Printer Friendly

The Hund Statues: Recent Discoveries between Gandhara and Kashmir.

A GROUP OF UNIQUE CLAY SCULPTURES was discovered some 20 years ago in the Greater Kashmir region of ancient Gandhara, between the Indus and Swat valleys, east of Peshawar, northern Pakistan (figure 2). The sculptures are believed to have been found at or near Hund, the ancient Hindu Shahi capital Udabhandapura. But according to other sources, these were found further northwest at Shabaz Garhi (Bakshali village), Mardan district. Of a hitherto unseen style and of superb quality, the statues have been presented on the international art market since they were found, but were documented as a preliminary survey only recently. "Hund", though in usage fairly recently as a place name, is henceforth also used as a common term for all images of the specific Kashmir-influenced style discussed here. (1)

The so far unconfirmed exact location of the excavation, an almost complete lack of comparable images, a distinctive style that reveals extraordinary cultural and art-historical uniqueness, the lack of serious scholarly attention, alongside the enigmatic iconographic identity of the images (whether Buddhist or Hindu), have led to rumours questioning their authenticity. This has perpetuated ignorance and prevented interpretation of these important objects. The suggested chronology by those who have handled and collected these statues so far, covering some 500 years from the 5th to 10th century, is evidence of a completely new and challenging chapter of Asian art history. These characteristics correspond well to the complex geo-cultural profile of the multi- and cross-cultural history of the Indo-Afghan borderlands from the post-Gandharan to the Islamic period, when Kashmir became the dominant political and cultural force in the northwest of India and Gilgit-Karakoram in the 7th and 8th centuries.

Regional kingdoms favoured and patronized Brahmanical and Buddhist orientations simultaneously, instead of successive periods being clearly identifiable with either religious tradition. This coexistence resulted quite often in an iconographic syncretism in Gandharan art, when far away from mainstream traditions in India, Hindu iconographies were modified both in motifs and in context to be incorporated into Buddhist image programmes. (2)

Absorbing earlier Hellenistic and Iranian influences, Afghanistan became a melting pot of different cultural traditions. Brahmanical centres existed in the northern neighbouring areas from Kabul to Khotan along the southern Silk Routes from circa 5th to 7th century. With regard to the growing influence of Brahmanism in northern Afghanistan, the Chinese Buddhist pilgrim Xuanzang mentions Kapisha (or Ka-pi-shi), in his famous travelogue from the years 629 to 645, referring to the ancient Kushan capital later known as Begram/Bagram, where Buddhism was about to decline and "stupas were desolate and rained". What emerged in place were "some ten temples of the (Hindu) devas, and thousand or so heretics" and "naked ascetics, and others who cover themselves in ashes". (3)

The increasing Brahmanic influence from the 6th to 8th century, in and from the Kabul region and further south, is well documented by many Hindu statues preserved in the Kabul Museum. This cultural flow continued in coexistence with Buddhist images in the whole Swat-Kashmir region during the 8th century. At that time, Gandharan Buddhism and its art had moved to Kashmir, then part of the Hindu Shahi territory, where it was to some extent assimilated, absorbed or inherited, not unlike how Greek-Roman art traditions were adapted into the medieval European art of Byzantium.

Besides the better-known Gandharan and Kashmiri sculptural styles, quite different Brahmanic and Buddhist statues were produced in the Shahi kingdom ateliers in the region between Swat and Kabul from the 7th through 10th centuries (figures 1 and 3). (4) They are mentioned and illustrated here to document "Hund in context", to reveal the different regional art traditions which were concurrent.

Following the royal Turki Shahi or Early Shahi rale which was predominantly Hindu-oriented, yet also supported Buddhism in the 7th century over what is today Afghanistan and northern Pakistan, the largely Brahmanic later Hindu Shahi dynasty moved, towards the end of the 9th century, for the next 100 years, from the Kabul region south to Swat-Gandhara and then to Hund-Udabhandapura.

Iconological Context

A large head once belonging to a probably life-size 7th or early 8th century clay statue of Durga Mahishasuramardini was discovered in 1968 at Tepe Sardar monastery, southwest of Kabul (figure 4), opposite a former central standing Buddha of the bejewelled Fondukistan type, and fragments from a Mahishasuramardini, with the legs of Durga resting on a buffalo head and her hand holding a vajra of the early "Gandhara" type (see pages 91, 92). (5) This popular manifestation of Durga in the northwestern region was already known since Gandharan times. As a foremost royal protecting goddess under the Hindu Shahis, she was incorporated into a Buddhist monastery, and provides a significant example of these Inner Asian cultural crossroads and of the syncretistic milieu during that period. How can we describe and identify the position of the images under review within the multicultural context as surveyed above?

Without reliable contexts to tell us specifically whether we are looking at a Hindu Tantric or Buddhist Vajrayana context, it is not always easy to distinguish them. This is because they share several iconographic features. Similarly, this unusual Buddhist-Hindu context can be recognized also among some male Hund statues. These figures with a slightly open mouth, the teeth visible, recall Tantric Buddhist dakini and yogini goddesses (figure 10). (6) Can these devatas, demigods between heaven and earth, be identified as semi-wrathful female Hindu Tantric divinities, or rather, as is written in the ancient texts, as goddesses with beautiful physiognomies and with a wrathful capacity that represents their holy violence and powerful protective energy? And were these images without any traditional Hindu or Buddhist attributes once acolytes around a central Mahadevi Durga Mahishasuramardini, the Great Mother, or mythical yakshas and yakshis, intermediaries between the divine and the human, associated with local field and forest deities, rural and vegetal fertility forces, which can appear in many different forms?

The iconological context of these worldly looking celestial attendants and "apsaras" remains, however, a mystery. No descriptive information exists yet on such "deva temples" (Xuanzang, 7th century) or about the specific site where these statues were found. Further no comparable full-figure statuary has come to light.

The formal overall characteristics, though not of the individual, rather "manneristic" Kashmir style of the Hund devatas, have parallels with the earlier painted clay statues from around 700 ad or later in the 8th century found at Fondukistan, (7) halfway between Kabul and Bamiyan along the grand trade routes from India to inner Asia, which also cannot be easily associated with specific religious schools. The elegant postures and elongated slim proportions, the slender waist and affected or even highly contrived hand gestures, the rich (often floral) jewellery, or the decorative headdress and sumptuous floral ornaments may well be regarded as stylistic antecedents (figure 5). Closer to "Hund" in style and geographic origin is a female terracotta head in the Peshawar Museum, excavated in 1912 at Sahri Bahlol (figure 7). This ancient town of Po-lu-sha (Bolusha) with Buddhist monasteries, south of Takht-i-Bahi as described by Xuanzang in the 7th century, is near modern Mardan, where the images from Hund discussed here were reportedly found. (8) The very similar headdress and facial features with the characteristic Kashmiri slit-eyes and engraved crescent-shaped eyebrows are essential features of the Hund sculptures (figure 6). This art-historical and stylistic attribution is confirmed by several terracotta fragments from the same excavation site, which provides a valuable context for our statuary group (figure 9). (9)

Of a very similar style is also an exceptional 67-cm-tall bronze image of Durga Mahishasuramardini attributed to Kashmir, from the 8th century, whose body details, crown, earrings and ornamental jewellery can well be compared with the Hund statues (figure 12). (10) The same stylistic profile is shown by various superb Kashmir "bronzes", for example the adoring figures at the pedestal of a seated Shakyamuni in the Los Angeles Norton Simon Foundation museum dated to around 700 (figure 14). (11)

Technical Aspects

The images are predominantly made of burnt clay (terracotta), the common statuary material in the post-Gandharan Indo-Afghan borderlands. Due to the lack of more systematic technical investigations it is presently difficult to say whether they were all fired originally, or in several cases burnt accidentally, when, like at Tepe Sardar, and no doubt at other sites, the shrines were set on fire in course of time. (12) Or were these statues originally composites of partially burnt terracotta portions in combination with unburnt decorative elements like crown details and other ornaments? Apparentiy thin layers of clay have been added for better "drawing" the facial features and decorative elements. As with the older stucco images at Hadda in the 4th century, the documentation by computed tomography reveals that the fine clay was applied around an inner core of coarse clay and other materials, which was then covered by still finer plaster-like layers. In Mes Aynak, Tepe Narenji, and eventually also in Fondukistan and at Tepe Sardar, north Afghanistan, where unburnt painted clay was the preferred material for sculptures, a coat of calcined lime plaster or gypsum (not unknown in early Gandhara and Bactria and found as a minor substance in a technically analysed Hund male devata) might have been used for surface layers. Stucco was already clearly distinguished from clay in Indian textual sources and was well known at many later Gandharan sites. It was also used for the outer layers of clay statues, though apparently not for Hund. Here, the head, neck, arms and legs were made separately and stabilized by wooden or, more rarely, metal armatures, of which the remaining rods can be seen inside a damaged statue or else with the help of tomographic radiographics (figure 13). (13)

The Hund statues that are in collections are in surprisingly good condition. Broken heads and legs, arms and fingers have been repaired and added to the figure again. The extent of renovation and conservation (like the manufacturing ateliers), is however, difficult to identify. When is a statue restored, re-done, overdone or reconstructed? And as the American art conservation scientist John Twilley told me, "the biggest problem with these sculptures is not that they contain repairs, but rather that the terracotta fragments themselves may not [and one can add, do not always] belong together".

While easy-going comments, without any facts about the authenticity, still come from respectable connoisseurs, this author had access to apparently "original" fragmentary sculptures and to photos of their presumable pre-restoration condition. (14) Both the original image and the photographic record were confirmed by an experienced informant in this field, and made it very likely that at least a major part of the early wave of Hund discoveries is basically antique. Although there is presently, even three years after my initial publication on the subject in 2016, no relevant technical documentation available for these statues, we can assume that some "repairs" have made their appearance more "beautiful" and too perfect. A pre-restoration photograph (from 2004) of a grosso modo well-preserved female devata upper half shows, if trustworthy, from whatever can be recognized, that the head was broken at the neck as were the two arms in the upper shoulder section. Interestingly some of the then still complete crown finials are missing from the sculpture as it was known after its restoration from 2012 (figure 11). However, when studied with a critical eye, even a photo of a sculpture prior to known restoration does not necessarily prove its authenticity.

Nine Hund statues and three heads (which this author knows of) have been analytically tested between 2005 and 2015 in British, Italian, German and Hongkong laboratories: thermoluminescence, radiocarbon, computed tomography, scanning electron microscopy, micro element and thermal analysis. The results warrant careful scrutiny and further analysis. Two items were certified as modern (one probably wrongly, as explained below), while the other nine objects, all of the same "Hund style", were declared as produced between the 4th and 9th century. Often an incorrect and slipshod "middle-way" of dealing with a time range of c. 500 years is used and people just take the mean date, which would land us at a c. 6th/7th century date. This would not fit these statues for stylistic and historical reasons; however the test actually gives us an equal possibility of a later date, and thus this should not question the reliability and the value of certain analytical examinations for dating a historical object. In one instance of the two items certified as "modern", the heterogeneous filling material of an examined female bust (15) was interpreted as recent and not part of the original composition, and some chemical elements from the surface layers were identified as modern substances, leading to the erroneous conclusion that the statue itself was a modern artefact. However, it is highly probable that a surface treatment was undertaken some years before, and that would have naturally included modern elements. Testing and sampling therefore need to be undertaken with far greater attention and care. Also in this case visual evidence from photographs taken before the final restoration (known to this author) suggest a largely authentic c. 9th-century statue. While radiocarbon-testing cannot be used for non-organic materials such as unbaked or baked (fired) clay, analysis of some pieces of straw found inside a devata image have indicated a date between the late 8th and mid-10th century, which corresponds well to a stylistic and historical 9th-century Hund chronology as given here.

Some sculptures are polychromed and covered by a shiny pearl-like mica varnish of which nothing comparable exists among any related statuary (figure 8). A few others, provided they are in largely original though apparently re-done condition, have traces of an ancient polychromy as well. (16)


These statues are of no small importance in the history of South Asian art, as they connect the long-standing traditions of Gandhara and Swat to Kashmir, and incorporate the sensibilities of a critical time that saw the rise of various Tantric elements in Hindu and Buddhist imagery. Rumours about their authenticity have prevented serious scholarship on their iconology, iconography, style and context, and that, in fact, has facilitated only depredations of the sites and over-conservation of the statues themselves. They deserve to be attributed to an unidentified Kashmir-style-influenced Brahmanic sanctuary of the c. 9th-century Hindu Shahi period in today's northern Pakistan Mardan district, once the ancient territory of Gandhara.


(1) Michael Henss, "The Mystery of Hund Statues. An Unknown Chapter of Central Asian Statuary", Arts of Asia, January/February 2016, pp. 28-43.

(2) Ibid., p. 30. For a 120-cm-tall Gandharan style schist statue of a Brahmanic "Nana-Durga", see Michael Henss, "Buddha: 2000 Years of Buddhist Art" (exhibition review, Volklingen/Saarbrucken 2016/2017), Arfs of Asia, March-April 2017, fig. 1.

(3) S. Beal, Si-Yu-Ki: Buddhist Records of the Western World, Vol. I, London: Trubner, 1884; Delhi: Motilal Banarsidass Publishers, 1981, pp. 55,91.

(4) A. Filigenzi, "The Buddhist Site of Tapa Sardar", in Fifty Years of Research in the Heart of Eurasia, Rome: Istituto Italiano per l'Africa e l'Oriente,-2009, pp. 41-57; "The Shahi Period: A Reappraisal of Archaeological and Art Historical Sources", in Coins, Art and Chronology, 11: The First Millenium c.e. of the Indo-Iranian Borderlands, Vienna: Austrian Academy of Sciences Press, 2010, pp. 407-27. For statues, see D. Stadtner, "Shahi Sculpture Revisited", Orientations, October 1999, pp. 68-73.

(5) For further references on the Tepe Sardar statues (chapel 23) see Henss, "The Mystery of Hund Statues", notes 3-5; M. Taddei, "The Mahisasuramardini image from Tapa Sardar, Ghazni", in South Asian Archaeology, edited by N. Hammond, London, 1971, pp. 203-12; C. Silvi Antonini, "Considerations on the Image of Mahisasuramardini of Tapa Sardar", East and West, 2005, pp. 313-28. I cannot recognize, however, from a stylistic perspective a certainly later date than of the "Chinese phase"(?) clay sculptures at Tepe Sardar as suggested by G. Verardi, "Issues in the Excavation, Chronology and Monuments of Tapa Sardar", in Coins, Art and Archaeology, II, p. 346. For another vajra-holding Tantric Kashmir-style Durga in the New York Metropolitan Museum, see P. Pal, Art and Architecture of Ancient Kashmir, Bombay: Marg Publications, 1989, p. 111, pl. 10.

(6) Henss, "The Mystery of Hund Statues", 2016, pp. 31f., figs. 4,5,29,32, and figs. 2 and 3, p. 32, for human skulls as part of the headdress decoration, provided that the heads are in authentic condition. At Tepe Narenji, Kabul, remains of a central homa fire shrine and of ("most likely") five Tathagata Buddhas were excavated in the centre of a stupa, documenting Tantric rites and iconographies in c. 8th century: cf. Z. Paiman, "Tepe Narenji: A Royal Monastery on the High Ground of Kabul", Journal of Inner Asian Art and Archaeology, 5, 2010/2012, pp. 33-53.

(7) S. Novotny, "The Buddhist Monastery of Fondukistan, Afghanistan: A Reconstruction", Journal of Inner Asian Art and Archaeology, 2,2007, pp. 31-37.

(8) Beal, Si-Yu-Ki, pp. inf. See also Henss, "The Mystery of Hund Statues", p. 32, figs. 7 and 8 for a male c. 8th-century head from Ushkur, west of Srinagar, Central Asian Museum, Srinagar.

(9) Henss, "The Mystery of Hund Statues", figs. 11,20,21.

(10) P. Pal, Goddess Durga. The Power and Glory, Bombay: Marg Publications, 2009, cover and pp. 85-87; Christie's New York, 22.3.2011, no. 220.

(11) Pal, Art and Architecture of Ancient Kashmir, no. 2.

(12) See Henss, "The Mystery of Hund Statues", figs. 17,18, and on technique and materials in more detail pp. 34-42.1 have to identify earlier clay images now more correctly as made from burnt clay (terracotta), with, as it seems, additional layers and sections of dry unburnt clay covering the surface and apparently also in combination with terracotta in certain sections. These issues, even most simple identifications as clay or terracotta, have been neglected in several scholarly publications (Antonini, "Considerations on the Image of Mahisasuramardini from Tapa Sardar"; Filigenzi, "The Buddhist Site of Tapa Sardar"; Verardi, "Issues in the Excavation, Chronology and Monuments of Tapa Sardar").

(13) Henss, "The Mystery of Hund Statues", figs. 17 and I9d.

(14) Ibid., figs. 17,19a, 20,21,33,33a. These photographs were not suitable for publication because of their modest quality.

(15) Ibid., figs. 19-19d.

(16) Ibid., figs. 31,32.

Caption: 1 Vishnu with attendants, Hindu Shahi period, Afghanistan, c. 8th century. Marble; height: 54 cm. Private collection. PHOTOGRAPH: ARCHIVES MICHAEL HENSS. SEE ALSO BELOW NOTE 4: STADTNER 1999, FIG. 7A.

Caption: 2 Map of the Gandhara-Swat area, northern Pakistan. FROM ADRIANA G. PROSER, THE BUDDHIST HERITAGE OF PAKISTAN, NEW YORK: ASIA SOCIETY, 2011, P. 33.

Caption: 3 Head of a Brahmanical deity, Turki Shahi period, c. 7th century. Marble; height: 18.3 cm. BRUSSELS ANCIENT ART FAIR 2013 (JOHN SIUDMAK). PHOTOGRAPH: ARCHIVES MICHAEL HENSS.

Caption: 4 Head of Durga Mahishasuramardini found atTepe Sardar monastery, Afghanistan (chapel 23), c. 5th century. Clay. AFTER TADDEI 1971, SEE NOTE 5 BELOW.

Caption: 5 Bodhisatva from Fondukistan (niche D), Ghorband valley, now at the Kabul Museum. Painted red unbaked clay; height: c. 72 cm. [C] NATIONAL MUSEUM OF AFGHANISTAN. K.M. INV. NO. 65-9-5.

Caption: 6 Head of a male devata from Hund, Pakistan, c. 9th century. Terracotta and clay; height: 27 cm. Private collection. COLLECTION OF AND COURTESY TONY ANNINOS.

Caption: 7 Female head from Sahri Bahlol (northwest of Hund, Pakistan), 8th/9th century, now at the Peshawar Museum. Terracotta; height: 14.5 cm. FROM ADRIANA G. PROSER, THE BUDDHIST HERITAGE OF PAKISTAN, NEW YORK: ASIA SOCIETY, 2011, P. 320.

Caption: 8 (above) Head of a male devata from Hund, Pakistan, c. 9th century. Polychromed terracotta and clay; total height: 93 cm. Private collection. PHOTOGRAPH: ARCHIVES MICHAEL HENSS.

Caption: 9 (below) Head of a female devata from Hund, Pakistan, c. 9th century. Terracotta; height: c. 20 cm. Private collection. PHOTOGRAPH: ARCHIVES MICHAEL HENSS.

Caption: 10 Seated male devata from Hund, Pakistan, c. 9th century, re-done to an unknown extent. Polychromed clay; height: c. 80 cm. Private collection. PHOTOGRAPH: ARCHIVES MICHAEL HENSS.

Caption: 11 Upper half of a female devata from Hund, Pakistan, c. 9th century [tl-test Oxford 2010 (6th-9th century)]. Terracotta and clay; height: 63 cm. AFTER CHRISTIE'S PARIS 12.6.2012, NO. 304.

Caption: 12 Durga Mahishasuramardini (detail), Kashmir style, c. 8th century. Bronze; height: 67 cm. Private collection. AFTER CHRISTIE'S NEW YORK 22.3.2011, NO. 220.

Caption: 13a and b Tomographies (CT) of a male devata bust seen from the back, and of the head in profile, from Hund, Pakistan, c. 9th century. Terracotta and clay; total height: 68 cm. PHOTOGRAPHS: ARCHIVES MICHAEL HENSS.

Caption: 14 Seated Buddha Shakyamuni, Kashmir, c. 700 ad or 8th century. Gilt brass; height (including pedestal section): 33.7 cm. Norton Simon Museum of Art, Pasadena, USA. FROM P. PAL, ART FROM THE HIMALAYAS AND CHINA, NEW HAVEN: YALE UNIVERSITY PRESS, 2003, P. 31.

Caption: U Shiva, probably from the Kabul valley, Hindu Shahi period, c. 8th century AD. Marble. Private collection, UK.

By the 6th and 7th centuries several temples dedicated to Surya, Shiva and Qanesh had been constructed in the Kabul valley. The style and iconography of their sculptures reveal continuities with the art of Kashmir a century or two later. In fact, at times they share such subtle iconographic correspondences that it seems they prefigure several complex "esoteric" ideas we normally associate with the developments of Tantras and Shaivism in Kashmir.

Please Note: Illustration(s) are not available due to copyright restrictions.
COPYRIGHT 2019 The Marg Foundation
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Henss, Michael
Publication:Marg, A Magazine of the Arts
Geographic Code:9AFGH
Date:Jun 1, 2019
Previous Article:The Sakra Sites and Their Enigmatic Coins.
Next Article:The Histories in Between: Afghanistan's Religious and Material Cultures.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters