A Cambodian masterpiece restored
Recent study and conservation of the Kimbell’s 7th-century Cambodian stone sculpture of the Hindu deity Harihara has renewed appreciation of its spectacular quality, and ended some doubts about its authenticity, writes Jennifer Casler Price.
Monday, 24th September 2007
Asian art comprises more than a third of the Kimbell Art Museum’s collection. It encompasses India, south-east Asia, the Himalayas, China, Korea and Japan, with works spanning a period of nearly 2,000 years. Asian art was not collected by the museum’s founders, Kay and Velma Kimbell, but was among the many areas of non-European art, ‘derived from any and all periods in man’s history’, sought out by the museum’s first director, Richard F. Brown.1 By the time the Kimbell opened in 1972, the collection included some 40 Asian works; since that time the Kimbell has continued to acquire works of rare quality and importance in this area.
Among the Asian works, the museum is fortunate to have a small but superlative collection of ancient Buddhist and Hindu sculpture from India, Nepal and south-east Asia that consists of several celebrated pieces. The high standard of aesthetic excellence in this area was set early on by Brown with two key acquisitions, both from Thailand: a late-8th-century pre-Angkor Bodhisattva Maitreya from Prakhon Chai; and an Angkor period Buddha Enthroned, c. 1180-1220, from Chaiyaphun province, acquired in 1965 and 1966 respectively. Brown also acquired a very early 7th-century Nepalese Standing Buddha Shakyamuni in 1979. Following Brown’s tenure, three important works were added in the 1980s: an Indian Gupta period Four-Armed Ganesha, 5th-6th century, in 1981; an Indian Kushan period Seated Buddha with Two Attendants, dated by inscription to ad 82, in 1986; and a Cambodian Harihara, from Kompong Cham, c. ad 675–700, in 1988, considered by many scholars to be one of the finest examples of pre-Angkor sculpture ever produced. More recently, an impressive Gandharan Standing Bodhisattva, 2nd–3rd century, was acquired in 1997. These works form the core of the museum’s Asian collection (Fig. 1).
Since its acquisition in 1988, the Kimbell Harihara in the style of Prasat Andet has been regarded as one of the finest works in the Asian collection Figs. 4 and 9).2 It was executed in the late 7th century, an artistic high point for the early Khmer kingdoms of Cambodia. The Khmers dominated Cambodia, as well as large areas of Vietnam, Thailand and Laos, from the 6th to the 14th centuries. Before the emergence in the early 9th century of the powerful Angkor kingdom, this region was divided into a number of early Khmer cultures. Contact with India through maritime trade early in the first millennium had already seen the introduction of Buddhism and Hinduism, together with associated sculptural and architectural styles.
By the 6th century, although this Indian heritage is still evident, Khmer sculpture clearly demonstrated the naturalism and mastery of sculpture in the round that was to distinguish future Cambodian art.
Pre-Angkor (6th to 8th century) sculptures already possessed many artistic qualities that differentiated Cambodian works from Indian models: sculptures were carved in the round as opposed to high relief, and were not adorned with elaborate jewellery or clothing; rather, decoration was limited to hairstyles, mitres and the fashion in which their simple, short garments were tied. In addition, a greater sense of naturalism was achieved by highly polishing the stone, which brought the flesh to life. Well modelled with smooth-flowing transitions between body masses, pre-Angkor sculptures possess an aura of elegance as well as power, and provide the stylistic formula for Cambodian sculpture for the next three centuries. The Kimbell image of the supreme Hindu deity Harihara, with its delicately modelled musculature, supple articulation of limbs and slender torso draped in a simple sampot, is characteristic of the finest pre-Angkor style.
The cult of Harihara, which was of great importance in early Cambodia, combined the potency of two of the most powerful Hindu gods, Shiva and Vishnu. Cambodia appears to have been dominated first by the cult of Vishnu until the late 7th century, when power shifted to the north and the majority of believers turned to the worship of Shiva.3 The ensuing differences in religious focus most likely stimulated the proliferation of images of Harihara, the conjoined deity that could serve as a compromise between the rival cults.4 In sculptural form, the deity is split vertically, with Vishnu (Hari), the benevolent preserver of the world, on the proper left and Shiva (Hara), the cosmic force of destruction and fertility, on the proper right. The embodiment of the two divine principles is physically indicated by the treatment of the headdress, which is split down the centre. The left half, which is smooth, depicts Vishnu’s tall, undecorated mitre (kiritamukuta). The right half is carved with loops that suggest Shiva’s ascetic chignon of matted and twisted locks (jatamukuta) and one half of Shiva’s vertical third eye is carved at the centre of the forehead. Three auspicious beauty folds circle the neck. The four arms, now broken above the elbow, would have held the appropriate identifying attributes. In the absence of these, the deity is identifiable only from its head details. Damage on the back of the head, near the top, indicates where a supporting arch was originally attached (see Figure 9).5
The figure wears a traditional Khmer sampot, a long rectangular cloth that could be tied and draped in various ways. It is tied in front, with one of the ends passing through the legs and fastened at the back, where it forms a small fan-shape above the waistline.6 The other side is tucked into the waist at the left hip as represented by a small oval bulge, and forms a pocket on the left thigh before it is secured at the waist below the navel, with its pleated ends falling gracefully on the right side.7 The sampot is adorned with a belt carved to imitate a linked chain fastened by a decorative clasp. Such belts fashioned out of gold exist in several collections of ancient Khmer jewellery.8 This ornamental belt had no function other than visual confirmation of wealth and power.9
This majestic sculpture displays naturalistic facial features – almond-shaped eyes, delicately traced brows, high cheekbones and subtly moulded lips with a curling moustache – that have the quality of expressive portraiture. The carving of the body, particularly the chest and back, is distinguished by its sensitive treatment of bone structure and musculature, enhancing the sense of a compact, powerful human figure. This individualised treatment may represent the royal patron who commissioned the sculpture, in the hope of identifying himself with the deity. Khmer monarchs chose posthumous names that incorporated the name of the deity with whom they wished to be identified, but in the absence of an inscription, the intention here can only be surmised.
The Kimbell Harihara has been described as one of the most beautiful and exceptional examples of pre-Angkor-period sculpture to survive. When the museum acquired the Harihara in 1988, Martin Lerner, then curator of south-east Asian art at the Metropolitan Museum of Art, wrote, ‘It is one of the most glorious and beautiful Asian sculptures I have ever seen’.10 Despite its evident quality, rarity and historical importance, however, some scholars of Cambodian sculpture have questioned its authenticity. Soon after its publication, the sculpture was condemned by the famous French scholar Jean Boisselier, who declared it a copy of the well-known Prasat Andet Harihara in the National Museum of Cambodia, Phnom Penh Fig. 3).11 Even though a number of connoisseurs accepted the Kimbell Harihara as a superb, genuine work of art, some art historians followed Professor Boisselier’s opinion in considering it a modern forgery.
In 2003, in an effort to establish more objective criteria for determining the authenticity of ancient Cambodian sandstone sculptures, Pieter Meyers, senior research chemist at the Los Angeles County Museum of Art, undertook a comprehensive research study of a broad range of this material. His aim was to identify technical characteristics of the types of stone employed, the patterns of weathering and the methods of manufacture that distinguish ancient works from modern forgeries. As no tests are known that allow the determination of the date of the carving of a stone sculpture, any conclusion on authenticity must be based on changes in and on the surface of the stone.12 Sculpture made 800 to 1,400 years ago has been subjected to a long-term natural erosion process during burial, or while being exposed to the atmosphere, and possibly also when in use as a spiritual or decorative object.13 Some of the changes to the stone surface can be so characteristic that they allow reasonably secure conclusions to be made about the authenticity of a sculpture, since the artificial ageing procedures to which modern forgeries have been exposed do not result in the surface alterations typical of authentic artifacts.14
The Kimbell Harihara was among the group of works examined in detail.15 This investigation revealed a number of specific physical properties, not visible to the naked eye, that confirmed that the Harihara had been subjected to a long-term natural weathering process, and that therefore the object could not be a modern forgery. Under stereoscopic examination of the stone’s surface, Meyers observed the following characteristics. First, there is a pronounced preferential surface corrosion. Once highly polished, the surface now exhibits significant ‘weathering’, exposing rounded quartz particles – the major constituent in all sandstones. The slow erosion process has resulted in a ‘swirl’ pattern on the chest, caused by alternating areas of less and more extensive erosion. Second, there is the presence of soil accretions with root fragments and root marks. Despite previous cleaning, dense, red clay soil accretions are still present in many micro-crevices in the stone. Root fragments indicate prolonged exposure in or on the ground, and root marks – the impression left on the surface of the stone by decomposed roots – are the result of a long-term process (Fig. 6). Third, there is the presence of a number of black ‘manganese oxide’ spots – small black deposits of non-crystalline manganese oxide. Their exact composition is unknown, but they are thought to be the result of microbiological action, and their presence indicates prolonged burial. All of the technical characteristics observed by Meyers are entirely consistent with the proposed date of manufacture in the late 7th to early 8th century. No evidence was found that would suggest a more recent date.16
In preparation for Meyers’s examination, an additional fragment of the left leg of the Harihara came to light in storage at the Kimbell (Fig. 7). The fragment had been stored at the museum since the sculpture’s acquisition in 1988 but had never been published with the object. In fact, no photographs could be found showing the leg fragment in place, but at least one museum document lists its existence.17 Meyers examined the fragment as well and determined that it clearly belonged to the sculpture: its shape, surface appearance and microstructure and soil accretions all matched those of the sculpture. The break surfaces, for the most part, formed a perfect match, except that small fragments had broken off, leaving voids when the fragment was put in place (Fig. 8). Because of the texture and accretions on the break surface, the break was determined to be old and almost certainly occurred in antiquity.
Once it was confirmed that the leg fragment belonged to the Kimbell Harihara, it was decided that it should be reattached. John Hirx, head objects conservator at lacma, expertly carried out the restoration and subsequent conservation work.18 First the leg fragment was reattached. This involved drilling both the fragment and upper left thigh with a water-cooled core drill bit in order to pin the pieces together. The two pieces were joined with an extended threaded nut adhered into the hole of the upper thigh and a long, threaded bolt that passes through the leg fragment into the threaded nut, securing the thigh fragment in place. After reattaching the fragment, the small losses to the break edge at the point of reattachment were compensated with pigmented resin.
The restoration of the leg provided further stylistic support for the sculpture’s authenticity. The now-complete bottom edge of the short skirt’s drapery folds disproves an argument made by some sceptics that a supposed modern forger did not understand how to finish the drapery and had therefore left the leg broken at mid-thigh.
In addition to reattaching the leg fragment, the issue of previous restoration work was addressed during treatment at lacma. Prior to being acquired by the Kimbell, the head had broken off at the neck and had been reattached with the line of restoration clearly visible – this was left untouched. Some areas with losses had been restored and filled with pigmented resins to match the surrounding surfaces. Most notably, the left nipple and a circular area above the left nipple had been filled and the nipple had been unsatisfactorily reconstructed. The compensation was removed and the small losses on the left nipple were minimally restored (Fig. 6). A line of infill applied to a crack appearing across the middle of the back at the waist was evidence that someone had tried at some point to cut the sculpture in half. The cut grazed the top of the sampot’s fanned bow, which had again been compensated. This was cleaned to remove the compensation, which turned out to be negligible. Areas on the back of the two upper arms and shoulders and scratches on the buttocks had also been infilled. The cleaning of these old restorations revealed gouges created when the sculpture was dragged along the ground in antiquity. These were left unfilled. After removing of all the previous compensations, the work’s entire surface was given a light cleaning, and is now much truer to the sculpture’s original appearance.
With the technical analysis, restoration and conservation work complete, the Harihara’s original attribution to late-7th or early-8th- century Cambodia is now greatly supported by scientific data. The added length of the left leg gives the sculpture a greater sense of contraposto and movement. Reaffirmed as one of the most outstanding examples of pre-Angkor sculpture, the Kimbell Harihara once again takes its place as the centrepiece of the museum’s Asian collection.
Jennifer Casler Price is Curator for Asian and Non-European Art at the Kimbell Art Museum.
LATEST NEWS & COMMMENT
Around the Galleries
Brussels plays host to a trio of outstanding fairs at the Place du Grand Sablon in early June, and the ever popular Carré Rive Gauche – now in its 36th year – returns to the Left Bank in Paris.
Architecture
The work of John Nash has often been overshadowed by that of his contemporary, John Soane. But his pragmatism, as well as his experiments with the picturesque, make him one of the most significant of all British architects.
Editors' Letter: Public outrage
Apollo is published in London, one of the world’s great art capitals and home to extraordinary, thrilling exhibitions such as last year’s ‘Bronze’ at the Royal Academy

