note 10, put it: “... the argument for an aniconic period of art would not be in terms of the symbols being the Buddha (such as the Bodhi Tree being the Buddha), nor of the symbol “replacing the Buddha,” but that the symbol indicated his presence in particular contexts.”

[5] Coinciding with the second iconoclastic phase in Byzantium from 815 to 843.