Show Summary Details

Page of
PRINTED FROM Oxford Art Online. © Oxford University Press, 2019. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single article in Oxford Art Online for personal use (for details see Privacy Policy).


  • D. W. MacDowall,
  • W. Ball,
  • Gregory L. Possehl,
  • Maurizio Taddei,
  • C. Fabrègues,
  • E. Errington,
  • N. Hatch Dupree,
  • Sheila S. Blair,
  • Jonathan M. Bloom
  •  and F. Tissot

Map of Afghanistan; those areas with separate entries in this dictionary are distinguished by Cross-reference type

View large

Country of some 647,500 sq. km in the middle of the steppe and desert zone of Eurasia. It is bounded on the north by the Amu (Oxus) River and the republics of Central Asia, on the west by Iran and on the south and east by the Indian subcontinent. In the Pamir Mountains to the north-east, a narrow tongue of land known as the Wakhan corridor links the country with China (see fig.). Located at the crossroads of major trade and migration routes between the Mediterranean, Central Asia, India and China, the region has been subjected to diverse cultural influences throughout its history.

I. Introduction.

1. Geography and climate.

The physical geography of Afghanistan is very varied and includes formidable mountain ranges, fertile valleys and barren deserts. The dominant mountainous core is the Hindu Kush, an extension of the Karakoram and Pamir mountains that stretches south-west for some 965 km and has peaks rising to some 5180 m in height. To the north, between the Hindu Kush and the Amu River lie the semi-desert plains of Turkestan. South of the Hindu Kush is a transitional zone of plateaux with broad mountain valleys. To the west and south-west the mountains gradually descend to the stony and sandy deserts of the Iranian plateau. North of Kabul the Kuh-e-Baba range (‘Grandfather Mountains’) of the Hindu Kush is the watershed for four great Afghan rivers: the Kabul River flowing east to the Indus, the Kunduz flowing north into the Amu River, the Hari Rud flowing west to Herat and the Helmand, which flows southwards into the marshy lake of Hamun Helmand in Sistan. There are several passes through the mountainous core of the country linking north to south and east to west, and traffic is also channelled along the rivers or round the mountain mass. The low-lying plains and deserts between Herat and Kandahar provide an easy route for traders and invaders travelling eastwards into the Indus Valley.

The climate is generally dry, with wide variations in temperature. Snow falls in the mountainous areas above c. 1830 m from October onwards and blocks the passes for much of the winter. In the plains of Turkestan most of the rain falls as spring thunderstorms, and there are sometimes disastrous floods when this water combines with melting snow from the mountains. In winter there is rain in the Herat area and the rivers are swollen with melt water in the spring, but in the Helmand basin there is virtually no rainfall in any season. The Jalalabad Valley has a winter rainfall and can, with irrigation, grow rich crops.

2. History.

The limited documentation of prehistoric sites in Afghanistan has produced evidence of a small Palaeolithic hunter–gatherer population, part of which was established by about 20,000–15,000 bp in the northern foothills of the Hindu Kush at cave sites such as Aq Kupruk II (Ghar-i Mar). The domestication of sheep, goats and possibly cattle seems to have taken place by about the 9th millennium bc. Some of the Neolithic sites produced pottery, but there are no architectural remains associated with this phase of the transition from hunter–gathering to food production. Knowledge of bronze technology and embossing by the late 6th millennium bc is indicated by finds of sheet metal from Aq Kupruk II. From about 3500 bc onwards there is evidence of trade links associated with the export of lapis lazuli, particularly at Shahr-i Sokhta in Sistan. The lapis lazuli, which seems to have derived from the Badakhshan region in north-east Afghanistan, especially the Kokcha Valley, was exported all over the Near East. Mundigak and the related sites of Deh Morasi and Said Qala in southern Afghanistan show the development of urbanization from around the early 4th millennium bc to the 2nd.

From the Ṛg veda (see Indian subcontinent, §I, 2, (i)), it appears that the Aryans passed through Afghanistan c. 1500 bc. Later Aryan migrants from Transoxania settled on the Iranian plateau and established the Achaemenid empire in the regions between the Mediterranean and the Indus River by the 6th century bc. In 330 bc the last Achaemenid, Darius III (reg c. 336–330 bc), was defeated by Alexander the Great, who went on to conquer the eastern satrapies in Afghanistan, bringing them into direct contact with Hellenism from the west. Alexander’s successors, the Seleucids, retained control in Bactria (the Amu River region north of the Hindu Kush and east of Merv and Herat) but lost the territories of Kabul and Kandahar c. 305 bc to the Mauryans from India.

About 250 bc Diodotus, the governor of Bactria, declared independence from the Seleucids. Subsequent Greco-Bactrian kings extended their territory south of the Hindu Kush to Kabul and Kandahar and invaded India (see Bactrian and Indo-Greek monarchies). After the initial successes of Menander (reg c. 170 bc), the kingdom became fragmented under several rulers, partly as a result of increasing pressure from nomadic Sakas or Scythians migrating southwards from Central Asia. One group, the Yueh-chih, occupied Bactria c. 130 bc. The remaining Indo-Greek kings to the south were replaced by Indo-Scythian rulers of the house of Azes in the 1st century bc and by Indo-Parthian under Gondophares (reg c. ad 20–50) in the 1st century ad.

In the 1st century ad the Kushana tribe united the Yueh-chih confederacy and established a powerful empire that expanded from Central Asia across Afghanistan to north-west India. The extent and stability of this empire encouraged the growth of international trade along the silk route from China across Afghanistan to the Indus River ports and thence by sea to Alexandria and Rome. Under the third king, Kanishka I, the Kushanas patronized Buddhism, stupas and monasteries were established throughout Afghanistan, and missionaries followed Kushana traders across Central Asia to China. The chronology of this period is still disputed, the era of Kanishka being attributed to various dates between ad 78 and the 3rd century (see also indian subcontinent, §IV, 5(ii)). What is clear is that the Kushanas were defeated by the rising power of the Sasanians (c. 224–651) in the 3rd century. Afghanistan suffered again in the 5th century from the invasions of the Hephthalites (White Huns), who in turn were overthrown in the mid-6th century by the Turki Shahis, allied to the Sasanians. Raids into western Afghanistan in the late 7th century gave the Arabs control of Sistan and Herat. From the 9th century, western Afghanistan was ruled by local Islamic dynasties: the Samanid (reg 874–999) based in Bukhara and the Saffarids (reg 867–1495) based in Sistan. Eastern Afghanistan remained an independent non-Muslim kingdom, centred at Kabul, under the Turki Shahis and their successors the Hindu Shahis, until conflicts with the Ghaznavid (reg 977–1186), an Islamic dynasty originating from Ghazna, forced a transfer of the Hindu Shahi capital to Hund, on the Indus River east of Peshawar.

The first Ghaznavid, Sebüktigin (reg ad 977–97), governed on behalf of the Samanids, but his son Mahmud (reg 998–1030) established an independent empire over Samanid territories south of the Amu River and expanded eastwards into India. By the mid-11th century, western Afghanistan had been relinquished to the Saljuqs (see Saljuq family, §1), Turkish nomads originating from the steppelands north of the Caspian and Aral seas, but the Ghaznavids retained control of eastern Afghanistan and northern India. In 1151 Ghazna was sacked by Ghurid chieftains from the inaccessible mountainous region east of Herat. Turkish tribes from the lower Syr (Jaxartes) River region overthrew the Saljuqs in 1153 and occupied Ghazna in 1163. In the next decades, the Ghurids gained control of Afghanistan and finally defeated the last Ghaznavid principality at Lahore in 1186.

Internal dynastic struggles and confrontation on the northern borders with the Khwarazmshahs resulted in the breakup of the Ghurid empire. From 1215–16 Ghurid territories were ruled by the Khwarazmshah Muhammad b. Takash (reg 1200–21), until the entire region was overrun in 1221 by the Mongols under Genghis Khan (reg 1206–27). Herat was restored in 1236 by his third son Chaghatay (Ögedey; reg 1227–41), while Ghazna and Kabul became military bases for Mongol raids into India. From 1250 onwards the different Afghanistan regions were controlled by independent Mongol rulers, such as the Neguderis at Ghazna.

Timur extended his Transoxanian steppe empire southwards into Afghanistan with the capture of Herat in 1380. Under his son Shahrukh, Herat became the Timurid family capital. From 1469 Afghanistan was divided into two Timurid principalities, one based at Herat, the other at Kabul. In the 16th century incursions by the Uzbek tribal confederacy of the Amu River region meant that frontier towns such as Herat frequently changed hands.

Following the loss of his Central Asian Timurid principality of Ferghana to the Uzbeks, Babur (reg 1526–30), the founder of the Mughal family dynasty, occupied Kabul in 1504. He began raids into India and captured Delhi in 1526. Afghan chiefs led by Sher Shah Sur (reg 1540–45; see Sur) forced Babur’s son Humayun (reg 1530–40, 1555–6) into exile in 1540. The capture of Kabul in 1545 gave Humayun a base from which to reconquer India in 1555. During the reign of Akbar (reg 1556–1605) boundaries between Afghanistan and the Uzbek territories to the north were demarcated, but control of Kandahar remained disputed with the Safavid dynasty (reg 1501–1732; see Safavid family, §I) to the west. Succeeding Mughals retained Kabul, but western Afghanistan came increasingly under Safavid control. In the early 18th century the Safavid governor, Mir Ways, declared independence. The Afghans occupied most of Iran from 1722 until expelled in 1727 by Nadir Shah (reg 1736–47), a Turkoman chieftain from Khurasan in service with the Safavids, who subsequently founded the Afsharid dynasty (reg 1736–95) of Iran. When Nadir was assassinated in 1747, Afghan soldiers in his army elected one of his leading commanders, Ahmad Khan of the Afghan Sardozay tribe, as Shah (reg 1747–73). The Durrani dynasty (reg 1747–1842), which Ahmad Shah founded, derived its name from his title Dur-i Durrān (‘Pearl of pearls’). He established an empire comprising Afghanistan and north-west India, including Sind, Baluchistan, part of the Punjab and Kashmir, but most of the Indian territories were lost during the reign of Zaman Shah (reg 1793–1800).

In 1819 Dost Muhammad of the Barakzay tribe (reg 1819–62) took Kabul and retained control of Afghanistan despite pressures from Iran, Russia and the British. The kingdom of Afghanistan survived as a political entity until overthrown by leftist urban groups in 1978. The ensuing civil war was not halted by Soviet military intervention in 1979 or by the withdrawal of Soviet troops in 1987–9.


  • C. E. Bosworth: The Islamic Dynasties: A Chronological and Genealogical Handbook (Edinburgh, 1967, rev. 1980)
  • N. Dupree: ‘Archaeology and the Arts in the Creation of a National Consciousness’, Afghanistan in the 1970s, ed. L. Dupree (New York, 1974), pp. 203–38
  • F. R. Allchin and N. Hammond, eds: The Archaeology of Afghanistan: From Earliest Times to the Timurid Period (London and New York, 1978)
  • R. Dor and C. Naumann: Die Kirghisen des afghanischen Pamir (Graz, 1978)
  • L. Dupree: Afghanistan (Princeton, 1980)
  • E. W. Anderson and N. H. Dupree: The Cultural Basis of Afghan Nationalism (London and New York, 1990)
  • A. B. Delmas and M. Casanova: ‘The Lapis Lazuli Sources in the Ancient East’, South Asian Archaeology, 1987, ed. M. Taddei (Rome, 1990), pp. 493–505
  • W. Ball and L. Harrow: Cairo to Kabul: Afghan and Islamic Studies Presented to Ralph Pinder-Wilson (London, 2002)
  • B. Dupaigne and G. Rossignol: Le carrefour afghan (Paris, 2002)
  • N. H. Dupree: ‘Cultural Heritage and National Identity in Afghanistan’, Third World Q., 23/5 (2002), pp. 977–89
  • J. W. Betlyon: ‘Afghan Archaeology on the Road to Recovery’, Near Eastern Archaeology, 67/1 (2004), pp. 59–60
  • M. K. Palat and A. Tabyshalieva: Towards the Contemporary Period: From the Mid-nineteenth to the End of the Twentieth Century (2005), vi of UNESCO History of Civilizations of Central Asia (Paris, 1992–), pp. 439–60, 757–93
  • J. B. Spurr: ‘Glimpses of an Eclipsed Heritage: Photography of Afghanistan in the Collections of the Fine Arts Library at Harvard’, Visual Resources, 21/1 (2005), pp. 55–71
  • C. H. Beaney and M. Á. Gallego: Afghanistan: A Bibliography, Handbook of Oriental Studies: Central Asia, 13 (Leiden, 2006)
  • J. van Krieken-Pieters: Art and Archaeology of Afghanistan: Its Fall and Survival, a Multi-disciplinary Approach, Handbook of Oriental Studies: Central Asia, 14 (Leiden and Boston, 2006)

3. Language and ethnic groups.

The earliest official inscriptions date from the Mauryan period. These comprise Aramaic inscriptions from Laghman and two bilingual inscriptions of Ashoka (reg c. 269–c. 232 bc) from Kandahar, one combining Aramaic with Greek and the other Aramaic with Indian Prakrit. Under the Greco-Bactrians at Ai Khanum, official inscriptions were in good Classical Greek. Under the Kushanas, inscriptions at Surkh Kotal, north of the Hindu Kush, were in Bactrian; inscriptions at Dasht-i-Nawar, 60 km west of Ghazna, were in Bactrian, Prakrit and an undeciphered local language, while the inscriptions of south-eastern Afghanistan were in Prakrit.

Afghanistan is not a single ethnic unit. In the pre-Muslim period, the Hindu Kush (‘Death to the Hindu’) Mountains formed a natural divide between the Hindu-dominated areas of the south and the Zoroastrian peoples of the north. The largest racial group in modern Afghanistan consists of the Pathans, a people of Turko-Iranian origin who speak Pushtu and are probably the descendants of the original inhabitants of the south. The Tajiks, who live north and east of Kabul, are of Iranian origin, speak Farsi and are thought to be descendants of the original northern inhabitants. The Hazaras who inhabit the central massif are thought to be descendants of the Mongols. There are also many minority groups, such as the Turkomans, Uzbeks and Nuristanis, who all speak their own dialects.

4. Religion and iconography.

  • D. W. MacDowall

The terracotta female figures from Mundigak (see §II, 1, (ii), (a) below) suggest that a fertility goddess resembling the great mother goddess of Mesopotamia was worshipped in the prehistoric period (c. 2000 bc). Zoroastrianism played an important role in Iranian lands such as Bactria before the conquest of Alexander the Great in 330 bc, but many remains of a pre-Zoroastrian religion also survived. The Sakas and Yueh-chih in Central Asia probably worshipped the sun and regarded Ahuramazda (the Zoroastrian supreme Good Spirit) as a god with a strong solar function (see Central Asia, §I, 1, (v), (a)).

The coins of the Greek rulers of Bactria showed a typical range of Greek divinities, with the same attributes and iconography as found throughout the Hellenistic world: Zeus with his thunderbolt, Artemis with her bow and Pallas Athena with her shield. According to Herodotus (Histories IV.lviii–lxix), the Scythians worshipped Greek gods but called them by Scythian names, and this also reflects the Greek approach to foreign deities. Native divinities were syncretized with Greek gods and represented in Greek iconographic forms. The votive pedestal at Takht-i Sangin, dedicated by the Iranian priest Atrosokes to the deity of the Oxus (Amu River) in the 2nd century bc, supported a statue of the Classical deity Silenus Marsyas, playing a double flute. Rare coins of Agathokles (c. 180 bc) from Ai Khanum follow the Classical tradition by depicting the Hindu deity Samkarshana in iconic form as a male figure in Oriental dress and winged headdress holding a plough, while his younger brother Vasudeva Krishna is similarly shown bearing the attributes of a conch shell (saṅkha) and wheel (cakra) (Kabul Mus.; see Allchin and Hammond, fig.). The enthroned and radiate Zeus on Indian tetradrachms of Hermaios (c. 90–70 bc) represents Zeus syncretized with Mithra, an Iranian pre-Zoroastrian solar deity (see 1992 exh. cat., p. 61, no. 24).

There were similar developments when the Kushanas extended their empire across Afghanistan in the 1st century ad. The first ruler, Kujula Kadphises, used Herakles, the Greek form of the Iranian war god and personification of victory, Verethragna. King Soter Megas used the rayed head of Mithra. All the gold and copper coins of his successor Vima Kadphises use the type of Shiva and his bull Nandi, showing the king’s personal devotion to the war god, appropriate to his campaigns of conquest and to the support he required from his new Indian subjects. Kanishka was more eclectic, and his coins show a pantheon of primarily Iranian gods. Initially he used Greek legends, even to the extent of using the female name Selene for the male Iranian moon god. The Greek names were subsequently replaced by Bactrian ones, such as Mao (moon god), Mioro (sun god), Athsho (fire god) and Nana (water goddess, the Bactrian Anahita).

The Mauryan ruler Ashoka (reg c. 269–c. 232 bc) promoted Buddhism in Afghanistan as elsewhere, and inscriptions giving his edicts have been found in Laghman and at Kandahar; but it was only from the late 1st century bc onwards, under the Indo-Parthians and Kushanas, that Buddhism spread widely and numerous stupas were built in Afghanistan. During this period there were some remarkable developments in Buddhist iconography. Whereas previously the Buddha’s presence had simply been indicated by symbols, such as a footprint or vacant throne, a new school of Buddhist thought stressing the miraculous life of the Buddha led to the representation of the Buddha in human form. A gold medallion from Tillya Tepe (see Sarianidi, pp. 188–9, no. 131), dated c. 50 bcc. ad 50, has a reverse design of a lion and a triratna (‘three jewels’ symbol representing Buddha, the Law and the Buddhist community). The Kharoshthi inscription siho vigatabhayo (‘the lion who chased away fear’) refers to the Buddha, as does the obverse dharmacakrapravatako (‘he who sets in motion the Wheel of the Law’; see Fussman, pp. 71–2), which is inscribed beside a bearded image of Herakles, complete with lion skin, pushing a wheel. The equation of a Classical deity with the Buddha suggests that the token was made in the aniconic period of Buddhist art, but within the Greco-Bactrian milieu, where religious images were traditionally represented in human form.

The gold reliquary from Bimaran Stupa 2 (see §II, 1, (iv), (c) fig. below) also belongs to an early phase and has niches containing the standing images of the Buddha, Indra, Brahma and a fourth figure, variously thought to represent a bodhisattva or a donor. The billon coins found with the reliquary were issued posthumously in the name of Azes and date from the time of Kujula Kadphises (c. mid-1st century ad). Kanishka, who convened a Buddhist council in Kashmir to settle doctrinal disputes, issued coins in both gold and copper bearing an image of the Buddha. It was, however, a rare coinage, perhaps intended simply to commemorate the council. In spite of Kanishka’s patronage of Buddhism, his dynastic shrine at Surkh Kotal provides evidence of an indigenous religion associated with the cult of fire. His successor, Huvishka, shows the same eclectic approach in his coinage, but later Kushanas reverted to the more limited representation of the Hindu god Shiva with his mount Nandi and the enthroned Ardochsho, the Iranian goddess of good fortune, increasingly represented as her Hindu equivalent, Lakshmi. Kushana and Kushano-Sasanian kings are often depicted on coins with flames issuing from their shoulders, a divine symbol in Iranian terms of a universal ruler. A relief of the 3rd century ad from Shotorak similarly depicts the Dipankara Buddha (the first of 24 predecessors of the historical Buddha) as a substantially larger figure with flames rising from his shoulders (see Snellgrove, fig.; see also §II, 1, (ii), (c) below).

The two colossal images (h. 55 m and 38 m) in rock-cut niches at Bamiyan were clearly intended to inspire respect and show the Buddha as Lord of the world. Paintings in the niche containing the 38 m figure depict Sasanian donors and Buddhas, with a solar divinity in a quadriga above to indicate Buddha’s solar character (see Tarzi, pls B8–27). The larger figure is set in a niche decorated with a pantheon of bodhisattvas encircling the cosmic Buddha (see Tarzi, pls B119–129). A late 7th-century ad Buddha statue from the monastery at Fondukistan is depicted wearing heavy earrings and a jewelled chasuble over his monastic garment, i.e. the ascetic Buddha transformed into the glorified transcendent form also seen in India in the later iconography of Maitreya (the future Buddha; see also §II, 1, (ii), (d)).

The ascendancy of Hinduism in the 7th–8th centuries ad, associated with the Hindu Shahi kings of Kabul, is represented by statues of the elephant god Ganesha from Koh Daman and Gardez; fragments of sculpture representing Shiva and his consort Durga killing the buffalo demon from Sa‛robi; and two Surya images from Khair Khana near Kabul. Islam, brought by the advancing Muslim armies from the 8th century ad onwards, became the dominant religion of Afghanistan by the 10th century.


  • B. Rowland: The Evolution of the Buddha Image (New York, 1963)
  • B. Y. Stavisky: Kushanskaya Baktriya: Problemy istorii i kul’tury (Moscow, 1977); Fr. trans. as La Bactriane sous les Kushans: Problèmes d’histoire et de culture (Paris, 1986)
  • Z. Tarzi: L’Architecture et le décor rupestre des grottes de Bāmiyān, 2 vols (Paris, 1977)
  • F. R. Allchin and N. Hammond, eds: The Archaeology of Afghanistan: From Earliest Times to the Timurid Period (London and New York, 1978)
  • D. L. Snellgrove, ed.: The Image of the Buddha (London, 1978)
  • V. Sarianidi: The Golden Hoard of Bactria from the Tillya-tepe Excavations in Northern Afghanistan (New York and Leningrad, 1985)
  • G. Fussman: ‘Numismatic and Epigraphic Evidence for the Chronology of Early Gandharan Art’, Investigating Indian Art, ed. M. Yaldiz (Berlin, 1987), pp. 67–88
  • The Crossroads of Asia: Transformation in Image and Symbol in the Art of Ancient Afghanistan and Pakistan (exh. cat., ed. E. Errington and J. Cribb; Cambridge, Fitzwilliam, 1992)
  • J. Harmatta: ‘Religion in the Kushan Empire’, The Development of Sedentary and Nomadic Civilizations, 700 bc to ad 250, ed. J. Harmatta, G. F. Etemadi and B. N. Puri (1993), ii of UNESCO History of Civilizations of Central Asia (Paris, 1992–)

II. Historical survey.

The study of art in Afghanistan can conveniently be divided into three broad periods. The rich and varied traditions of early historic Afghanistan gave way in the late 1st millennium ad to the influence of Islam, which dominated artistic output for 1000 years. Twentieth-century art has tended to imitate Western styles.

1. Before 900.

(i) Architecture.

At different periods in its history, Afghanistan has been subject to Harappan, Greek, Persian, Indian, Central Asian, Chinese, Islamic and Russian influences. The architecture of Afghanistan is the syncretic fusion of these divergent traditions, the origins of which can be traced in the prehistoric, Achaemenid and Hellenistic periods, culminating in the 1st millennium ad in the predominantly Buddhist architecture of Gandhara (see also Indian subcontinent, §III, 3, (ii), (a)).

(a) Prehistoric.

Excavations at the Bronze and Iron Age town of Mundigak (founded c. 4000 bc) near Kandahar revealed a long history of simple mud and mud-brick domestic architecture. Monumental town walls were constructed in Period IV (later 3rd millennium bc), a date contemporary with the Urban Phase of the Harappan civilization at the great Indus Valley cities of Mohenjo-daro and Harappa to the east. Two buildings, tentatively identified as a ‘Palace’ and a ‘Temple’, have massive exterior walls and buttresses built of fired brick with a surface coating of plaster. The ‘Palace’, located on the highest point of the site, has a façade on the north-west side, surviving some 35 m in length and nearly 3 m in height. The line of semicircular buttresses or ‘colonnade of pilasters’ fronting this façade was capped by a frieze of stepped merlons. The interior plan, in contrast to the regularity and solidity of the exterior, appears haphazard and comprises small rooms of flimsy mud-brick construction around a central courtyard. Much, however, had been lost through erosion, particularly since the building was first excavated in the 1950s. The ‘Temple’ is more regular in plan, with a façade of decorative triangular buttresses traceable on three sides, although erosion had reduced the height of the building almost to foundation level.

When extant, these buttressed façades must have presented a very impressive aspect. This type of structure appears to have been a favourite form of architectural adornment, as even the ramparts are distinguished by more closely spaced, square buttresses than would be required for either structural or defensive reasons. The Late Bronze Age palace at Dashli 3 in Bactria (northern Afghanistan) is similarly decorated with regular lines of square buttresses, on both the interior and exterior façades. The form recalls late 4th-millennium bc monumental architecture at Uruk (see Mesopotamia, §II, 3), where semicircular, triangular and square buttressing were all used to decorate exterior façades. The semicircular buttresses of the ‘Palace’ at Mundigak seem to have a particularly close affinity with the ‘Pillar Hall’ at Uruk, while the merlons perhaps recall Mesopotamian prototypes.

The tradition of monumental building continued in the 1st millennium bc. Massive structures, usually on immense brick platforms, include Nad-i Ali in Sistan, south-west Afghanistan, the citadels at Maiwand and Kandahar in the south-east, and fortifications at Altin 1, Altin Dilyar Tepe, Kutlug Tepe, Dilberdjin and Dashli in Bactria to the north (see Central Asia, §I, 2, (i), (a)). The same monumentality is also found in Iron Age and Achaemenid buildings of Pakistan (e.g. Bannu) and Iran (e.g. Tureng Tepe and the immense palace platform of Persepolis).

Altin 1 and Altin Dilyar Tepe in Bactria are both fortified towns with a high citadel surrounded by a town and massive outer defensive walls. This type of urban layout was to characterize Central Asian town planning for several millennia. The ramparts of Altin Dilyar are, furthermore, in the form of an immense circle, a feature of subsequent Parthian and Sasanian town planning that was still utilized in ad 762 at Baghdad. The Achaemenid citadel at Dilberdjin is also circular (though the surrounding town is largely later), as is a building at Kutlug Tepe, tentatively identified as a temple, that comprises three massive concentric mud walls pierced by embrasures.

Altin 1 was probably the administrative centre for a group of settlements. The most significant settlement was Altin 10, which contained two buildings, probably palaces, the one consisting of two porticoed courtyards with roofs supported by massive brick pillars, the other comprising a single courtyard dominated by a wide central entrance on one side. Both features subsequently evolved into two important architectural elements of the Middle East: the columned hall and the monumental portal.

The 1st millennium bc, therefore, was one in which the prototypes of certain basic architectural features were established: monumental buildings, the circular form, the columned hall and the monumental portal. That so many standard architectural elements of Afghanistan and neighbouring Iran were formulated in Afghanistan assumes significance in the light of the Central Asian origins of the Iranians themselves: when they arrived on the Iranian plateau, they already had a developed and vigorous architectural tradition.

(b) c. 4th–c. 1st century bc.

New architectural traditions from the west arrived with the conquests of Alexander of Macedon and his Seleucid successors in the 4th–3rd century bc. The survival of the subsequent Greco-Bactrian and Indo-Greek kingdoms in the Central Asian and Indo-Iranian borderlands ensured the continuation and development of an independent Hellenistic artistic and architectural tradition that was open to more Oriental influences.

The only Greek city that has been excavated in Afghanistan is Ai Khanum (probably anc. Alexandria Oxiana). The city, founded at or soon after the time of Alexander’s conquest (329–325 bc), occupies a naturally fortified position at the confluence of the Amu and Kokcha rivers. A low ridge on the third side is further reinforced by ramparts and, in places, a ditch. A north–south main thoroughfare divided the upper and lower town. The upper town has a necropolis and a citadel and is separated from the lower town by a main thoroughfare running from north to south. The lower town consists of three distinctive parts: an almost empty area to the north; a central administrative quarter and a palace; and living quarters to the south. The adjacent plain has an extensive irrigation system. The remains belong well within standard Hellenistic architectural traditions.

The earliest building is the temple dedicated to Kineas (c. 325–c. 300 bc), a leading citizen and probably one of the founders of the city. Other excavated buildings include a gymnasium, theatre, monumental mausoleum, propylaeum, monumental fountain, arsenal, several temples and some houses. The architectural decoration is typically Hellenistic, with stone columns bearing Corinthian, Ionic and Doric capitals throughout. A number of Oriental architectural features were nevertheless retained. Mud-brick was the primary building material, with stone used only for important structural purposes, such as columns. Even many of the Corinthian columns are Asiatic variations of the form, rather than true Greek. The great courtyard of the palace is dominated in the centre of one side by a monumental opening and reception hall supported by three rows of six columns each, in which it is easy to see the local elements of monumental portal and columned hall. While the mausoleum is a standard peripteral style temple, the Temple of Kineas and the temple ‘à redans’ or ‘à niches indentées’ are each situated in a large enclosure, or temenos. Such large temple enclosures are typical Asian features (see also Central Asia, §I, 2, (i), (a)). The layout of Ai Khanum conforms to standard Hellenistic town planning principles, with its grid system of streets and main north–south thoroughfare, though the division into citadel and lower town might be seen as a local influence. But elsewhere in Hellenistic Bactria, the Iranian circular plan remained popular: for example Jiga Tepe is a circular fortification, while Emshi Tepe is an entire circular city, like the earlier site of Altin Dilyar.

Evidence of major conflagrations at Ai Khanum in about 130 bc and again in 90 bc point to the final destruction of the city by invaders, possibly the Kushanas. But the unique blend of western and eastern elements found at the site continued to influence the subsequent cultures of Afghanistan for almost a millennium. Before the discovery and excavation of Ai Khanum in the 1960s, knowledge of this important formative period when the two cultures first became fused was almost non-existent.

(c) c. 1st–c. 5th century ad.

The arrival of the originally nomadic Kushana from Central Asia added a new element to the already eclectic nature of the architecture of Afghanistan. The establishment of the Kushana empire invigorated existing traditions and re-asserted international connections, particularly with the Indian subcontinent, thereby encouraging the spread of Buddhism into Afghanistan. Under the Kushanas, the widely divergent traditions of architecture in Afghanistan became a cultural and political whole. The first Kushana political capital was at Begram, north of Kabul. Subsequently the capital was moved to Taxila in Pakistan and eventually to Mathura in India. The site of Khalchayan in Tajikistan has been identified by its excavators as an early Kushana dynastic shrine. The Kushanas established a similar dynastic cult centre at Surkh Kotal, north of the Hindu Kush. The complex, dated c. 1st–c. 4th century ad, is reached by a monumental mud-brick and masonry staircase, flanked by four massive terraces cut from the hillside. The main temple has a cella facing east, which is enclosed by a corridor on the three other sides and contains a square masonry platform with the remains of four Hellenistic column bases at each corner. The building was first thought to be a fire temple, but more recent research suggests that it may have had a more public ceremonial function. Abutting the south side of the temple is a later complex of two small fire temples linked by a courtyard. These each have a square central sanctuary enclosed by corridors. A Greek-letter inscription in the Bactrian language describes the construction of a well and the restoration of the complex by an official called Nokonzok in the year 31 of the Kanishka era (c. first half of the 2nd century).

The great staircase of Surkh Kotal is unique in the architecture of the region. The only contemporary parallel occurs at Wadi Hadhramaut in southern Arabia, where the monumental staircase approaches to a series of shrines bear many similarities to Surkh Kotal, although there is no apparent connection between the two sites. The associations with fire worship at Surkh Kotal have obvious Zoroastrian affinities, while the ambulatory in the form of corridors around the sanctuary is an essential feature of Buddhist architecture at sites such as Adzhina Tepe. The syncretic fusion of different religious ideas and elements into a cult focused on the person of the emperor was apparently intended to symbolize the unity of different regions of the empire and the cultural tolerance of the Kushanas.

During the Kushana period Buddhism was the dominant religion in Afghanistan. The main architectural manifestation of Buddhism, the Stupa, is one of the most distinctive architectural features of eastern Afghanistan. In the long, complex evolution of the stupa form, Afghanistan played a major role, not least because contact with Hellenism gave rise to the distinctive art and architectural style of Gandhara. Many new forms of Buddhist architecture evolved, the Buddha image was given artistic and architectural expression, and numerous monastic complexes were built in the regions east of Kandahar in the south to Balkh in the north. Most importantly, it was during this formative period that Buddhism and its associated architecture spread from Afghanistan northwards into Central Asia and ultimately eastwards to China, Japan and South-east Asia, rather than directly from the Indian heartland.

The stupa, originally built to house relics associated with the Buddha, became both the centre for great monastic communities and the symbol of far-reaching philosophical ideas. In eastern Afghanistan the stupas number in hundreds, if not thousands. A well-preserved and architecturally outstanding example of the stupa form is found at Guldara, south-east of Kabul. The stupa consists of a dome and two drums positioned on a high, square platform and socle with a stairway on the south-west side. The core of the structure is faced with the distinctive ‘diaper masonry’ that is the hallmark of architecture of the Gandharan period throughout the region: large stones, dressed on one side with small, flat stones filling the interstices. The platform and both stupa drums are decorated with blind arcades of Indo-Corinthian pilasters (an Oriental variation of the Greek prototype). A fortified monastery to the north comprises a central unexcavated courtyard, which possibly contained a central stupa and smaller votive stupas surrounded on four sides by an ambulatory and cells for monks. To the south of the principal stupa stands a second stupa of similar style also constructed of diaper masonry. The complex is dated on stylistic, epigraphic and numismatic grounds to the 2nd–4th century.

The Buddhist sites of Hadda, near Jalalabad in eastern Afghanistan, comprise numerous stupas, monasteries and artificial cave complexes extending over an area of approximately 39 sq. km. The major monastic complexes often have more than one large stupa and always have numerous votive stupas. The stupas, shrines and caves were decorated with a wealth of clay, stone and stucco sculptural decoration and wall paintings (see §II, 1, (ii), (c) below). The most spectacular sites were Tepe Kalan (three main stupas surrounded by dozens of decorated votive stupas) and Tepe Shotor (one main stupa with many votive stupas, richly decorated niches and shrines). All the excavated sites at Hadda have been destroyed, either soon after excavation in the 1920s or during fighting in the 1980s.

Although Afghanistan was incorporated into the Sasanian empire in the 3rd century ad, under the Kushano-Sasanian rulers local cultural forms continued uninterrupted, albeit evincing increasing Sasanian influence. The Hephthalites who migrated southwards from Central Asia in the 5th century are often associated with great destruction, but in Afghanistan the Buddhist civilization not only continued but flourished following their invasion. Indeed, it is often difficult to distinguish any differences in the material culture of the Kushana, Sasanian and Hephthalite periods.

(d) c. 6th–c. 9th century ad.
  • W. Ball

The numerous minor principalities that emerged in eastern Afghanistan during the 7th century ad were often little more than city states ruled by Hephthalite or Turkish lords who enjoyed considerable independence and frequently encouraged the religious arts. Although there was no great central unifying power, Buddhism imposed a cultural unity throughout, and some of the greatest Buddhist monuments were built during this period. At Fondukistan, for example, the monastery contains a courtyard with niches that were elaborately decorated with painted clay sculptures and frescoes. Near Ghazna (probably the site of a minor principality), excavations also revealed a stupa and monastery complex at Sardar, Tepe that was almost Baroque in the richness of its sculptural decoration. The main stupa, the largest in Afghanistan, is surrounded by many votive stupas and chapels, richly decorated in clay reliefs. The remains in the sanctuary include clay fragments of several colossal Buddha statues. Most significantly, there is also a Brahmanical shrine in the Tepe Sardar complex, where a statue of Durga Mahishasuramardini was found. Following the establishment of the Hindu Shahis at Kabul in the 8th century ad, many significant Hindu works of art were created, but there are no great works of architecture belonging to this period, except perhaps the controversial Minar-i Chakri, a stone pillar on a mountain overlooking the Kabul valley.

By far the most spectacular Buddhist site in Afghanistan is Bamiyan, in the central Hindu Kush. The cliffs on the north side of the valley are honeycombed for a length of about 1800 m by some 750 artificial caves, forming part of a Buddhist centre. Among the caves are two colossal, almost free-standing statues of the Buddha (55 m and 38 m high) and a smaller, seated Buddha, carved out of niches in the cliff face in high relief. Details of the robes are applied in stucco over a framework of ropes (representing the folds) attached to the bedrock with wooden pegs. The top halves of the faces have been destroyed by iconoclasts. In the niches surrounding the two standing Buddhas are the remains of frescoes. Frescoes and sculptures also decorate many of the caves, and plaster applied to the bedrock to represent architectural details often imitates wooden prototypes. Indeed, rock-cut architecture is characteristic of Bamiyan, in contrast with other eastern regions of Afghanistan, where architectural structures, particularly stupas, predominate. The only stupa discovered at Bamiyan was excavated to the east of the 38 m Buddha, though mounds of debris might indicate other examples.

In the architecture and art generally of Bamiyan there exists a surprising paradox. For Bamiyan was only a minor principality of the federation known as the Empire of the Western Turks, the capital of which was at Qunduz in north-eastern Afghanistan, where no monuments on such a scale exist. It has been suggested that Bamiyan may have been a dynastic centre for the Western Turks, in much the same way as Surkh Kotal functioned for the Kushanas. Whether or not this is true, Buddhist architecture reached a peak at Bamiyan. The caves cut out of the mountainside housed thousands of devotees; the colossal statues were the ultimate embodiment of the Buddha image, while the painting and sculpture combined Hellenistic, Iranian and Indian elements that influenced subsequent Chinese and Islamic art.


  • J. Barthoux: Les Fouilles de Haḍḍa, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 4–5 (Paris and Brussels, 1930–33)
  • J. Hackin and J. Carl: Recherches archéologiques au col de Khair Khaneh près de Kâbul, Mém.: Dél. Archéol. Fr. Afghanistan, 7 (Paris, 1936)
  • J. Meunié: Shotorak, Mém.: Dél. Archéol. Fr. Afghanistan, 10 (Paris, 1942)
  • J. Carl, J. Hackin and J. Meunié: Diverses recherches archéologiques en Afghanistan, 1933–1940, Mém.: Dél. Archéol. Fr. Afghanistan, 8 (Paris, 1959)
  • J.-M. Casal: Fouilles de Mundigak, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 17 (Paris, 1961)
  • S. Mizuno, ed.: Haibak and Kashmir-Smast: Buddhist Cave Temples in Afghanistan and Pakistan Surveyed in 1960 (Kyoto, 1962)
  • B. Dagens, M. Le Berre and D. Schlumberger: Monuments préislamiques d’Afghanistan, Mém.: Dél. Archéol. Fr. Afghanistan, 19 (Paris, 1964)
  • S. Mizuno, ed.: Durman Tepe and Lalma: Buddhist Sites in Afghanistan Surveyed in 1963–1965 (Kyoto, 1968)
  • S. Mizuno, ed.: Chagalag Tepe: Fortified Village in North Afghanistan Excavated in 1964–1967 (Kyoto, 1970)
  • S. Mizuno, ed.: Basawal and Jelalabad-Kabul: Buddhist Cave Temples and Topes in South-east Afghanistan Surveyed Mainly in 1965 (Kyoto, 1971)
  • P. Bernard and others: Fouilles d’Aï Khanoum, 9 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 21, 26–31, 33 (Paris, 1973–92)
  • I. T. Kruglikova: Dil’berdzhin (raskopki 1970–1972 gg.) [Dilberdjin (excavations 1970–1972)], 1 (Moscow, 1974)
  • G. Fussman and M. Le Berre: Monuments bouddhiques de la région de Caboul, I: Le Monastère de Gul Dara, Mém.: Dél. Archéol. Fr. Afghanistan, 22 (Paris, 1976)
  • I. T. Kruglikova, ed.: Drevnyaya Baktriya [Ancient Bactria], 3 vols (Moscow, 1976–84)
  • Z. Tarzi: ‘Hadda à la lumière des trois dernières campagnes de fouilles de Tapa-é-Shotor (1974–1976)’, Académie des inscriptions et belles-lettres: Comptes rendus des séances (1976), pp. 381–410; Eng. trans. by A. A. Motamedi with C. Grissman as ‘Hadda after the Three Last Seasons of Excavations at Tepe Shotor (1974–1976)’, Afghanistan Quarterly, 32/2 (Sept 1979), pp. 60–89
  • I. T. Kruglikova and G. Pugachenkova: Dil’berdzhin (raskopki 1970–1973 gg.) [Dilberdjin (excavations 1970–1973)], 2 (Moscow, 1977)
  • Z. Tarzi: L’Architecture et le décor rupestre des grottes de Bāmiyān, 2 vols (Paris, 1977)
  • F. R. Allchin and N. Hammond, eds: The Archaeology of Afghanistan: From Earliest Times to the Timurid Period (London and New York, 1978)
  • H. G. Franz: ‘Das Chakri Minar als buddhistische Kultsäule’, Afghanistan Journal, 5/3 (1978), pp. 96–101
  • M. Taddei and G. Verardi: ‘Tapa Sardār: Second Preliminary Report’, East and West, 28 (1978), pp. 33–136
  • H.-P. Francfort: Les Fortifications en Asie centrale de l’âge du bronze à l’époque kouchane (Paris, 1979)
  • W. Ball: Archaeological Gazetteer of Afghanistan/Catalogue de sites archéologiques d’Afghanistan, 2 vols (Paris, 1982)
  • D. Schlumberger, M. Le Berre and G. Fussman: Surkh Kotal en Bactriane, I: Les Temples, architecture, sculpture, inscriptions, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 25 (Paris, 1983)
  • T. Higuchi, ed.: Bamiyan: Art and Archaeological Researches on the Buddhist Cave Temples in Afghanistan, 1970–1978, 4 vols (Kyoto, 1983–4)
  • S. Kuwayama: ‘Tapa Shotor and Lalma: Aspects of Stupa Court at Hadda’, AION, 47 (1987), pp. 153–76
(ii) Sculpture.

The sculpture of Afghanistan reflects the diverse cultural traditions brought by trade and a series of foreign invaders to the region. Stylistic links with Iran and the Indian subcontinent are already evident in the prehistoric period. From the late 4th century bc onwards, the predominant Hellenistic and subsequent Buddhist influences formed the basis of Gandharan art (c. 1st–5th century ad); later sculptures provide increasing evidence of the resurgence of Hinduism. Apart from a few free-standing statues, the sculpture primarily comprises friezes and relief images that were designed to be placed against a wall. Stone, particularly schist and limestone, was used until c. 3rd century ad; stucco and particularly unbaked clay were also used extensively throughout all periods, later to the almost total exclusion of stone. A proportion of the finds from the French excavations at Bamiyan, Begram, Hadda, the Kapisa sites and Fondukistan are in Paris, Musée Guimet. Tepe Shotor, Hadda, was enclosed as a protected archaeological site, with the sculptures retained in situ (since destroyed). The Bamiyan rock-cut images and large items such as the columns and capitals at Ai Khanum were also left in situ, but the majority of sculptures were placed in the Kabul Museum (destroyed by bombing in 1993).

(a) Prehistoric.
  • Gregory L. Possehl

A carving of a human head (c. 635×381 mm) on a soft limestone pebble was found during the 1965–6 excavation of the cave Aq Kupruk II (Ghar-i Mar or ‘Horse Cave’) on the Balkh River, south of Mazar-e Sharif (see Dupree, 1968, 1972; Marshak, figs 138–40). Despite the crude carving, the eyes are clearly depicted and the curve of a distorted mouth can also be seen. This object is associated with an Upper Palaeolithic blade industry (Kuprukian A) and can tentatively be dated 20,000–15,000 bp. The site of Mundigak on the upper reaches of the Helmand River near Kandahar became a major centre in Period IV (c. 2500–c. 2000 bc; see §II, 1, (i), (a) above). Finds from this period include terracotta figurines, compartmented seals and copper or bronze tools (see Shaffer, figs 3.39–40). A male head of white limestone (h. 915 mm), broken high at the neck and assigned to Period IV, 3, was found in a complex of rooms associated with a niched wall (see Casal, pp. 76–7, 255; pls XLIII–XLIV). It is a rather colourless representation of a beardless man with a full (now broken) nose. The chin has been defaced and only faint traces remain of a mouth. Ears are shown as simple ‘C’ shapes, much like those on Mohenjo-daro sculptures. The eyes are oval and rather large, could not have accommodated inlays and have prominent eyebrows above which a hairline is indicated. There is a slight indication that the hair may have been parted in the middle and worn wide and full at the shoulders. There is also a fillet around the forehead, extending down the back of the head in twin flat bands, just as with the ‘Priest–King’ (see §II, 1, (ii), (a)). This piece of sculpture fits well within the corpus of Harappan material in terms of size, material and some stylistic features.

An Afghan tribal leader from Sistan owns a small stone head said to come from the environs of the village of Khwabgah (see Dales, p. 219). This is also broken at the neck and survives to a height of 94 mm. It is fashioned from a soft, creamy buff stone, with many pits and white veins. The beardless head has a small, tight mouth above a small chin and lacks vivacity. The nose comes from a steeply slanted forehead in a rather direct way and is slightly broken. Large, ovate eyes, capped by distinctive eyebrows, seem to have been without inlay. There is a prominent hairline, which could denote the original existence of a cap or some other headgear. A slight indentation towards the front of the top of the head may indicate a central parting. Details are lacking, but the hair seems to have been worn wide and full at the shoulders, as on the Mundigak head. A fillet goes around the forehead and extends down the back of the head in twin flat bands.

A team of Italian archaeologists exploring Iranian Sistan in 1977 found a small limestone head on the surface of the small site of Chah-i Torogh, c. 15 km south of Shahr-i Sokhta (see Jarrige and Tosi, pp. 131–3). Chah-i Torogh is a Shahr-i Sokhta IV site, dating to the earliest centuries of the 2nd millennium bc. The head, broken at the neck, survives to a height of 35 mm and is thus significantly smaller than the stone sculptures from the Indus sites or Mundigak. The expressionless countenance has a full but broken nose. The ears seem to be oval rather than ‘C’ shapes. The hair appears to have been parted in the middle and held in place with a fillet that hangs down the back in a pair of flat bands as in other examples.


  • J. -M. Casal: Fouilles de Mundigak, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 17 (Paris, 1961)
  • L. Dupree: ‘The Oldest Sculptured Head?’, Natural History, 77/5 (1968), pp. 26–7
  • L. Dupree: ‘Prehistoric Research in Afghanistan (1959–1966)’, Transactions of the American Philosophical Society, 62/4 (1972), pp. 3–84
  • A. Marshak: ‘Aq Kupruk: Art and Symbols’, in L. Dupree: ‘Prehistoric Research in Afghanistan (1959–1966)’, Transactions of the American Philosophical Society, 62/4 (1972), pp. 66–72
  • P. Basaglia and others: La città bruciata del deserto salato (Venice, 1977)
  • J. G. Shaffer: ‘The Later Prehistoric Periods’, The Archaeology of Afghanistan: From Earliest Times to the Timurid Period, ed. F. R. Allchin and N. Hammond (London and New York, 1978), pp. 71–186
  • C. Jarrige and M. Tosi: ‘The Natural Resources of Mundigak: Some Observations on the Location of the Site in Relation to its Economic Space’, South Asian Archaeology, 1979, ed. H. Härtel (Berlin, 1981), pp. 115–42
  • V. M. Masson: Altyn Depe (Leningrad, 1981); Eng. trans. by H. N. Michael (Philadelphia, 1988)
  • G. F. Dales: ‘Stone Sculpture from the Protohistoric Helmand Civilization, Afghanistan’, Orientalia Iosephi Tucci memoriae dicta, 2 (Rome, 1985), pp. 219–24
  • P. Amiet: L’Age des échanges interiraniens, 3500–1700 avant J.-C. (Paris, 1986)
(b) c. 4th–c. 1st century bc.

The earliest historical sculptural production from the Afghanistan region is remarkably Hellenistic in both style and content. Until the 1960s Greco-Bractrian art was known only from large numbers of coins, as all attempts to locate the Greek cities of the region had been unsuccessful. The discovery of Ai Khanum (Alexandria Oxiana) at the confluence of the Kokcha and Amu (Oxus) rivers and subsequent excavations at the site by the Délégation Archéologique Française en Afghanistan (1965–78) dramatically increased knowledge of the period from 330 to c. 100 bc. Apart from a few metalwork pieces (see §II, 1, (iv), (c) below), the most impressive sculptures from Ai Khanum are of limestone or marble.

A hermlike bust of a bearded man (§II, 1, (ii), (b)) from the northern square of the gymnasium is dated archaeologically to Phase II (c. 150 bc), although the palaeographic characteristics of the dedicatory inscription point to an earlier date (see Veuve, pls 52–3). Though the bust resembles images of Herakles or the bearded Hermes, the strong individual characterization of the facial features suggests an actual portrait, perhaps of a gymnastic master, a typical Hellenic character.

The largest number of sculptures came from the temple ‘à redans’ (or ‘à niches indentées’) and its sanctuary. These sculptures included the colossal left foot of an acrolithic cult image; the unfinished statuette of a naked young athlete that is Lysippan in its posture and treatment of volumes (see Lysippos); an unbaked clay female head and a stucco male head from statues that stood on either side of the cella entrance; an unbaked clay mould of a female bust (see Francfort, pl. XVII); and two ivory or bone figurines of an unknown naked goddess, one of which was perhaps a puppet, with movable jointed forearms (see Francfort, pl. V). The naked athlete and foot from the cult image (see Stavisky, pls VIIIb–IXa) are unmistakably Hellenistic in style and typology. The evidence provided by the athlete is particularly important, since the piece is unfinished and therefore certainly a local product. Even the two stucco and clay heads and the female bust are definitely in the Hellenistic tradition, though the treatment of their features, especially the rather heavy chins and eyelids, display local characteristics that survived until c. 5th century ad, at such sites as Tepe Sardar (Tapa Sardar) and Tepe Shotor, Hadda. The two ivory goddesses, although presumably derived from Mesopotamian prototypes, appear most closely comparable to Indian examples.

A lion-head gargoyle, perhaps from a fountain in the same sanctuary, is Greek in type but shows such pronounced ‘provincial’ features that it may be linked stylistically with Gandharan lion heads and protomes (see Francfort, p. 91, pl. XL). Other gargoyles from Ai Khanum are the comedy mask (see Stavisky, pl. IXb) and the dolphin and lion protomaí that decorated a fountain on the bank of the Amu River (see Leriche and Thoraval, figs 19–21). Although the latter gargoyles more closely resemble Greek prototypes than the lion-head example from the sanctuary, they are certainly local products and not of such a high quality.

The use of ivory in the Greco-Bactrian period was probably much wider than the finds at Ai Khanum indicate. It has also been suggested that the famous rhyta from Nisa in Turkmenistan were produced in Greek Bactria or in the Indo-Greek region. Afghanistan at the end of the 1st millennium bc was also greatly receptive to surrounding cultures, and it appears that Indian ivories in the Begram treasure may have been imported into the region during the Greco-Bactrian or Indo-Greek period (see also §II, 1, (ii), (c) below and Central Asia, §I, 8, (iv)).

(c) c. 1st–c. 3rd century ad.

Only a few sites in Afghanistan are associated with the art production of the Kushana period, but the archaeological potential of the country is too little exploited for the apparent paucity of finds to be meaningful. The underlying culture is naturally Bactrian Greek, but the region certainly remained open to western influences even in the Kushana period, so that the presence of stylistic or iconographic elements of Hellenistic derivation cannot always be explained as Greco-Bactrian art. At the same time the region received ideas (and, in some instances, perhaps even artefacts) from the eastern Gandharan area, where some of the most vital centres of the ‘Greco-Buddhist’ art were situated: Taxila, the monasteries of Peshawar, Mardan and Swat (e.g. Shah-ji-ki-Dheri, Sahri Bahlol, Butkara). From India came not only new artistic stimuli but also luxury goods destined for customers of high rank. Symbolic of this receptiveness is the treasure from Begram (anc. Kāpishī), north of Kabul, which probably has a terminus date of the 3rd century ad. The hoard includes Indian ivories, Chinese lacquers, Hellenistic bronzes and plaster casts of metalware and Roman glass, which range in date between the 1st century bc and the beginning of the 3rd century ad. The princely owner of the collection seems to have had exquisite artistic interests, for some of the objects, especially the plaster casts (see Hackin and others, figs 274–320), would have had little intrinsic value or worth as status symbols, but great value as models for reproduction, or inspiration for locally manufactured pieces. The most ancient objects of the treasure are almost certainly those from India. It has been suggested that some of the ivories might be the produce of a local Indo-Parthian school. However, although an intermingling of Indian and, to a much lesser extent, Iranian motifs is certainly recognizable in the ivories, it is far more likely that they were all made in an Indian environment (see Hackin and others, figs 1–239). Perhaps the only object really posing a difficult problem of interpretation is the glazed kinnarī-shaped askos, the stylistic details of which are purely Indian (presumably of the first half of the 1st century bc), though the technique of glazing definitely refers to the Parthian world (see Hackin and others, figs 241–2).

In the Jalalabad region of eastern Afghanistan, Gandharan sculpture in stone, presumably attributable to the Kushana period, is indistinguishable from the more widely documented Gandharan output of the north-west Indian subcontinent (present-day north-west Pakistan). Dark grey schist reliefs were discovered at Hadda by the Délégation Archéologique Française en Afghanistan (1926–8), but the stratigraphic context and even the precise site provenance (either Tepe Kafariha or Bagh Gai) of the sculptures are unfortunately unknown (see Dagens, Le Berre and Schlumberger, pp. 11–34, pls I–XXI). Hadda is more famous for later stuccowork, but the discovery of this group of schist sculptures is important because it documents a substantial stylistic unity in the regions to the east and west of the Khyber Pass. However, the question of whether the schist sculptures at Hadda were produced locally or were imported from the Peshawar region, as certain circumstances suggest, is still open to debate. It should be noted that the so-called ‘toilet trays’ that abound in Pakistani sites of the 2nd century bc to at least the 1st century ad are almost completely missing from Afghanistan (see indian subcontinent, §IV, 5(ii)).

Limestone relief of the Conversion of Sundarananda, 521×299×70 mm, from Hadda, 2nd–3rd century ad (London, British Museum); photo © The British Museum

View large

It is certain, however, that the few limestone reliefs discovered at Hadda were locally produced, and they do not show any marked differences in style from the schist group. The formal elements taken from the late Hellenistic repertory can be clearly identified, while the architectural structures that frame the scenes are those of the mature phase of Gandharan sculpture (see fig.). Nor are any stylistic links missing between the limestone reliefs and the stuccowork, even though the latter is mostly attributable to a later period (from the 4th century onwards).

Schist relief of the Miracle of Shravasti, from Paitava, c. 2nd–c. 3rd centuries ad (Paris, Musée Guimet); photo © Allan T. Kohl/AICT

View large

Another more consistent group of schist sculptures was discovered during excavations at Paitava, Shotorak and Kham-e Zargar in Kapisa province, north of Kabul. The work from Kapishi is distinguished by a large number of images of the Buddha standing with flames on his shoulders and water flowing from his feet. The iconography has been related to the so-called Miracle of Shravasti (see fig.). It has also been suggested that the area was one of localized Iranian influence and was therefore strongly receptive to religious phenomena of a luminous nature. There is, however, no reason to consider a specifically Iranian origin for the iconography, since India provides an equally natural ideological context. Another characteristic element of these sculptures is the high occurrence of stiff, frontal compositions, with figures that stand out against the background of the relief, thus eschewing superimpositions and perspective effects, and suggesting a close link with works belonging to the so-called dynastic art of the Kushanas at Surkh Kotal, near Pul-i Khumri, Baghlan province (see Schlumberger, Le Berre and Fussman, pls 58–64). This dynastic art is characterized by images of Kushana kings that are absolutely un-Indian from an iconographic point of view but are connected rather with the nomadic tradition of Central Asia whence the dynasty itself originally emerged. Such royal images (Kabul Mus., see Allchin and Hammond, fig.) have counterparts in the coinage, the Kushana images of Mathura (Uttar Pradesh, India) and in a few other examples.

Stylistic affinities with this ‘dynastic’ production, confirmed by the frequent presence of figures in ‘Kushana’ dress, occur in another group of sculptures that are nevertheless wholly Gandharan in their choice of themes and composition. The sculptures comprise greyish limestone reliefs from the temple of Surkh Kotal (almost exclusively architectural elements, see Schlumberger, Le Berre and Fussman, pls 53–5, 66–8), Sham Qala near Baghlan, Chaqalaq Tepe near Kunduz and from other sites in the same region. The same stylistic characteristics and materials are shared by sculptures from beyond the Amu River in present-day Uzbekistan at such sites as Ayrtam, Kara Tepe and Fayaz Tepe (see Termez, §2). The small sample of clay sculptures from Surkh Kotal seems to be contemporary with the limestone work, even if it is not really possible to date either category with any precision.

Widely diffused stucco and unbaked clay production, belonging at least in part to the Kushana period, is documented from numerous sites at Hadda (e.g. Tepe Kalan, Tepe Kafariha, Bagh Gai, Tepe Shotor), the site of Lalma, due south of Hadda, and Basawal on the Kabul River east of Jalalabad. At Tepe Sardar near Ghazna there is some stuccowork that perhaps dates to the 3rd century ad. Stucco from Afghanistan, particularly the eastern regions, exhibits the same iconographic and stylistic characteristics as work in the same medium from north-west Pakistan. The narrative scenes tend to disappear and isolated images of the Buddha and bodhisattvas predominate, repeated on the walls of the stupas to signify the multiplication of both historical and transcendental Buddhas, a visual expression of the divine experience as professed in the doctrine of Mahayana Buddhism.

Excavations from the 1960s onwards have revealed unbaked clay sculptures, especially at the Buddhist sites of Tepe Sardar, Tepe Maranjan near Kabul and Tepe Shotor II–IV at Hadda. Minor discoveries have also been made at Basawal, Guldara, Bamiyan and Dilberdjin. Only a part of this production falls within the 2nd and 3rd centuries ad, but some of the early clay sculptures are real masterpieces. The same stylistic tendency already encountered in stucco is also visible in the clay images, but coupled with a genuine revival of the Hellenistic style. The immediate impression is that the Mediterranean ‘sources’ are more cultured and courtly than those that formed the earlier basis for so many Gandharan images in schist. The Vajrapani–Herakles and the Tyche of Niche V2 in the ‘Great Vihāra’ of Tepe Shotor (2nd–3rd century), for example, exhibit the reappearance after many centuries of iconographic types and stylistic forms that seem to have their roots in the Hellenism of the Greco-Bactrian and Indo-Greek kingdoms.

(d) c. 4th–c. 9th century ad.

The quantity of Gandharan stuccowork, discovered particularly at Hadda in the 1920s, has been studied in detail, but no satisfactory chronological classification has yet been reached. It is, however, generally held that the majority of these stuccos must be later than the 3rd century ad. The fragments of stucco ornamentation, mainly from stupas, depict images of the Buddha, bodhisattvas and, more rarely, narrative scenes. The images were made by applying progressively refined layers of plaster to an inner core. Colour applied to surface details is still preserved on many examples.

The earliest phase of stucco production reworked images derived from Hellenistic stylistic prototypes, but there was a simultaneous development towards a greater idealization of the human figure. The Buddha image in particular lost many realistic traits in order to make room for an elegantly balanced play of simplified volumes, but certain subsidiary figures still retained the vivacity of Hellenistic naturalism. Clay images from the Buddhist sites follow the same stylistic trend, but display a greater preference for stylistic conventionalization and are often more ‘Indianized’. Meaningful stylistic comparisons can be made with late examples from Taxila and the random collection of heads from Akhnur in Kashmir (see indian subcontinent, §IV, 6(ii)). A relatively late date for the work is affirmed not only by the archaeological evidence but also by the appearance of many iconographic and stylistic elements reflecting the ideal of serenity that formed the basis of Indian art in the Gupta period: the facial features of the Buddha image, for example, show a strong reduction to geometric forms, in marked contrast to the fluent richness of the drapery.

Some sculptural groups at Tepe Shotor that also seem to belong to this period on the basis of their technical characteristics demonstrate a capacity for stylization of the human figure that anticipates the solutions of successive centuries. In particular, the ‘aquatic niche’ (?4th century ad), one of the greatest masterpieces of the region, shows a group of figures modelled in the round, including a Buddha (now missing) standing on a lotus flower, against a highly effective and dramatic background of fantastic fish and other aquatic animals amid the eddies of a pool. A date of the 4th–6th century has been suggested for Tepe Shotor V–VII, while the 6th–7th century has been proposed for the few sculptures from Tepe Maranjan. Excavations at Sardar, Tepe, however, tend to confirm that this production of unbaked clay work took place between the 4th and the 6th century.

The Tepe Sardar excavations are also important for showing conclusively that the so-called school of terracotta sculpture of the north-west Indian subcontinent and Afghanistan never existed. The sculptures were always, in origin, unbaked clay, any terracotta characteristics being merely the result of a later conflagration of the entire monument. The Tepe Shotor and Tepe Sardar excavations have also permitted a detailed analysis of the techniques employed in the manufacture of clay sculptures. In certain respects the process was similar to that used for stucco. The clay sculptures were often supported by a wooden framework, were also coloured and, occasionally, perhaps during the 5th and 6th centuries ad, gilded. Decorative details, such as jewellery, architectural motifs, curls of hair, and also the heads of small figures, were produced in moulds and then applied to the sculptures.

The period also marks the beginning of a trend towards gigantism in images of the Buddha. Tepe Sardar, in particular, provides eloquent examples; but the phenomenon was not exclusive to sites in Afghanistan, even if it was destined to become more important there (see also Adzhina Tepe, Takht-i-Bahi). Representations of the Buddha increasingly appeared between two or more bodhisattvas of smaller size: an iconographic scheme that was perpetuated even in the most complex representations of the next period. Also perhaps assignable to the 4th–6th century ad is the small head of ‘plâtre au grain grossier, de qualité médiocre’ (i.e. ?stucco, see Hackin, in Carl, Hackin and Meunié, 1959, pp. 19–21, fig.) discovered in a monastery near Kunduz, a beautiful example of the working of Hellenistic motifs, which were subsequently destined to merge into more refined solutions such as those at the 7th-century site of Fondukistan. At Qul-i Nader, Kapisa province, some fragments show the use of yet another technique, also seen at Tepe Shotor, of raw earth covered with painted stucco. Fragments of stucco sculpture from Kama Dakka, between Jalalabad and the Pakistani border, perhaps dating from the 6th–7th century, provide possibly the latest evidence for this phase of artistic production in Afghanistan.

In the second half of the 6th century ad, political and military struggles, including the decline of the Hephthalites, undermined the north-west territories of the Indian subcontinent, formerly the heartland of Gandharan architectural and artistic production. Conversely, Kapisa and the Bamiyan region acquired greater importance, due to the opening of a new road, and thus assumed the function of artistic vanguard. The travel diary of the Buddhist Chinese pilgrim Xuanzang gives some idea of the importance of the kingdom of Bamiyan c. ad 632. The two colossal images of the Buddha (55 m and 38 m high), sculpted out of the rockface at Bamiyan and finished in stucco, are without doubt the most famous works in Afghanistan. For a long time it was thought that the 55 m Buddha (considered the younger of the two) was made between the 5th and 6th centuries, partly because paintings in the surrounding niche recalled subjects also present at Ajanta (see indian subcontinent, §V, 3(i)(a)). The 38 m Buddha was variously thought to date from the 2nd–3rd century or the 4th. A more recent hypothesis suggests that the two Buddhas, whatever the time difference between them, should not be dated much before Xuanzang’s visit, i.e. in the second half of the 6th century, for an undertaking of such importance and influence would only be justified after Bamiyan had emerged as a major centre on an important traffic route.

The 7th century was a turning-point in the artistic production of Afghanistan, both architecturally and in sculpture. Tepe Sardar is a typical example: following the destruction of the ancient sanctuary a partial reconstruction took place, using building techniques that are to a certain degree innovative and sculptural decoration that is also new in its conception. Stratigraphical evidence and stylistic comparisons with Fondukistan place the beginning of these activities in the second half of the 7th century, presumably following one of the Arab raids that began shortly after the middle of the century. Both Fondukistan and Tepe Sardar have yielded a remarkable quantity of unbaked clay sculptures of very high quality and distinctive technique: a core of common clay, prepared round a wooden stick or pole, was covered with an outer layer of modelled red clay, on which colour and/or gilding were applied. The sculptures are innovative in both technique and style, though whether there was an actual gap between this phase of production and the sculptures of the earlier period is uncertain. The use of moulded details greatly increased, not only for parts of the jewellery and clothing but also, ever more prevalently, for curls, locks and tresses of hair and for architectural features such as small pilasters and scroll patterns. Moulds were also commonly used for architectural figural decoration, particularly the small atlantids and numerous small Buddha figures in various postures that decorated the unbaked clay stupas and row of ‘thrones’ along two sides of the main stupa. Stylistically, some patterns drawn from the late Gupta repertory became more flamboyant and the human figures more elongated, the Buddha image being practically the only one occasionally to show an archaizing trend. The sculptural compositions within the niches or chapels are generally more complicated: paradise scenes at Tepe Sardar, for example, comprise colossal images of the Buddha and bodhisattvas in the midst of myriad minor figures. Colour increasingly became an integral part of the sculpture. Although only a few fragments preserve their original polychromy (including gilding), both Fondukistan and Tepe Sardar have yielded enough specimens to provide some idea of the original overall effect, which was presumably closer to the strong contrasts of the earliest known works from the Himalayan regions than to the more elementary chromatism of Gandharan stucco and clay examples of the earlier periods.

In the field of Buddhist iconography, besides the ever-increasing importance of representations that can be labelled ‘paradise’ scenes, there was a greater occurrence, both in sculpture and painting at Tepe Sardar, Fondukistan and Bamiyan, of image types unknown or at least unusual in the earlier period, such as the so-called ‘Buddha paré’ (bejewelled Buddha), and the introduction into Buddhism of deities taken from the Hindu pantheon, such as Durga Mahishasuramardini at Tepe Sardar. Colossal images such as the Parinirvāṇa Buddha at Tepe Sardar (l. 16 m) were constructed by modelling unbaked clay around a brick core, a technique that continued from the preceding period. As the cores of votive stupas were also brick, it is apparent that the sculptures employed a masonry technique derived from architecture, while architectural structures (chiefly the stupas) were finished using typical modelling techniques. Since the discovery in 1934 of the Surya seated in a chariot from Khair Khana (Kabul Mus.), several marble sculptures have come to the light in Afghanistan, chiefly in the areas of Kabul, Tagao and Gardez, as well as in the North-west Frontier Province of Pakistan. The marble sculptures, while showing affinities with Buddhist clay images from Fondukistan and Tepe Sardar, also exhibit some idiosyncratic features, such as more rigidly conventionalized details of hair and clothing (usually Sasanian in origin), while the subjects represented belong, almost without exception, to the Hindu and not the Buddhist repertory. Sculptures in this category include a standing Surya, also from Khair Khana, Kabul; the inscribed Umamaheshvara from Tepe Skandar, c. 30 km north of Kabul; two images of Durga Mahishasuramardini, one from Gardez (Kabul Mus.), the other known as the ‘Scorretti marble’ after its former owner (Rome, Pal. Brancaccio); the Ganesha image with an inscription referring to a Khiṃgāla Uḍḍiyāna Śāhi from Gardez in the Dargah Temple at Pir Rathan Nath, Kabul; another Ganesha, originally from Shakar Dara, in a Hindu temple in the Shur Bazaar, Kabul; and a third seated on a lion-vāhana (vehicle), allegedly from somewhere near Tagao, an area from which other cognate sculptures have also been found.

It has been convincingly suggested that this group of sculptures should be assigned to the period of the Turki Shahi dynasty (7th–8th century ad), while the closely linked marble sculptures primarily from the area of Hund, Pakistan, should be dated later (9th–10th century). It is interesting to observe that the Ghazna area appears to have remained outside the boundaries of this politico-cultural entity, for no Turki Shahi coins were found at Tepe Sardar, nor have any marble sculptures been reported from the region apart from the Brahma stele in the palace of the Ghaznavid ruler Mas‛ud III (reg 1099–1115) and the Jaina stele (Kabul Mus.), presumably from the same site, which were taken as war booty from elsewhere. The production of these sculptures apparently came to an abrupt end during the 9th century, perhaps with the conquest of Ghazna by the Saffarid Ya‛qub b. Layth in 869–70. As the heads from Akhnur were the eastern counterparts of the earlier clay sculptures from such sites as Tepe Sardar and Tepe Maranjan, so also do fragments from Ushkar, again in Kashmir, provide the closest comparisons for the later works from Afghanistan. The stylistic and iconographic elaborations that took place both in the 4th–5th and the 7th–8th centuries in Afghanistan were, moreover, of great importance for contemporary and later developments in Central Asia and China.


  • D. Schlumberger: ‘Les Descendants non-éditerranées de l’art grec’, Syria: Revue d’art oriental et d’archéologie, 37 (1960), pp. 131–66, 254–318
  • J. M. Rosenfield: The Dynastic Arts of the Kushans (Berkeley, 1967)
  • M. Hallade: The Gandhara Style and the Evolution of Buddhist Art (London, 1968)
  • D. Schlumberger: L’Orient hellénisé (Paris, 1970)
  • B. Rowland: The Art of Central Asia (New York, 1974)
  • S. Gaulier, R. Jera-Bezard and M. Maillard: Buddhism in Afghanistan and Central Asia, 2 vols, Iconography of Religions, 13/14 (Leiden, 1976)
  • B. Y. Stavisky: Kushanskaya Baktriya: Problemy istorii i kul’tury (Moscow, 1977); Fr. trans. as La Bactriane sous les Kushans: Problèmes d’histoire et de culture (Paris, 1986)
  • G. A. Pugachenkova: Iskusstvo Baktriy epokhi Kushan [The art of Bactria in the Kushana period] (Moscow, 1979)
  • B. Stawiski: Mittelasien: Kunst der Kuschan (Leipzig, 1979)
  • S. L. Huntington and J. C. Huntington: The Art of Ancient India: Buddhist, Hindu, Jain (New York and Tokyo, 1985), pp. 125–62
  • L. Nehru: Origins of the Gandhāran Style: A Study of Contributory Influences (New Delhi, 1989)
  • S. Kuwayama: Kāpishī–Gandāra shi kenkyū [Studies in the history of Kāpishī–Gandhāra] (Kyoto, 1990)
Archaeological reports
  • J. Barthoux: Les Fouilles de Haḍḍa, III: Figures et figurines, Mém.: Dél. Archéol. Fr. Afghanistan, 6 (Paris and Brussels, 1930)
  • J. Hackin and others: Nouvelles recherches archéologiques à Begram (ancienne Kâpicî), 1939–1940, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 11 (Paris, 1954)
  • J. Hackin, J. Carl and J. Meunié: Diverses recherches archéologiques en Afghanistan, 1933–1940, Mém.: Dél. Archéol. Fr. Afghanistan, 13 (Paris, 1959)
  • B. Dagens, M. Le Berre and D. Schlumberger: Monuments préislamiques d’Afghanistan, Mém.: Dél. Archéol. Fr. Afghanistan, 19 (Paris, 1964)
  • S. Mustamandi and M. Mustamandi: ‘The Excavation of the Afghan Archaeological Mission in Kapisa’, Afghanistan Quarterly, 21/4 (1968), pp. 67–79
  • P. Bernard: ‘Quatrième campagne de fouilles à Aï Khanoum (Bactriane)’, Académie des inscriptions et belles-lettres: Comptes rendus des séances (1969), pp. 313–55
  • M. Mustamandi and S. Mustamandi: ‘Nouvelles fouilles à Haḍḍa (1966–1967) par l’Institut afghan d’archéologie’, Arts asiatiques [prev. pubd as Rev. A. Asiat.], 19 (1969), pp. 15–36
  • P. Bernard: ‘Fouilles de Aï Khanoum (Afghanistan): Campagnes de 1972 et 1973’, Académie des inscriptions et belles-lettres: Comptes rendus des séances (1974), pp. 280–308
  • G. Fussman: ‘Nouvelle découverte à Bamiyan’, Afghanistan Quarterly, 27/2 (1974), pp. 57–78
  • P. Bernard: ‘Campagne de fouilles 1975 à Aï Khanoum (Afghanistan)’, Académie des inscriptions et belles-lettres: Comptes rendus des séances (1976), pp. 287–322
  • P. Leriche and J. Thoraval: ‘La Fontaine du rempart de l’Oxus à Aï Khanoum’, Syria: Revue d’art oriental et d’archéologie, 56 (1979), pp. 170–205
  • V. M. Sokolovsky: ‘Rekonstruktsiya dvukh skul’pturnykh izobrazheniy iz Dil’berdzhina (raskop X)’ [A reconstruction of two sculptural images from Dilberdjin (excavation 10)], Materialy sovetsko–afganskoy arkheologicheskoy ekspeditsii [Findings of the Soviet–Afghan archaeological expedition], ii of Drevnyaya Baktriya [Ancient Bactria] (Moscow, 1979), pp. 113–19
  • D. Schlumberger, M. Le Berre and G. Fussman: Surkh Kotal en Bactriane, I: Les Temples, architecture, sculpture, inscriptions, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 25 (Paris, 1983)
  • H.-P. Francfort: Le Sanctuaire du temple à niches indentées, II: Les Trouvailles, Mém.: Dél. Archéol. Fr. Afghanistan, 27 (1984), iii of Fouilles d’Aï Khanum (Paris, 1973–92)
  • M. Taddei and G. Verardi: ‘Clay Stūpas and Thrones at Tapa Sardār, Ghazni’, Zinbun: Memoirs of the Research Institute for Humanistic Studies, Kyoto University, 20 (1985), pp. 17–32
  • S. Veuve: Le Gymnase: Architecture, céramique, sculpture Mém.: Dél. Archéol. Fr. Afghanistan, 30 (1987), vi of Fouilles d’Aï Khanum (Paris, 1973–92)
  • Z. Tarzi: ‘Tapa-e-Top-e-Kalān (TTK) of Haḍḍa’, South Asian Archaeology, 1987, ed. M. Taddei (Rome, 1990), pp. 707–26
  • Z. Tarzi: Haḍḍa à la lumière des fouilles afghanes de Tapa-é-Shotor et Tapa-é-Top-é-Kalan (diss., U. Strasbourg II, 1991)
Specialist studies
  • J. Hackin: ‘Sculptures gréco-bouddhiques du Kapiça’, Monuments Piot [Fondation Eugène Piot], 28 (1925–6), pp. 35–44
  • A. C. Soper: ‘Aspects of Light Symbolism in Gandharan Sculpture’, Artibus Asiae, 12 (1949), pp. 252–83, 314–30; xiii (1950), pp. 63–85
  • K. Fischer: ‘Gandharan Sculptures from Qunduz and Environs’, Artibus Asiae, 21 (1958), pp. 231–49
  • S. Kuwayama: ‘The Turki Śāhis and Relevant Brahmanical Sculptures in Afghanistan’, East and West, 26 (1976), pp. 375–407
  • M. M. Rhie: ‘Some Aspects of the Relation of 5th-century Chinese Buddha Images with Sculpture from N. India, Pakistan, Afghanistan and Central Asia’, East and West, 26 (1976), pp. 439–61
  • G. Verardi: ‘Gaṇeśa Seated on Lion: A New Śāhi Marble’, East and West, 27 (1977), pp. 277–83
  • P. Bernard and F. Grenet: ‘Découverte d’une statue du dieu solaire Surya dans la région de Caboul’, Studia iranica, 10 (1981), pp. 128–46
  • M. Taddei and G. Verardi: ‘Buddhist Sculptures from Tapa Sardâr, Ghazni’, Parola del passato, 199 (1981), pp. 251–66
  • G. Verardi: ‘Osservazioni sulla coroplastica di epoca kuṣāṇa nel nord-ovest e in Afghanistan in relazione al materiale di Tapa Sardâr’, AION, 43 (1983), spp. 479–504
  • G. Verardi: ‘The Kuṣāṇa Emperors as Cakravartins: Dynastic Art and Cults in India and Central Asia, History of a Theory, Clarifications and Refutations’, East and West, 43 (1983), pp. 225–94
  • C. Mustamandy: ‘Herakles, Ahnherr Alexanders, in einer Plastik aus Hadda’, Aus dem Osten des Alexanderreiches, ed. J. Ozols and V. Thewalt (Cologne, 1984), pp. 176–80
  • M. Taddei: ‘Neue Forschungsbelege zur Gandhāra-Ikonographie’, Aus dem Osten des Alexanderreiches, ed. J. Ozols and V. Thewalt (Cologne, 1984), pp. 154–75
  • K. Tanabe: ‘Iranian Origin of the Gandharan Buddha and Bodhisattva Images’, Bulletin of the Ancient Orient Museum, 6 (1984), pp. 1–27
  • G. Verardi: ‘Gandharan Imagery at Tapa Sardâr’, South Asian Archaeology, 1981, ed. B. Allchin (Cambridge, 1984), pp. 257–62
  • Z. Tarzi: ‘La Technique du modelage en argile en Asie centrale et au nord-ouest de l’Inde sur les Kouchans: La Continuité malgré les ruptures’, Ktema, 11 (1986), pp. 57–93
  • S. Kuwayama: ‘Literary Evidence for Dating the Colossi in Bāmiyān’, Orientalia Iosephi Tucci memoriae dicata, 2, ed. G. Gnoli and L. Lanciotti (Rome, 1987), pp. 703–27
  • D. Klimburg-Salter: ‘Bāmiyān: Recent Research’, East and West, 38 (1988), pp. 305–12
  • M. M. Rhie: ‘Interrelationships between the Buddhist Art of China and the Art of India and Central Asia from 618–755 ad’, AION, suppl. no. 54 (1988)
  • D. Klimburg-Salter: The Kingdom of Bāmiyān: Buddhist Art and Culture of the Hindu Kush (Naples, 1989)
  • S. Kuwayama: ‘L’Inscription du Gaṇeśa de Gardez et la chronologie des Turki-Ṣāhīs’, Journal asiatique, 279 (1991), pp. 267–87
  • M. Taddei: ‘The Bejewelled Buddha and the Mahiṣāsuramardinī: Religion and Political Ideology in Pre-Muslim Afghanistan’, South Asian Archaeology, 1989, ed. C. Jarrige (Madison, 1992), pp. 457–64
  • M. Taddei: ‘La plastica buddhistica in argilla in Afghanistan e nel nordovest del subcontinente indiano’, Oxus: Tesori dell’Asia Centrale (exh. cat., Rome, Pal. Venezia, 1993–4), pp. 118–22
(iii) Painting.

Apart from a few remnants of the Greco-Bactrian period (c. 4th–c. 1st century bc), most of the evidence for painting in Afghanistan dates from the 3rd century ad onwards and comes from the sites of Hadda and Bamiyan. The medium used throughout is tempera. A large proportion of the paintings at Bamiyan was retained in situ; unless otherwise specified, the remaining finds were placed in the Kabul Museum. Following the bombing of the museum in 1993, their fate is uncertain.

(a) c. 4th–c. 1st century bc.

The excavations at Ai Khanum, so rich in architecture and sculpture, provide little direct evidence of figural painting. There is, however, clear evidence that the technique of painting was extensively used. Several architectural pieces, mostly from Rooms 6 and 9 of the Administrative Quarter’s southern complex (mid-2nd century bc), still preserve traces of paint on their whitewashed stone surfaces. These traces show that heart-and-dart motifs were painted on mouldings and different artists’ ‘hands’ are even recognizable (see e.g. Bernard and others, i, pl. 80). Another precious relic of painting, found in the sanctuary of the temple ‘à niches indentées’ (or ‘temple à redans’) at Ai Khanum, is attributed to the earliest phase of the city (4th century bc). The few fragments of painted canvas, pasted on a wooden frame (a typical Greek technique), depict a frieze of walking lions that presumably constituted the background scene for a cult image.

More substantial evidence of painting in Afghanistan only occurs at the end of the Greco-Bactrian period. Flanking either side of the entrance to the ‘temple of the Dioscuri’ at Dilberdjin, north of the Hindu Kush, are the remains of a two-tiered wall painting defined at the bottom by a band of Greek fretwork. The scene depicts two naked young men, one on either side of the door and each accompanied by a horse. The painting, identified by its excavators as a representation of the Dioscuri, belongs to the first phase of the sanctuary (founded c. 150 bc), though it is not necessarily contemporary with the period of construction of the temple. The painting is in a very poor state of preservation; nevertheless it is apparent even from the published line drawings and watercolours that the painter’s cultural background was intensely Greek. According to the excavation reports, the Dioscuri remained visible during the second phase of the sanctuary (dated to the period of Kujula Kadphises, 1st century ad) but were immured behind a new wall during the third phase (period of Kanishka, c. late 1st century to early 2nd). This important point suggests that the Greek religious tradition was not abandoned in this part of the Kushana empire until the time of Kanishka. There is no reason to think of a Vedic connection (the Ashvins) for the two young heroes. The painting technique consisted of the preparation of a loess surface, a sketch of the whole composition in diluted red colour, the use of a comparatively poor range of colours (no blue or green pigments) without chiaroscuro and finally the outlining of the faces and naked bodies in black or brown.

(b) c. 1st–c. 5th century ad.

Afghanistan has yielded a fairly large number of wall paintings of the Kushana period. Close stylistic and technical comparisons are possible between these examples and other specimens found in Pakistan and north of the Amu River, for example at Kara Tepe. Although dating the various paintings is extremely difficult, all these remains bear witness to a highly sophisticated technique, and it is clear that the use of painting was far from exceptional in this period.

If the use of colour on stucco sculpture is not taken into consideration, painting of the Kushana period at Buddhist sites is only known from very poorly preserved fragments at Hadda. Some examples in Vihāra B. 56 at Bagh Gai, Hadda, were ochre sketches that were never completed: the most impressive is a representation of the story of Angulimala, which follows the same iconographic models employed by Gandharan artists in stone reliefs. Another sketch in ochre of a moustached head was found on a stupa at Tepe Shotor, Hadda, in 1966–7. A niche from Tepe Kalan (TK.17; Paris, Mus. Guimet) is a good example of symbiosis of sculpture and painting: a modelled Buddha image is flanked by painted images of donors and surmounted by winged cupids who hold a wreath above his head. The typically Greco-Roman iconography, somewhat reminiscent of 3rd-century ad painting at Palmyra, underwent what might be called a Buddhist interpretation: a certain parallelism is noticeable in the paintings from Miran, Chinese Turkestan, which are also attributed to the 3rd century.

Panel fragment with the God Shiva/Oesho, terracotta, gouache, h. 56.8 cm, w. 52.3 cm, d. 5.4 cm, Kushana, AD 3rd century (New York, Metropolitan Museum of Art, Gift of Isao Kurita, 2000, Accession ID:2000.42.4); photo © The Metropolitan Museum of Art

View large

In the ‘temple of the Dioscuri’ at Dilberdjin a painting belonging to the fourth period of the sanctuary depicts a standing, high-ranking official beside a divine pair seated on a crouching bull. Since the male divine figure is ithyphallic, the pair can be identified as Shiva and his spouse Parvati. Though it is often doubtful whether this and other ‘Hindu’ images from Central Asia and Afghanistan (e.g. the Durga Mahishamardini figure at Tepe Sardar and fragment, 3rd century ad; New York, Met., see fig.) should actually be interpreted as purely Hindu subjects, the derivation from an Indian iconographic model is clear and suggests a date for the painting not earlier than the 4th century ad, though the excavators attribute it to the time of Vasudeva I (3rd century). The painting technique differs from that of the Dioscuri, particularly since it includes the use of chiaroscuro in drapery and perhaps also in the naked parts of the human bodies.

The later north-east religious complex at Dilberdjin contained three rooms with important remains of wall paintings. The painting technique is closer to that of the wall paintings from Central Asia and of the contemporary sites in Afghanistan (e.g. Bamiyan, Fondukistan, Tepe Sardar) insofar as the preparation of the mud-plaster coating and whitewash are concerned. The colour range includes lapis lazuli blue, which also occurs in the Shiva and Parvati painting in the ‘temple of Dioscuri’.

The most problematic painting, in Room 12 of the complex, depicts a helmeted character between attendant figures. It has been suggested that the main figure, seated front view with knees apart in a posture strongly reminiscent of the iconography of Sasanian kings, is a victorious ‘hero’ to whom homage is paid, but others have labelled this character a ‘goddess’. Though the excavators suggest that Dilberdjin was abandoned immediately after the Hephthalite invasion in the 5th century ad, the helmeted figure may be attributed to the 6th–7th century or, if compared with the bodhisattva Maitreya in the soffit of Cave K3 at Bamiyan, late 7th century, or even, if compared with Maitreya in the soffit of Niche E at Bamiyan, dated to the 8th century.

(c) c. 6th–c. 9th century ad.
  • Maurizio Taddei

A discovery in 1976 at Tepe Shotor, Hadda, provided the first impressive document of Buddhist painting south of the Hindu Kush: in a meditation cave at the site, ten disciples of the Buddha are shown seated cross-legged on either side of a skeleton. The exact meaning of this representation is unclear, though the presence in the scroll frieze below the monks of what appear to be floral elements resembling phalli, seems to suggest a Tantric reading of the whole composition. The style can still be placed in the Greco-Buddhist tradition that gave rise to such compositions as the Buddha with monks and trees from Kara Tepe (Uzbekistan; §II, 1, (iii), (c)), attributed to the 2nd–3rd century. The scroll frieze is also very ‘Hellenistic’ in its free naturalism; nevertheless many details show signs of evolution: the heads are comparatively smaller, the drapery more elaborate, the faces less naturalistic and the shape of the eyes reminiscent of late works such as the manuscript covers from Gilgit (Pakistan). The archaeological evidence confirms a 6th-century date.

In the 6th century ad there was a westward shift of the main roads connecting India to Central Asia. The importance of the sites in Gandhara declined while those in the Hindu Kush flourished. The main archaeological centre in the Hindu Kush is Bamiyan, with four minor centres: Foladi and Kakrak, quite close to Bamiyan, Fondukistan, c. 130 km to the east, and Nigar (Dukhtar-i Nushirwan) to the north beyond the Kara Kotal Pass in the direction of Haibak (Samangan). Both Bamiyan and Fondukistan were investigated by the Délégation Archéologique Française en Afghanistan. In the 1960s and 1970s the wall paintings at Bamiyan were surveyed by scholars chiefly from Afghanistan and Japan and extensively restored by an Indian team.

The chronology of Bamiyan has often been the subject of controversy, also in connection with the formative process of various styles of painting, such as Gandharan, Syrian, Iranian (chiefly Sasanian) and Indian (Gupta). The early dating (Kushana period to the 4th–5th century ad) for the colossal Buddhas has now been abandoned, and most scholars are inclined to believe that neither image is earlier than the second half of the 6th century (see §II, 1, (ii), (d)). The paintings decorating their huge niches cannot therefore be any earlier in date. Research (see Klimburg-Salter, 1989) suggests the following working hypotheses: firstly, that the art of the Hindu Kush belongs to a single cultural period (7th–9th century); secondly, that it probably resulted from a consistent form of patronage; and thirdly, that the formal development indicates a possible movement towards a later Mahayana perspective.

The painting technique is common to Buddhist sites throughout Afghanistan and Central Asia. Like the wall paintings in India, those from Bamiyan and other sites in Afghanistan are not frescoes (as they are often unduly styled even in scientific literature), but rather tempera paintings. The wall surface first received a dressing of clay mixed with vegetable fibres (kâhgel), which functionally corresponds to the arriccio of the Italian medieval and later wall paintings. This dressing was coated with a thin white ground layer of burnt gypsum or plaster of Paris, on which the pigments were directly applied in a binding of animal glue. At other sites, such as Tepe Sardar, the ground layer was made of clay mixed with fibres obtained from the inflorescences of a kind of marsh reed (simgel); at Dunhuang kaolin was employed for the same purpose.

A chronological classification of the wall paintings in the Hindu Kush has been attempted (see Klimburg-Salter, 1989) on the basis of style (morphology and representational conventions) and iconography. Phase Ia (mid- to second half of 7th century ad) is represented by paintings in the niche of the 38 m Buddha of Bamiyan; Phase Ib (c. second half of the 7th century) by the niche of the 55 m Buddha; Phase IIa–b (late 7th century–early 8th) by Bamiyan Cave K, Niches H and ‘i’, Fondukistan and Kakrak; Phase IIc–d (8th century or later) by Bamiyan Niche E and Foladi.

Phase Ia appears to be linked to the 7th-century ad Sogdian paintings at Pendzhikent (see also Central Asia, §I, 4, (iv), (a)). Compositions are rather two-dimensional: there is no indication of shading and contours are obtained by using bands of colour separated by brown or black lines. Phase Ib, though essentially a linear style, is richer in shading and sometimes, at least in some subjects, even reflects conventions elaborated at Ajanta in India (see indian subcontinent, §V, 3(i)(a)), though the most significant parallels are with the art of Kizil and Khotan in Chinese Turkestan (see also Central Asia, §II, 4, (ii), (b)).

Phase II can be considered as the typical Bamiyan style, with such clear links with eastern Central Asia, especially Kizil and Kucha, as to suggest that after the Chinese invasion of Kucha c. ad 647, painters from there migrated to Bamiyan. The concurrent Chinese conquest of Kizil (647–8), the subsequent Tibetan conquest and Chinese reconquest perhaps also induced Kizil artists to flee their former large, cosmopolitan and free town. The tendency in Phase II is towards a high conventionalization of forms that abandons all interest in realism and develops into a somewhat linear elegance reminiscent of contemporary sculpture at Fondukistan. Characteristic features are thin double lines employed to define drapery, a swirl under the arms, and the use of contrasting black and white figures. At the end of the development, the human form is reduced to diagrammatic patterns.

Among the most interesting compositions at Bamiyan is one in the niche of the 38 m Buddha, which represents a Pañcavārṣika ceremony (as described by the Chinese Buddhist pilgrim Xuanzang in about ad 632), at precisely the point when the king presents the Buddha with all his worldly possessions. The ideological value of this ceremony is highly significant for understanding the role assumed by the kings of Bamiyan and the relationship between political and religious power. In the vault of the same niche a solar deity (possibly Surya) is also depicted.

The Parinirvāṇa in Cave K3, although poorly preserved, is one of the most impressive representations of this central scene of Buddhist devotion, with an abundance of detail that makes it and the companion painting in the soffit of Maitreya encircled by fifteen roundels, each containing six Buddhas, a rich source of information concerning jewellery, metalwork and (royal) garments (see Tarzi, figs 3–4, 6–12). Presumably the Parinirvāṇa was a ‘reproduction’ of the colossal reclining Buddha c. 300 m long that Xuanzang saw at Bamiyan (see Tarzi). The other Parinirvāṇa scenes at Bamiyan (Caves F, J and 222; see Miyaji, 1978) are not so rich or complicated and are rather reminiscent of traditional Gandharan representations of the same episode.

The remains of painting from more minor sites include the ‘hunter king’ in Sasanian style from Kakrak (Kabul Mus.) and the seated bodhisattva Maitreya from Niche E at Fondukistan. The Maitreya, together with a considerable number of modelled sculptures in unbaked clay in the same sanctuary, provide one of the most reliable links between sculpture and painting in the Hindu Kush region.

Various sites in other parts of Afghanistan preserve the remains of wall paintings. Those found by the Archaeological Mission of Kyoto University in Cave 130 at Basawal near Jalalabad depict a row of pairs of Buddhas facing each other. A similar setting is found in a painting excavated by the Italian Archaeological Mission in Room 52 at Tepe Sardar, Ghazna, where a row of alternating Buddhas and richly bejewelled bodhisattvas is perhaps intended to represent the revelation of Buddhas to bodhisattvas. Stylistically, these and other wall paintings from Tepe Sardar are close to those from Fondukistan, but comparisons are also possible with Khotan, Kizil and Pendzhikent. A dating in the 8th century ad seems quite reasonable.

The painting of pre-Muslim Afghanistan and Iran, even after the sanctuaries had been abandoned or (more rarely) destroyed, had a great impact on both art and literature of the subsequent period. At the end of the 10th century, the anonymous Ḥudūd al-‛ālam refers to the existence at Bamiyan of wall paintings ‘in the Indian style’ (see Melikian-Chirvani, p. 23). Even in the 13th century, Arab travellers still showed a lively interest in Bamiyan and its paintings: a description by Yaqut, of wall paintings depicting birds and other animals, is copied from an earlier account written by Sam‛ani in the mid-11th century (see Melikian-Chirvani, pp. 25–6).


  • A. Godard, Y. Godard and J. Hackin: Les Antiquités bouddhiques de Bāmiyān, Mém.: Dél. Archéol. Fr. Afghanistan, 2 (Paris, 1928)
  • J. Hackin and J. Carl: Nouvelles recherches archéologiques à Bāmiyān, Mém.: Dél. Archéol. Fr. Afghanistan, 3 (Paris, 1933)
  • R. Gettens: ‘Materials in the Wall Paintings of Bāmiyān, Afghanistan’, Technical Studies, 6 (Jan 1938), pp. 186–93
  • B. Rowland and A. Coomaraswamy: The Wall Paintings of India, Central Asia and Ceylon (Boston, 1938)
  • J. Hackin: ‘Le Monastère bouddhique de Fondukistân: Fouilles de J. Carl, 1937’, in J. Hackin, J. Carl and J. Meunié: Diverses recherches archéologiques en Afghanistan, 1933–1940, Mém.: Dél. Archéol. Fr. Afghanistan, 13 (Paris, 1959), pp. 49–58
  • U. Scerrato: ‘A Short Note on Some Recently Discovered Buddhist Grottoes near Bāmiyān’, East and West, 11 (1960), pp. 94–120
  • K. Yamasaki: ‘Saiiki hekiga no ganryo ni tsuite’ [Pigments in the wall paintings of Central Asia], Bijutsu kenkyū [Art research], 212 (1960), pp. 31–3
  • M. Bussagli: Painting of Central Asia (Geneva, 1963)
  • B. Dagens: ‘Monastères rupestres de la vallée de Foladi’, in B. Dagens, M. Le Berre and D. Schlumberger: Monuments préislamiques d’Afghanistan, Mém.: Dél. Archéol. Fr. Afghanistan, 19 (1964), pp. 41–8
  • P. Bernard and others: Fouilles d’Aï Khanoum, I: Campagnes 1965, 1966, 1967, 1968, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 21 (Paris, 1973)
  • A. S. Melikian-Chirvani: ‘L’Evocation littéraire du bouddhisme dans l’Iran musulman’, Le Monde iranien et l’Islam, 2 (1974), pp. 1–72
  • V. P. Buryzh: ‘Tekhnika zhivopisi’ [Painting technique], Materialy sovetsko–afganskoy ekspeditsiy, 1969–1973 gg. [Findings of the Soviet–Afghan expedition], i of Drevnyaya Baktriya [Ancient Bactria] (Moscow, 1976), pp. 111–24
  • I. T. Kruglikova: ‘Nastennye rospisi Dil’berdzhina’ [The wall paintings of Dilberdjin], Materialy sovetsko–afganskoy ekspeditsiy, 1969–1973 gg. [Findings of the Soviet–Afghan expedition], i of Drevnyaya Baktriya [Ancient Bactria] (Moscow, 1976), pp. 87–110
  • A. Miyaji: ‘Wall Paintings of Bāmiyān: A Stylistic Analysis’, Japan–Afghanistan Joint Archaeological Survey in 1974 (Kyoto, 1976), pp. 17–31
  • S. Sengupta: ‘Preservation and Conservation of Bamiyan’, Bamiyan: Crossroads of Culture, ed. A. Miyaji (Tokyo, 1976), pp. 9–15
  • A. Miyaji: ‘The Parinirvāṇa Scenes of Bāmiyān: An Iconographical Analysis’, Japan–Afghanistan Joint Archaeological Survey in 1976 (Kyoto, 1978), pp. 13–22
  • P. Bernard and H.-P. Francfort: ‘Nouvelles découvertes dans la Bactriane afghane’, AION, 39 (1979), pp. 119–48
  • V. P. Buryzh: ‘Tekhnika rospisey pomeshcheniya 16’ [The technique of the paintings of site 16], Materialy sovetsko–afganskoy arkheologicheskoy ekspeditsiy [Findings of the Soviet–Afghan archaeological expedition], ii of Drevnyaya Baktriya [Ancient Bactria] (Moscow, 1979), pp. 146–65
  • I. T. Kruglikova: ‘Nastennye rospisi v pomeshcheniy 16 severo-vostochnogo kul’tovogo kompleksa Dil’berdzhina’ [The wall paintings of site 16 at the north-eastern cultic complex of Dilberdjin], Materialy sovetsko–afganskoy arkheologicheskoy ekspeditsiy [Findings of the Soviet–Afghan archaeological expedition], ii of Drevnyaya Baktriya [Ancient Bactria] (Moscow, 1979), pp. 120–45
  • Z. M. Zhelninskaya and others: ‘Analisy krasok nastennykh rospisey Dil’berdzhina’ [An analysis of the colours of the wall paintings of Dilberdjin], Materialy sovetsko–afganskoy arkheologicheskoy ekspeditsiy [Findings of the Soviet–Afghan archaeological expedition], ii of Drevnyaya Baktriya [Ancient Bactria] (Moscow, 1979), pp. 166–72
  • A. Miyaji: ‘The Wall Paintings of Bamiyan Caves (Continued): Stylistic Analysis’, Japan–Afghanistan Joint Archaeological Survey in 1978 (Kyoto, 1980), pp. 16–26
  • D. Klimburg-Salter: ‘Ritual as Interaction at Bamiyan’, Systems of Communication and Interaction in South Asia, ed. P. Gaeffke and S. Oleksiw (Philadelphia, 1981), pp. 65–9
  • C. Silvi Antonini and M. Taddei: ‘Wall Paintings from Tapa Sardār, Ghazni’, South Asian Archaeology, 1979, ed. H. Härtel (Berlin, 1981), pp. 429–38
  • D. E. Klimburg-Salter, ed.: The Silk Route and the Diamond Path: Esoteric Buddhist Art on the Trans-Himalayan Trade Routes (Los Angeles, 1982)
  • Z. Tarzi: ‘La Grotte K3 de Bâmiyân’, Arts Asiatiques, 38 (1983), pp. 20–29
  • I. T. Kruglikova: Dil’berdzhin: Khram Dioskurov [Dilberdjin: the temple of the Dioscuri] (Moscow, 1986)
  • F. Grenet: ‘L’Athéna de Dil’berdžin’, Cultes et monuments religieux dans l’Asie centrale préislamique, ed. F. Grenet (Paris, 1987), pp. 41–5
  • D. Klimburg-Salter: ‘Dukhtar-i Nushirwan: An Ideology of Kingship’, Kusumañjali: Sh. C. Sivaramamurti Commemoration Volume, 1, ed. N. Rao (New Delhi, 1987), pp. 62–76
  • D. Klimburg-Salter: The Kingdom of Bāmiyān: Buddhist Art and Culture of the Hindu Kush (Naples, 1989) [with app. on conservation by R. Sengupta]
  • M. Mode: ‘The Great God of Dokhtar-e Noshirwān (Nigār)’, East and West, 42 (1992), pp. 473–83
(iv) Other arts.
(a) Coins.
  • D. W. MacDowall

The official coinage of the Achaemenid (see Achaemenid, §2) period (550–331 bc) was the Persian gold Daric and silver siglos based on a gold:silver ratio of 13.3 to 1, but these coins are rarely found in Afghanistan (see Ancient Near East, §II, 8, (i)). Normal currency during the 4th century bc consisted of worn Greek silver coins and their copies. The earliest of these coins had been exported from Athens and other cities of the Greek world as bullion. The hoard discovered in 1966 at Balkh comprised 170 old Greek silver coins of this category, 150 Athenian tetradrachms and coins from 13 other Greek states down to c. 380 bc. The group of Greek coins in the Oxus Treasure had a similar composition. The 1933 hoard from Chaman-i-Hazuri, Kabul, dating from the mid-4th century bc, contained worn Greek silver with subsidiary denominations of bent bar silver coins weighing c. 11.7 gm and punch-marked silver (see Curiel and Schlumberger, pp. 31–45). Other hoards of similar bent bar silver coins with a wheel or sun symbol at each end of the bar have been found at Jalalabad and at Mir Zakah near Gardez (see Curiel and Schlumberger, pp. 65–91), as well as at sites in the Indus Valley such as Taxila. Double the weight of a silver siglos and possibly derived from the bar coinage of Media, they appear to have constituted the silver coinage of the Achaemenid satrapies in north-west India before the invasion of Alexander the Great in 331 bc.

The fine coinage bearing the head of Herakles that Alexander issued was apparently based on a gold:silver ratio of 10 to 1. Another major change was the introduction of an Attic weight silver tetradrachm tariffed at 20 drachmae per gold stater that ensured the circulation of the new silver issues throughout the east as coins, not bullion. After Alexander’s death in 323 bc, Bactria, as a Seleucid province, had a typically Hellenistic coinage with fine obverse portraits of the kings in gold and silver. The Greco-Bactrian kings who subsequently governed Bactria as an independent state after about 250 bc inherited this tradition. The tetradrachms of the first ruler Diodotos and his successor Eukhydemos are well engraved and show the king with a diadem. Demetrios, who invaded India in the early 2nd century bc, is shown with an elephant headdress. Antimachos Theos is depicted with a Greek travelling cap (petasos) and Eucratides with a helmet. The reverse types of the coins show Greek divinities such as Zeus, Herakles, Artemis and the Dioscuri.

Bronze coin of Agathokles showing the goddess Subhadra (obverse) and a panther (reverse), h. 24 mm, from Begram, c. 190–180 bc (London, British Museum); photo © The British Museum

View large

South-east Afghanistan was initially part of the Mauryan empire and used the Mauryan square silver punch-marked coinage and square copper coins. When the Greeks extended their rule to the south (c. 180 bc), they struck a distinctive coinage incorporating features from the existing Mauryan currency for their new Indian provinces. Pantaleon, Agathokles and their successors struck square copper coins with a reverse legend, initially in Brahmi and subsequently in Kharoshthi, which repeated the information given on the Greek obverse (see fig.). Apollodotos I issued square silver drachms modelled on the Mauryan silver denomination, and all the Greek rulers of the regions south of the Hindu Kush struck a bilingual silver coinage on the reduced Indian weight standard. More than 80,000 coins collected in the 1830s by Charles Masson (see §III) at Begram, 60 km north of Kabul, and the hoard found in 1947 at Mir Zakah, 73 km north-east of Gardez, provide excellent evidence for the currency of this period in these regions. The Yueh-chih successors to the Greeks in Bactria from c. 130 bc onwards struck crude copies in base metal of the tetradrachms of Heliokles, the last Greco-Bactrian king, while the invaders of the Kabul Valley struck increasingly crude copies in copper of Hermaios, the last Indo-Greek king of the region. About ad 20 the Indo-Parthian king Gondophares established a powerful empire extending from Sistan across south Afghanistan to the Punjab. His coinage in Arachosia (Kandahar region) combined a striking portrait bust of the king with the Classical figure of Victory on the reverse. Successors to Gondophares in Arachosia and Sistan continued to strike copper tetradrachms until the Sasanian conquest by Ardashir and Shapur I (mid-3rd century ad).

The establishment of the Kushana empire in the 1st century ad led to the introduction of a standard coinage throughout Bactria and the Punjab. The nameless king Soter Megas struck a copper coinage with the radiate head of Mithra (see §I, 4) on the obverse and the mounted king on horseback on the reverse. His successor Vima Kadphises introduced the Kushana gold dinar and a large copper tetradrachm with the distinctive obverse motif of the king standing, sacrificing at a low fire altar. During the 2nd century Kanishka and Huvishka introduced a wide range of reverse types, such as Miiro (sun god), Mao (moon god), Oado (wind god) and Oesho (an Iranian wind god associated with the Hindu god Shiva). Under the later Kushanas the range of divinities was reduced to Shiva and Ardochsho, in an increasingly stylized form.

In the absence of an indigenous silver coinage, Sasanian silver drachms came to play an increasing role in the currency of Afghanistan during the 3rd century ad. The Sasanian coins have a portrait of the king on the obverse and a fire altar flanked by two attendants on the reverse, as do some associated Kushano-Sasanian gold and copper issues of Bactria and Gandhara. During the 4th century Hephthalite invaders of Bactria copied the types of the Sasanian king Shapur II (reg 309–79). Several Sasanian silver drachms from Afghanistan have Hephthalite countermarks and seem to have been coins paid in tribute in the time of Firuz (reg 457/9–84), which were countermarked to serve as Hephthalite currency. Rulers in the Turki Shahi period (early 7th century) followed the same tradition, as can be seen by the silver and copper drachms of Napki Malka and the coins of Vrahitigin. The silver currency of the Hindu Shahis of Kabul (750–1000) used a bull on the obverse and a horseman on the reverse. Most of the coins have an obverse legend in Sharada script, either Sri Spalapati Deva or Sri Samanata Deva (which are titles rather than personal names), and a letter or symbol in the field. The later debased coins of the currency, struck in billon, were extensively copied by the Islamic dynasties of Ghazna, Kanauj, Ajmer and Delhi (see indian Subcontinent, §VII, 6(ii)).


  • H. H. Wilson: Ariana Antiqua: A Descriptive List of the Antiquities and Coins of Afghanistan (London, 1841/R Delhi, 1971)
  • R. Curiel and D. Schlumberger: Trésors monétaires d’Afghanistan, Mém.: Dél. Archéol. Fr. Afghanistan, 14 (Paris, 1953)
  • D. W. MacDowall and M. Taddei: ‘The Early Historic Period: Achaemenids and Greeks’; ‘The Pre-Muslim Period’, The Archaeology of Afghanistan: From Earliest Times to the Timurid Period, ed. F. R. Allchin and N. Hammond (London and New York, 1978), pp. 201–14, 245–55
  • D. W. MacDowall: ‘The Hazrajat Hoard of Indo-Greek Silver Drachms’, Pakistan Archaeology, 26/1 (1991), pp. 188–98
  • The Crossroads of Asia: Transformation in Image and Symbol in the Art of Ancient Afghanistan and Pakistan (exh. cat., ed. E. Errington and J. Cribb; Cambridge, Fitzwilliam, 1992)
(b) Jewellery.
  • C. Fabrègues

Northern Afghanistan has produced the richest finds of jewellery from the Achaemenid to the early Kushana period. The wealth of gold material suggests that during this time (c. 5th century bc–1st century ad) the region may have been a centre for goldwork of high artistic quality, but the lack of comprehensive excavation evidence makes the precise distinction between imported and indigenous wares unclear. Ornaments in the Oxus Treasure, discovered near the Amu River, depict male figures, animals or mythical beasts. Prototypes for the stylized figures of fantastic animals occur in Achaemenid art, but the more vigorous appearance of the Oxus Treasure examples links them to the Animal style of the Central Asian steppes (see Ancient Near East, §II, 4, (ii), (a)). Insets for stones on the haunches of the animals take the form of a dot between two curved triangles, a design that probably derives from the dot and comma motif on animal representations at Persepolis (see Achaemenid, §2).

Six tombs of the 1st century bc–1st century ad at Tillya Tepe yielded more than 20,000 gold objects, particularly buckles, clasps, collars, pendants and clothing plaques (see also §II, 1, (iv), (c)). The material shows that a local artistic tradition with roots in Achaemenid art continued until the beginning of the 1st century ad. The jewellery further reveals the co-existence of two other trends, the one influenced by the Animal style of the steppes, the other by Greek culture. Borrowed images are rather debased, as they usually lack the characteristic elegance and vitality of their sources of inspiration. A further distinct group shows a new stylistic development that combines Oriental, Greek and nomadic traditions. Elements borrowed from the decorative repertories of India and China appear on some items, while indigenous traditional motifs occur on others. Small drop-shaped insets of turquoise decorate the objects. Jewellery found in tombs on the opposite bank of the Amu River, in Uzbekistan, are, by contrast, all executed in a homogeneous style that follows well-defined local artistic traditions.

Jewellery of the 2nd–3rd century ad, recovered in excavations of the Kushana levels at Begram (see Ghirshman, pl. XVI.1–11), shows an increasing preference for polychromy and for ornaments encrusted with coloured, geometrically cut stones. Similar jewellery has also been found at Palmyra, Hatra and in south Russia (see Ancient Near East, §II, 4, (ii), (b)). Although few extant examples have been found in Afghanistan, the wealth of ornaments depicted on the bodhisattva statues of Gandhara provide a comprehensive record of the jewellery of the Kushana period (see indian subcontinent, §IV, 5(ii)(d)).


  • O. M. Dalton: The Treasure of the Oxus with Other Examples of Early Oriental Metalwork (London, 1905, rev. 3/1964)
  • R. Ghirshman: Bégram: Recherches archéologiques et historiques sur les kouchans, Mém.: Dél. Archéol. Fr. Afghanistan, 12 (Cairo, 1946), pp. 58–65
  • J. Hackin, J. Carl and J. Meunié: Diverses recherches archéologiques en Afghanistan, 1933–40, Mém.: Dél. Archéol. Fr. Afghanistan, 8 (Paris, 1959), figs 228–55
  • A. M. Mandel’shtam: Kochevniki ha puti v Indiya: Tradui tadzhikskoi arkheologicheskoi ekspeditsii [The nomads on the Indian routes: results of the Tajik archaeological expedition] (Moscow, 1966)
  • A. M. Mandel’shtam: ‘Archäologische Bemerkungen zum Kuschana-Problem’, Beiträge zur alten Geschichte und deren Nachleben: Festschrift für Franz Altheim, 2 vols, ed. R. Stiehl and H. E. Stier (Berlin, 1969–70), pp. 525–34
  • G. Pugachenkova: Les Trésors de Dalverzine-tepe (Leningrad, 1978), 47, pls 63–79
  • B. A. Litvinsky and I. R. Pichikyan: ‘The Temple of the Oxus’, Journal of the Royal Asiatic Society of Great Britain and Ireland (1982), no. 2, 163
  • V. Sarianidi: The Golden Hoard of Bactria from the Tillya-tepe Excavations in Northern Afghanistan (New York and Leningrad, 1985)
(c) Metalwork.
  • Gregory L. Possehl

In September 1966, a hoard of gold and silver vessels was uncovered near Sad Hazard village, between Kunduz and the lapis lazuli mines of Badakhshan, north-east Afghanistan. The hoard apparently came from an archaeological mound known as Fullol or Khosh Tapa (‘happy mound’). The exact circumstances of discovery are not known; it is likely, however, that the hoard came to light while villagers were collecting earth from the site for their fields. Some of the vessels were cut up and sold to local goldsmiths and jewellers. Five gold and twelve silver vessels, almost all fragmentary and weighing c. 2 kg and 1 kg respectively, were placed in the Kabul Museum. Incised designs on the vessels include the stepped cross, so prevalent on Quetta ware and ceramics from Turkmenistan. Incised snakes recall examples from Sialk, Tepe, Iran, and those on Anjira ware from Baluchistan. Two vessels have boars. Also present are the bearded bull, shown turned with a full face, as in finds from the Royal Cemetery of Ur, and rows of romping bulls. One vessel has an eight-armed curl pattern and confronted bulls. None of the bovids depicted is of the humped Indian zebu type. The general conclusion of those who have handled the material is that the objects seem to come from diverse localities and different periods. Dupree, Gouin and Omer even entertain the possibility that the villagers assembled materials from several different find spots. Tosi and Wardak suggest that a local chieftain assembled the materials in antiquity. The bearded bulls suggest a date of c. 2500 bc; other motifs place the date somewhat later, even extending into the early centuries of the 2nd millennium bc.


  • L. Dupree, P. Gouin and N. Omer: ‘The Kosh Tapa Hoard from North Afghanistan’, Afghanistan, 24/1 (1971), pp. 44–54
  • M. Tosi and R. Wardak: ‘The Fullol Hoard: A New Find from Bronze-Age Afghanistan’, East and West, 22/1 (1972), pp. 9–17
  • K. R. Maxwell-Hyslop: ‘The Khosh Tapa–Fullol Hoard’, Afghan Studies, 3–4 (1982), pp. 25–37
Later periods.
  • E. Errington

Excavated material from Greco-Bactrian and early Kushana sites (c. 3rd century bc–2nd century ad) provides most of the information on metalwork in the historical periods before c. ad 900, supplemented by a few stray finds from other periods. The material fits broadly into three categories: imported wares, local imitations (principally of contemporary Near Eastern and Iranian wares) and indigenous items.

The Oxus Treasure, a hoard of primarily Achaemenid gold and silver objects, dated c. 550–c. 330 bc, was found in the banks of the Amu River, perhaps at Takht-i Kubad, in 1877. Achaemenid metal artefacts have also been excavated at Takht-i Sangin, the fortress on the opposite bank of the river in Tajikistan. The earliest objects in the Oxus Treasure are in an assyrianizing style: a gold acinaces sheath decorated with hunting scenes and a gold phiale with addorsed lions. The hoard includes vessels (one in the form of a fish), a model of a chariot, human and animal figurines (deer, a goose, a silver handle in the form of an ibex), a silver disc with horsemen hunting reindeer, ibex and a hare depicted in Animal style, jewellery, coins and numerous votive plaques. The wide diversity of objects suggests the treasure may have been a temple hoard, but details concerning its precise context and original composition are lacking (see §II, 1, (iv), (b)).

Excavated finds from the Greco-Bactrian city of Ai Khanum (c. 330–c. 100 bc) include a bronze cauldron handle with two female busts emerging from vine leaves and a copper crescent (probably a pectoral) with a female face (see Francfort, pp. 56–8, pl. XX.IV, nos 27–8). A bronze statuette of a thickset, wreathed Herakles is iconographically correct, but disproportionate, and was probably made locally (see Bernard, p. 302, fig.). A 3rd-century bc gilded silver medallion, thought to be imported from northern Syria, depicts Cybele and a winged Nike in a chariot pulled by two lions facing a figure on a stepped altar, with Helios, a crescent moon and star above (see Francfort, pp. 93–104, pl. XLI).

More than 20,000 metalwork objects, displaying a rich diversity of styles (e.g. Hellenistic, Greco-Bactrian, Animal style) and variously dated c. 1st century bc or early 1st century ad, were discovered in the royal burials at Tillya Tepe. Most of these finds were pieces of gold jewellery (see also §I, 4 above), but also included were Chinese mirrors, a gold fluted phiale and a cylindrical container, both inscribed in Greek, several bowls and pots, some decorated with bands of Hellenistic vegetal motifs, a gold figure of a goat and, in the male grave, a gold handled dagger and two sheaths, inset with turquoise and decorated with animals, mythical winged beasts and dragons in Animal style.

At Begram, bronze balsamaria or unguent vases in the form of busts of Athena, Hermes and Ares were probably imported from the west c. 1st century bcc. 1st century ad, as were a bronze Harpokrates and an unusual syncretic image combining Serapis with Herakles. The bronze finds (Kabul Mus. and Paris, Mus. Guimet; see Ghirshman, pl. XII; Hackin and Hackin, figs 47–59; Hackin and others, figs 322–5) also include statuettes of a monk, a winged Eros and two horsemen, one Greek, the other of Scythian type, a cockerel with a human head, a mask of Silenus and a shield decorated with a gorgon head encircled by repoussé dolphins. Imported 1st–2nd century ad Alexandrian plaster mouldings of Hellenistic subjects are thought to have served as a source of reference for silversmiths at Begram. Several small bronze figurines of uncertain Afghan provenance (see 1992 exh. cat., nos 102–3, 105–6, 108, 110–12, 117) exhibit a similar range of Classical influences and subjects (e.g. a herm, Herakles, Demeter).

Wardak vase, bronze with Kharoshthi inscription, from Wardak stupa deposit, h. 178 mm, Gandhara period, late 2nd century ad (London, British Museum); photo © The British Museum

View large

A quantity of metalwork was uncovered in the relic deposits of stupas near Kabul and Jalalabad in the 19th century (see §III below). Most famous is the Bimaran gold reliquary (see fig.), which is decorated with standing images of the Buddha, Indra, Brahma and perhaps a bodhisattva (London, BM; see 1992 exh. cat., pp. 186–92; see also §I, 4 above and indian subcontinent, §IV, 5(ii)). These are the earliest datable Buddha images in Gandharan art, for which coins of c. mid-1st century ad included in the stupa deposit provide a terminus ante quem. Other reliquaries important for dating purposes are the gold amulet case from Ahinposh that contained Roman aurei of ad 100–136 and the inscribed bronze Wardak vase, which is dated year 51 of the Kushana king Kanishka and mentions his successor Huvishka (London, BM; see 1992 exh. cat., nos 170–71; see fig.).

Two silver pateras, heirlooms of the Mirs of Badakshan, were acquired by a British officer, Dr P. Lord, in 1838. One is decorated with embossed and gilded sheet silver figures depicting the Triumph of Dionysos (London, BM; see Dalton, no. 196). Certain misunderstood details suggest a non-western, possibly local workshop of c. 1st century ad. The second patera, a Sasanian piece of c. mid-4th century, depicted a prince on horseback attacking a lion but was lost during the British retreat from Kabul in 1840 and is now known only from a drawing (see Harper, p. 134, pl. 11). A number of bronze Buddha and bodhisattva statues, purported to be principally from the eastern Afghanistan Hindu Kush region and variously dated c. 4th–c. 7th century (see von Schroeder, pl. V, figs 3a, 5a–b), are closely related to images from the Swat and Peshawar valleys, Pakistan. While Afghanistan is the most likely source of copper, tin and zinc for all these figures, preliminary technical analysis suggests there are identifiable regional variations in composition and manufacturing techniques (see 1992 exh. cat., pp. 241–56).


  • O. M. Dalton: The Treasure of the Oxus with Other Examples of Early Oriental Metalwork (London, 1905, rev. 3/1964)
  • J. Hackin and J.-R. Hackin: Recherches archéologiques à Begram, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 9 (Paris, 1939)
  • R. Ghirshman: Bégram: Recherches archéologiques et historiques sur les kouchans, Mém.: Dél. Archéol. Fr. Afghanistan, 12 (Cairo, 1946)
  • J. Hackin and others: Nouvelles recherches archéologiques à Begram, 1939–1940, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 11 (Paris, 1954)
  • P. Bernard: ‘Fouilles de Aï Khanoum (Afghanistan): Campagnes de 1972 et 1973’, Académie des inscriptions et belles-lettres: Comptes rendus des séances (1974), pp. 280–308
  • P. Harper: Royal Imagery, i of Silver Vessels of the Sasanian Period (New York, 1981)
  • U. von Schroeder: Indo-Tibetan Bronzes (Hong Kong, 1981)
  • H.-P. Francfort: Le Sanctuaire du temple à niches indentées, II: Les Trouvailles, Mém.: Dél. Archéol. Fr. Afghanistan, 27 (1984), iii of Fouilles d’Aï Khanoum (Paris, 1973–92)
  • V. Sarianidi: The Golden Hoard of Bactria from the Tillya-tepe Excavations in Northern Afghanistan (New York and Leningrad, 1985)
  • G. Fussman: ‘Numismatic and Epigraphic Evidence for the Chronology of Early Gandharan Art’, Investigating Indian Art, ed. M. Yaldiz (Berlin, 1987), pp. 67–88
  • G. A. Pugachenkova and L. I. Rempel: ‘Gold from Tillia-tepe’, Bulletin of the Asia Institute, 5 (1991), pp. 11–25
  • The Crossroads of Asia Transformation in Image and Symbol in the Art of Ancient Afghanistan and Pakistan (exh. cat., ed. E. Errington and J. Cribb; Cambridge, Fitzwilliam, 1992) [technical analysis app. by C. Reedy, pp. 241–63]
(d) Pottery.
  • W. Ball

Prehistoric pottery (c. 4000–1000 bc) is characterized by painted decoration, which develops from simple curvilinear designs towards bolder and more complex geometric motifs and finally includes representations of stylized animals towards the end of the period. There are considerable stylistic links with southern Turkmenistan and Baluchistan throughout the prehistoric period. Burnishing, a distinctive new style of surface treatment, first appeared in the Achaemenid period (c. 530–330 bc) and soon became predominant. The initial simple wavy line or radial burnishes were superseded in the Hellenistic period (c. 3rd–1st century bc) by ‘red spiral-burnished ware’, that is spiral or horizontal line burnishing, usually on red/orange fabrics. Achaemenid and Hellenistic pottery shapes share some characteristics with Middle Eastern pottery of the same date, and Greek influence in the Hellenistic period can be seen with local imitations of Greek black-polished wares. Thereafter, foreign influences seem to recede, and the indigenous spiral-burnished wares predominate.

Red spiral-burnished ware was used almost exclusively in the Kushana period (c. 1st–3rd century ad). The ware occurs all over Afghanistan, much of Pakistan and in adjacent Parthian areas as far west as Khorasan in Iran. Although often considered a ‘hallmark’ for the Kushana period, it cannot be associated exclusively with the Kushanas, as was once thought, but is also found in the subsequent Sasanian, Hepthalite and Turki Shahi periods. Red spiral-burnished ware thus remained the most important pottery style, showing very little variation, throughout the 1st millennium ad until the beginning of the Islamic period (c. early 10th century).


  • J.-C. Gardin: Ceramiques de Bactres (Paris, 1957)
  • L. Dupree: Shamshir Ghar: Historic Cave Site in Kandahar Province, Afghanistan (New York, 1958)
  • J.-M. Casal: Fouilles de Mundigak, 2 vols, Mém.: Dél. Archéol. Fr. Afghanistan, 17 (Paris, 1961)
  • A. W. McNicoll and W. Ball: Excavations at Kandahar (Oxford, 1995)

2. 900–c. 1900.

Afghanistan was of particular importance for the arts of Islam between the 10th century and the 16th. Western Afghanistan was often included in the province of Khurasan (now limited to Iran). Although Nishapur was the provincial capital in early Islamic times, Herat became increasingly important after the Mongol conquests in the 13th century and served as a capital of the Timurid family empire in the 15th. In the 11th–12th century, southern Afghanistan became important under the Ghaznavid dynasty, which had capitals at Ghazna and Lashkari Bazar. Eastern Afghanistan, with its centre at Kabul, flourished from the early 16th century under the patronage of the Mughal emperors of the Indian subcontinent. Northern Afghanistan, particularly the region of Balkh, became important from the 16th century under the Shaybanid dynasty and their Uzbek successors.

The earliest Islamic buildings in Afghanistan date from the 9th century ad, when builders already employed materials, techniques and styles associated with the metropolitan centres of the Abbasid caliphate in Iraq. The nine-bay mosque at Balkh (9th century; see islamic art, §II, 4(i)(c)), for example, was built of brick following a plan also found in North Africa and Spain and was decorated with carved stucco in the Bevelled style. Under the Ghaznavids and Ghurids (reg c. 1000–1215) there developed a progressive style of architecture distinguished by massive scale, occasional use of stone, distinctive arch profiles, new types of glazed tilework, terracotta decoration and inscriptions written in angular and cursive scripts (see islamic art, §II, 5(i)(c)). Perhaps the most evocative example of this style is the towering late 12th-century minaret at Jam (§II, 2) in a remote valley in the centre of the country. Herat flourished under the Timurids, when the city and its environs were graced with splendid brick buildings enveloped in glittering webs of glazed tile (see islamic art, §II, 6(i)(b)). Many of the most important monuments, however, were destroyed in the 19th century when the region was contested by the Russians and the British. The complex of Abu Nasr Parsa at Balkh exemplifies how shrines developed in the later period with the addition of madrasas and hospices around the grave of an earlier saint.

Many types of Islamic decorative art were produced in Afghanistan from early times. A group of hemispherical basins made of high-tin bronze is associated with the Ghaznavids because of the bifurcated hats worn by the courtiers depicted on them. Herat was a major centre of metalworking for centuries, to judge from the inlaid Bobrinski Bucket (Islamic art, fig.) made in 1163, the huge cast bronze basin (dated 1374–5) in the congregational mosque, and the inlaid jugs made for the Timurid ruler Husayn Bayqara and other 15th-century patrons (see Islamic art, fig.). The arts of the book also flourished from an early date. Some idea of early manuscript illumination can be gained from a book (Leiden, Rijksuniv. Bib., MS. 437) formerly in the library of the Ghaznavid amir ‛Abd al-Rashid (reg 1049–51) and a copy of the Koran made at Bust in 1111–12 (Paris, Bib. N., MS. arab. 6041; Islamic art, fig.). The apogee of the arts of the book in Afghanistan, and one of the great moments in all Islamic art, occurred in Herat under the patronage of the Timurid princes Baysunghur and Husayn Bayqara (see islamic art, §III, 4(v)(d)). Splendid calligraphy by such masters as Mir ‛Ali Husayni Haravi was embellished with beautiful illumination and paintings by such artists as Bihzad and contained within sumptuous bindings (§II, 2). After the collapse of Timurid power in the early 16th century, Herat ceased to be a major centre of patronage, but some painters worked in Kabul, temporary home of the Mughal court, in the middle of the century.

For bibliography see the individual articles cited in the text.

3. After c. 1900.

Creative expression in 20th-century Afghanistan resides in the artistry of diverse ethnic and tribal groups living mainly in rural villages and semi-nomadic camps. Western art styles, popularized during the early years of the 20th century by educated urban élites, tend to be imitative rather than innovative.

(i) Architecture.

The introduction of a more Western style of domestic architecture was accompanied by innovations in interior décor, furniture making, painting, landscape gardening and dress styles. When Amir Abdur Rahman (reg 1880–1901) acceded to the throne after living in exile in Central Asia for over a decade, he abandoned the traditional house plan with its interior courtyards and personally designed vaulted and domed palaces that faced outwards on to English gardens adorned with fountains.

The first ‘European’ home was built according to the same specifications as Dorchester House of Park Lane, London, in Kabul. This building heralded a period characterized by British Indian designs. These verandahed colonial styles fell out of fashion after the short 1919 war between Afghanistan and England, and King Amanullah (reg 1919–29) turned for inspiration to 18th-century European grand styles with their exuberant and eclectic mix of Neo-classical and pseudo-Rococo elements. By the mid-20th century this ebullience gave way to the utilitarian Soviet and Central European models that still dominate both domestic and public buildings.

Interior transformations mirrored changing lifestyles. In traditional homes each room served several purposes. White walls were decorated with floral ornamentation in pressed, moulded or carved stucco. Furnishings comprised richly coloured Afghan carpets and embroidered door hangings. There was no need for furniture, as mattresses and bolsters doubled for sitting and sleeping, fabric runners were spread on the floor for dining, and traditional clothing folded easily for storage in wall niches or decorated boxes.

Rooms in modern homes in Kabul, on the other hand, were set aside for specific purposes and filled accordingly with massive, ornately carved furniture, including commodious wardrobes to accommodate European clothes. While carpets were retained, stuccowork gave way to flocked and textured wallpaper, or stencilled approximations and daubed simulations of luxurious wood and marble wainscoting that provided a backdrop for a wealth of imported Victorian clutter.

(ii) Painting and sculpture.

From the late 19th century onwards artists experimented with novel Western techniques, yet their landscapes and vivid abstract paintings incorporated no recognizable Afghan characteristics; even when the scenes were Afghan, the styles were clearly derivative. Sculpture, an innovation introduced many years later by students returning from Italy and the Soviet Union, was coolly received by this Muslim society. However, birds and animals carved from marble and lapis lazuli by artisans trained in Kabul by Chinese masters became popular with tourists.

Contemporary Afghan artists and sculptors have yet to enjoy either private patronage or public support despite official promotion by the government since the 1970s. In 1978 leftist urban leaders overthrew the élites who had set trends for almost 100 years. Their rise, closely followed by Soviet military intervention, ushered in a period dominated by Socialist Realism. Devastation caused by ground and air offensives forced more than a third of the Afghan population into exile in neighbouring Pakistan and Iran. Nevertheless, since 1989 an intrepid group of young artists has attracted growing numbers of students in their attempts to revive the Herati traditions of miniature painting and calligraphy renowned in Afghanistan during the 15th century.

Within the public domain, stylized floral mural painting enjoys a certain popularity in the decoration of mosques and teahouses. The most distinctive painting tradition in this genre, however, is displayed on truck bodies completely embellished with a wide range of themes including Swiss chalets, lovely ladies, trains, boats, telephones, animal combat scenes, birds and contemporary battle scenes.

Nevertheless, conservative elements in Afghan society remained suspicious of art. Islamists clerics cracked down on ‘un-Islamic’ traditions, banning all forms of imagery, music and sports, including television, in accordance with what they considered a strict interpretation of Islamic law. In March 2001 they blew up the Buddhas of Bamiyan, two monumental statues (h. 55 and 37 m) of the standing Buddha, carved into the sandstone cliffs of the valley in central Afghanistan some time between the 5th and the 9th century. The act spurred an international outcry, and various attempts have been made to correct what all now agree was a horrendous mistake. The Afghan government has commissioned Japanese artist Hiro Yamagata (b 1948) to recreate the Bamiyan Buddhas using 14 solar- and wind-powered laser systems to project the images of the statues onto the cliff where they once stood. Pending the decision whether to rebuild the statues on the site, a $1.3 million UNESCO-funded project is sorting out the chunks of clay and plaster, ranging from boulders weighing several tons to fragments the size of tennis balls, and sheltering them from the elements.

(iii) Other arts.

Of all Western innovations, dress probably had the most far-reaching, durable influence upon society. In the late 19th century hoops, bustles and wide-brimmed plumed and beribboned hats were introduced for the ladies, along with splendid emblazoned uniforms, frock coats and tweeds fancied by the gentlemen at court. As fashions closely followed European changes, including mini-skirts and the ubiquitous T-shirt and jeans, Western dress became a measure of modernity throughout the educated urban populations and, for women, it symbolized emancipation.

More enduring indigenous examples of art and craftsmanship are found among the diverse creative traditions brought to Afghanistan over many centuries by artisans travelling to this pivotal Central Asian land from east, west, north and south along the routes of conquest and commerce. From 1978 onwards, however, the disruptions of war hastened the decline of crafts already affected by the introduction of modern materials, production methods, imports and commercialization. The art of ornamental stucco has all but disappeared; ikat weaving from an already limited number of northern workshops suffers equally; Nuristani wood-carving, Pushtun painted and lacquered wood decoration, tilemaking, copperwork and pottery and Herati glassblowing, silversmithing and silk weaving are all threatened. Nevertheless, since few items produced in Afghan villages and semi-nomadic camps are purely decorative and since most express personal and/or group identification and status, there is reason to hope for a craft revival once peace permits life to return to normal. The exquisite embroidery made by refugees for their personal use, in contrast to the lamentable pieces seen for sale, allows this note of optimism.

(a) Carpets and textiles.

The richness of form and colour of the flat-woven, hand-knotted and felt carpets made by the Turkomen, Uzbek, Hazara, Aimaq, Kirgiz and Baluch place them among Afghanistan’s most renowned artistic products. Ranking fifth among the country’s exports before 1978, the carpet trade has continued on a reduced scale throughout the war, although the difficulties in obtaining quality raw materials coupled with local market demands on design and pricing have adversely affected the production of Afghan carpets by refugees in Pakistan and Iran. Carpet production contributes significantly to family income and is highly valued. Quality products are more particularly esteemed because they add to individual status. A bride gains heightened respect from her husband’s family when her dowry includes fine examples of her own handiwork; a man’s wealth and status is gauged by the quantity and quality of his household’s production. Furthermore, since distinctive structures, designs, symbols and colours are proudly associated with specific groups, the excellence of the work of its individual producers enhances the reputation of the entire community.

While carpets represent a major portion of woven articles for sale, an inexhaustible variety of other items are made for both utilitarian and decorative use within the household. Long narrow woven bands both strengthen and decorate the wooden lattice framework of round felt-roofed yurts (see Tent, §II, 2). The Kirgiz, among others, entwine the reeds forming the skirting of yurts and interior partition screens with yarn in variegated designs. Almost all semi-nomadic groups use hand-knotted, often fringed, door hangings.

Furniture in most sedentary and semi-nomadic homes scarcely extends beyond an occasional wooden stool, one or two wooden chests and perhaps a cradle (see §II, 3, (iii), (d) below). Possessions from clothing to food supplies are stored in flat-woven or knotted bags of various shapes and sizes. Uzbeks distinctively wrap bedding and clothing in flat-woven squares (2×2 m), which are stacked on top of chests. Finely embroidered V-shaped pieces and beaded tassels hung against these bundles provide the final adornment. Elsewhere, strings of pompoms and woollen tassels are used as cornices and wall hangings. By the mid-20th century, hand-woven textiles for clothing had largely given way to imported and locally manufactured materials. Exceptions are prized silk turbans, a speciality of the Herat area, and the popular striped cottons used for long-sleeved robes throughout the north.

Embroidery motifs and stitches serve to distinguish ethnic and regional goups. Embroidered items are made for family use and only rarely offered for sale. This handicraft is most importantly associated with marriage. Each male family member attending a wedding is presented with a finger-woven trouser drawstring with silk tassels and an intricately embroidered cummerbund. In addition to embroidered clothing, the bride’s dowry typically includes up to 20 types of embroidered household items, from spoon bags, tray covers, sachets for money and make-up to dust covers for Korans and radios. Soft, knee-high leather boots embellished with fine embroidery by Turkic-speaking women of the north are especially prized. Every woman in the family takes part in spinning wool and silk, weaving, stitching and embroidering in order to amass this extensive collection. Kin-related girls often work together on the large embroidered and patchwork pieces. Mothers devise baby bonnets festooned with feathers, pompoms and protective charms. Intricately embroidered and beaded hats worn under men’s turbans and women’s headscarves are distinctive symbols of group identity. Ornaments, mirrorwork, gold braid, elaborate beading and fine embroidery decorate the high-waisted bodices, elbow-length cuffs and deep hems of women’s dresses, the skirts of which may, among some groups, contain as much as 12 m of velvet or flowered cotton.

Among semi-nomadic groups, individual artistry is publicly displayed during their annual migrations in a variety of ornamental trappings for camels, donkeys and horses. In addition to saddle bags and blankets there are decorated leather harnesses, silver-studded saddles and neckpieces. A bridal camel, bedecked with a heavy, glass-beaded headdress and reins takes pride of place in any caravan. These sumptuous accoutrements proclaim wealth, status and power.

(b) Jewellery.

Items for personal adornment have been zealously developed by all groups. Distinctive patterns distinguish each ethnic group, place of origin and, particularly among the Pushtun, tribe and sub-tribe. The most popular jewellery is made of silver, at times fire-gilded, a technique most employed by the Turkomen. Heavy, embossed torques of twisted silver are a speciality of Nuristan, but all bracelets, armlets, earrings, temple pendants and headdress ornaments tend to be massive and liberally hung with pendants. Insets of coloured glass or hardstones, including cornelian, turquoise and lapis lazuli, are frequently embued with symbolic meaning to avert sorrow, danger and disease or bring joy, serenity and marital bliss. Generous sprinklings of silver beads, discs, coins, medallions, dress fastenings, amulets and talismans are also sewn on to clothing in great profusion.

(c) Pottery.

The shapes and designs of the primarily utilitarian pottery have survived unchanged for 5000 years. Glazes are rare, except for those found at Istalif, a hillside village just north of Kabul. The clear, bright blue pottery with black incised floral decorations and the bird and animal figurines from Istalif are unique. The Istalif double-headed horse must surely represent a tradition of considerable antiquity, although the potters themselves attribute their creations simply to ‘custom’.

(d) Woodwork.
  • N. Hatch Dupree, Sheila S. Blair and Jonathan M. Bloom

Lacquered wooden boxes, stools, bed-legs and cradles are regional specialities, most notably in the east. Lacquerwork is normally restricted to colourful banding, but some artisans have developed a technique of applying several layers of different colours, which are then cut away to reveal intricate floral designs. The art of wood-carving in general is largely applied to such architectural elements as window-frames and panels, doorframes, lintels and pillars in homes and mosques. Much of this work exhibits affinities to Kashmiri floral, geometric and curvilinear traditions; unique motifs from Nuristan include animistic symbols that pre-date the conversion of this area to Islam in 1895. Distinctive snuff-boxes, made from small gourds, are traditionally shaped in wooden moulds as they ripen on the vine and are then highly polished, painted or adorned with silver stoppers and decorative collars.


  • M. Kohzad: ‘L’Inauguration du salon d’automne à Kaboul’, Afghanistan Quarterly, 1/4 (1946), pp. 30–34
  • B. Dupaigne: ‘Aperçus sur quelques techniques afghans’, Objets et mondes, 8 (1968), pp. 41–84
  • S. P. Seherr-Thoss and H. C. Seherr-Thoss: Design and Color in Islamic Architecture: Afghanistan, Iran, Turkey (Washington, 1968)
  • A. Friedman: ‘The Handicrafts of Afghanistan’, Afghanistan Quarterly, 25/2 (1972), pp. 11–12
  • N. Dupree: ‘Archaeology and the Arts in the Creation of a National Consciousness’, Afghanistan in the 1970s, ed. L. Dupree (New York, 1974), pp. 203–38
  • C. Naumann: ‘Pamir und Wakhan’, Afghanistan Journal, 1/4 (1974), pp. 91–104
  • G. O’Bannon: The Turkoman Carpet (London, 1974)
  • W. Bechhoefer and T. B. Katz: Serai Lahori: Traditional Housing in the Old City of Kabul (College Park, MD, 1975)
  • P. Centlivres: ‘Les Uzbeks du Qattagnan’, Afghanistan Journal, 2/1 (1975), pp. 28–36
  • M. Centlivres-Demont: ‘Les Peintures sur camions en Afghanistan’, Afghanistan Journal, 2/2 (1975), pp. 60–64
  • J.-C. Blanc: Afghan Trucks (London, 1976)
  • M. Centlivres-Demont: Popular Art in Afghanistan: Paintings on Trucks, Mosques and Tea-houses (Graz, 1976)
  • L. Dupree: Afghan Women (Hannover, 1976/R 1994 VHS) [film; Afghanistan Ser., iv]
  • N. Dupree: ‘Early Twentieth-century Afghan Adaptations of European Architecture’, Art and Archaeology Research Papers [AARP], 12 (1977), pp. 15–21
  • R. Dor and C. Naumann: Die Kirghisen des afghanischen Pamir (Graz, 1978)
  • A. Janata: ‘Ikat in Afghanistan’, Afghanistan Journal, 5/4 (1978), pp. 130–39
  • A. Stucki: ‘Horses and Women’, Afghanistan Journal, 5/4 (1978), pp. 140–49
  • N. Dupree: ‘A Building Boom in the Hindukush [Boom edilizio nell’Hindukush]’, Lotus International [prev. pubd as Lotus], 26 (1980), pp. 115–21
  • M. Klimburg: ‘A Collection of Kafir Art from Nuristan’, Tribus: Zeitschrift für Ethnologie und ihre Nachbarwissenschaften vom Linden-Museum Stuttgart, 30 (1981), pp. 155–202
  • J. Kalter: ‘Die Sammlungen des Linden-Museums aus Afghanistan und der Nachbargebieten’, Afghanistan Journal, 9/3 (1982), pp. 76–85
  • I. Rittmeyer: ‘Die Sammlung Rittmeyer’, Afghanistan Journal, 9/4 (1982), pp. 112–14
  • R. Parsons: The Carpets of Afghanistan (Woodbridge, 1983)
  • N. Dupree: ‘National Museum of Afghanistan’, Art Museums of the World, 1 (Westport, 1987), pp. 26–30
  • M. Klimburg: ‘Notes on the Architecture of Nuristan’, Archv Vlkerknd, 41 (1987), pp. 41–52
  • N. Dupree: ‘Victoriana Comes to the Haremserai in Afghanistan (Viktorianischer Stil erobert den Haremserail)’, Bauen und Wohnen am Hindukush, Stiftung Bibliotheca Afghanica, 7, ed. P. Bucherer-Dietschi (Liestal, 1988), pp. 111–49
  • The Decorative Arts of Central Asia (exh. cat., ed. J. Graham and H. Sandys; London, Zamara Gal., 1988)
  • J. Frembgen: Naswar: Der Gebrauch von Mundtabak in Afghanistan und Pakistan, Stiftung Bibliotheca Afghanica, 8 (Liestal, 1989)
  • A. Szabo and T. Barfield: Afghanistan: An Atlas of Indigenous Domestic Architecture (Austin, 1991)
  • B. Dupaigne and F. Cousins: Afghan Embroidery (Lahore, 1993)
  • O. Tirard–Collet: ‘After the War: The Condition of Historical Buildings and Monuments in Herat, Afghanistan’, Iran, 36 (1998), pp. 123–38
  • J. W. Frembergen: Lebensbaum und Kalaschnikow: Krieg und Frieden im Spiegel afghanischer Bildteppiche [Tree of Life and Kalashnikov: War and Peace Reflected in Afghan Pictorial Carpets] (Blieskastel, 2000)
  • B. O'Kane: ‘The Uzbek Architecture of Afghanistan’, La Mémoire et ses supports en Asie centrale, Cahiers d’Asie Centrale, viii, ed. V. Fourniau and C. Poujol (Tashkent and Aix-en-Provence, 2000), pp. 123–60
  • M. Klimburg: ‘The Arts and Culture of Parun, Kafiristan's “Sacred Valley”’, Arts asiatiques [prev. pubd as Rev. A. Asiat.], 57 (2002), pp. 51–68
  • H. Bechna: ‘Fine Arts in Afghanistan: from Ancient Times until the 20th Century’, Art & Thought: Fikrun wa Fann, 78 (2003), pp. 20–26
  • W. Floor: ‘Iran and Afghanistan’, Towards the Contemporary Period, from the Mid-nineteenth to the End of the Twentieth Century, ed. M. K. Palat and A. Tabyshalieva (2005), vi of UNESCO History of Civilizations of Central Asia (Paris, 1992–), pp. 757–93
  • C. H. Bleaney and M. Á. Gallego: ‘Arts: Traditional Crafts’, Afghanistan: A Bibliography, Handbook of Oriental Studies: Central Asia, 13 (Leiden, 2006), pp. 102–10
  • C. H. Bleaney and M. Á. Gallego: ‘Islamic Architecture in Afghanistan’, Afghanistan: A Bibliography, Handbook of Oriental Studies: Central Asia, 13 (Leiden, 2006), pp. 116–23
  • S. Paine: Embroidery from Afghanistan (London, 2006)
  • Turquoise Mountain Foundation: [organization dedicated to reviving traditional arts] (accessed 5 August 2009)

III. Historiography.

  • D. W. MacDowall

In the first centuries ad under Kushana rule, Afghanistan became established as a major centre of Buddhism, with many monasteries and stupas. Between the 5th and 8th centuries, a number of these Buddhist sites were visited and described by Chinese pilgrims such as Faxian (c. 400) and Xuanzang (c. 630).

Western interest in Afghanistan stems from the discovery in the 18th century by numismatists such as Theophilus Bayer, J. Pellerin and M. Mionnet that some fine portrait coins could be attributed to several Greek kings of Bactria (see Bactrian and Indo-Greek monarchies) who were mentioned in Classical texts (see Wilson, pp. 3–4). At the beginning of the 19th century, the British in India became increasingly interested in Afghanistan, initially in order to forestall any potential designs of Napoleon I, and later for fear of Russian expansion in Central Asia. Various travellers who visited the country and recorded the monuments they saw were William Moorcroft and George Trebeck, on an ill-fated journey in 1825; Alexander Burnes and J. G. Gerard, who travelled to Bukhara in 1831; and Martin Honigberger, a Transylvanian doctor, formerly in the service of the Sikh emperor Ranjit Singh, who explored several stupas in the neighbourhood of Kabul and Jalalabad in 1833. A major contributor was Charles Masson, a deserter from the East India Company army, who from 1826 onwards travelled extensively in the regions to the north of British India. He spent six years in Afghanistan (1832–8), during which time he surveyed and explored the archaeological remains near Kabul, Jalalabad and Hadda. The East India Company granted him a pardon in 1834 and funding to continue his investigations. He amassed a collection of more than 80,000 coins, primarily from Begram, while his excavations of numerous stupas produced such remarkable objects as the Bimaran casket. His principal discoveries were published in Ariana Antiqua (see Wilson, pp. 55–118). He was also an early contributor to the journal (J. Asiat. Soc. Bengal) published from 1832 onwards by the Asiatic Society of Bengal (founded 1784), which under the editorship of James Prinsep became a principal source of information on the antiquities of India and neighbouring regions. Work on coins and their inscriptions, particularly by Prinsep and the German scholar Christian Lassen, resulted in the decipherment of the Kharoshthi script. But after the first Anglo–Afghan War (1839–42), little further fieldwork or research was undertaken, other than William Simpson’s excavation of the Ahinposh Stupa near Jalalabad at the beginning of the second Anglo–Afghan War (1878–9) and incidental reports by members of the British border commissions in 1896–7 and 1903–5.

In the early 20th century the leading art historian in the field was undoubtedly Alfred Charles Auguste Foucher, who, fascinated by the extension of Hellenism in the east, contributed the monumental study, L’Art gréco-bouddhique du Gandhâra. The Franco–Afghan cultural convention of 1922 gave the French a virtual monopoly of archaeological research in Afghanistan for 30 years and a permanent institute in Kabul, with Foucher as first director. The Délégation Archéologique Française en Afghanistan has undertaken a major series of excavations and studies of the art and archaeology of Afghanistan. The results, published in more than 30 volumes (Mém.: Dél. Archéol. Fr. Afghanistan), include comprehensive excavation reports on Hadda, Begram, Surkh Kotal and Ai Khanum. From the 1960s until 1978 Afghan archaeologists conducted extensive excavations at Tepe Shotor and Tepe Kalan, two important monastic complexes at Hadda. Since the 1960s other foreign missions have been allowed to work in Afghanistan: the Italian Istituto per il Medio ed Estremo Oriente at Ghazna (see Sardar, Tepe), German and American teams in Sistan, a Japanese team from Kyoto University at Tepe Skandar, a Russian mission at Tillya Tepe and the British Institute at Kandahar. This has led to much wider international interest in the art history of Afghanistan.


  • A. Burnes: ‘On the Colossal Idols of Bamian’, Journal of the Asiatic Society of Bengal [see J. Asiat. Soc.], 2 (1833), pp. 561–4, pl. XIX
  • A. Burnes: Travels into Bokhara, 3 vols (London, 1834/R Karachi, 1973)
  • J. G. Gerard: ‘Memoir on the Topes and Antiquities of Afghanistan: From Jelalābād, 4th December 1833’, Journal of the Asiatic Society of Bengal [see J. Asiat. Soc.], 3 (1834), pp. 321–9
  • E. Jacquet: ‘Sur les découvertes archéologiques faites par M. Honigberger dans l’Afghanistan’, Journal asiatique, n. s. 2, 2 (1836), pp. 234–77; iv (1837), pp. 401–40; v (1838), pp. 163–97; vii (1839), pp. 385–404
  • W. Moorcroft and G. Trebeck: Travels in the Himalayan Provinces of Hindustan and the Panjab from 1819 to 1825, 2 vols, ed. H. H. Wilson (London, 1841); intro. G. J. Alder (R Karachi, 1979)
  • H. H. Wilson: Ariana Antiqua: A Descriptive List of the Antiquities and Coins of Afghanistan (London, 1841/R Delhi, 1971)
  • W. Simpson: ‘Buddhist Architecture of the Jelalabad Valley’, Transactions of the Royal Institute of British Architects (1879–80), pp. 37–58
  • A. Foucher: L’Art gréco-bouddhique du Gandhâra, 2 vols (Paris, 1905–18)
  • M. Taddei and G. Verardi: ‘The Italian Archaeological Mission in Afghanistan: Brief Account of Excavation and Study, 1976–1981’, Studi di storia dell’arte in memoria di Mario Rotili (Naples, 1984), pp. 41–70
  • G. Whitteridge: Charles Masson of Afghanistan (Warmington, 1986)

IV. Museums and collections.

  • F. Tissot

The collections of material from Afghanistan are few in number but well published. The majority of pieces are located in the Kabul Museum and the Musée Guimet, Paris, including finds by the Délégation Archéologique Française en Afghanistan (DAFA), the sole organization with the right to survey and excavate in Afghanistan from 1922 to 1964. The earliest collections date from the 19th century, when, notwithstanding political unrest and wars, a number of primarily British travellers and pioneers visited Afghanistan and sent a few stray finds, principally coins, to the British Museum, London. A major collection of coins and stupa relic deposits, made by Charles Masson in the 1830s (see §III above), was initially deposited in the East India Company Museum, London, then in 1880 divided between the British Museum and the Indian Museum, Calcutta.

After World War I cultural co-operation between the khan (later king) Aman-Allah (reg 1919–28) and France resulted in 1922 in the creation of the Délégation Archéologique Française en Afghanistan. The Délégation received exclusive rights to survey and excavate in Afghanistan for a period of 30 years. Following initial surveys by Alfred Charles Auguste Foucher at Balkh and Hadda (1922–3), finds from these investigations were shared between the Kabul Museum and the Musée Guimet, under the respective control of the king and the director of the Délégation. Even during World War II the scheme continued to benefit both countries. The majority of discoveries thus preserved were of outstanding importance, such as the well-known series of schist reliefs from Paitava (see fig. above) and Shotorak, the stucco reliefs from Hadda, fragments of wall paintings, reliefs and clay statues from Bamiyan, Kakrak and Fondukistan (see figs 12 and 9 above), the famous hoard from Begram and finds from the prehistoric city of Mundigak (see figs 2 and 5 above) and the large dynastic temple of Surkh Kotal.

The French finally lost their exclusive rights to excavation in 1962, and from 1964 onwards no archaeological finds were legally allowed out of Afghanistan. Foreign teams worked under the control of the newly created Archaeological Survey of Afghanistan. The finds from British, American, German and Japanese excavations of prehistoric and historic sites, the French at Ai Khanum, the Italians at Ghazna (see Sardar, Tepe) and Afghan and Soviet teams at Emshi Tepe and Tillya Tepe were all placed in the Kabul Museum. In addition, site museums were created at Bamiyan, following restoration of the site (1974–8), and at the Buddhist monastery of Tepe Shotor, Hadda (destroyed during bombing in 1979).

While Joseph Hackin was Director of the Musée Guimet in the 1930s, groups of 20 selected sculptures from the Hadda collection were distributed on permanent loan to various museums worldwide, in order that the material might become better known (Brussels, Musées Royaux A. & Hist.; Kansas City, MO, Nelson–Atkins Mus. A.; London, BM; Luxembourg, Mus. N. Hist. & A.; New Haven, CT, Yale U. A.G.; St Petersburg, Hermitage; Stockholm, Östasiat. Mus.; Tokyo, Ueno Royal Mus.; and a few others). In addition, about 12 ivory pieces from Begram were sent to India in 1960–63 as an exchange loan for some rare pieces of ancient Indian art. These ivories, the only examples outside Kabul or Paris, are in the National Museum, New Delhi. In the 1990s a programme was launched to expand the exhibit of Afghan material in the Musée Guimet beyond the original two galleries, in order to display additional, newly restored pieces from the reserve collection. Many other museums worldwide have Gandhara stucco pieces of uncertain provenance that are stylistically attributable to Afghanistan or Pakistan. It is to be hoped that the site origin of many of these pieces may be determined through the use of highly sophisticated methods of technical analysis that are increasingly available.

The Kabul Museum ranked among the most opulent depositories in the world, with a collection that recorded 50,000 years of the cultural history of Afghanistan. Although the artefacts were all boxed in 1991 for safe-keeping during the civil war, the museum building was extensively damaged during bombing in 1993. Soon afterwards artefacts from the museum began to appear on the international art market, but the museum staff had apparently hidden some of the museum’s treasures and 22,000 objects, including the Begram ivories and coins, have been recovered.


  • O. Monod-Brühl: Guide to the Musée Guimet (Paris, 1966)
  • Ancient Art from Afghanistan: Treasures of the Kabul Museum (exh. cat. by B. Rowland; New York, Asia Soc. Gals; Los Angeles, CA, Co. Mus. A.; Washington, DC, N. Col. F.A.; 1966/R 1976)
  • J. Auboyer: L’Afghanistan et son art (Prague, 1968)
  • N. Hatch Dupree and others: The National Museum of Afghanistan: An Illustrated Guide (Kabul, 1974)
  • F. R. Allchin and N. Hammond, eds: The Archaeology of Afghanistan: From Earliest Times to the Timurid Period (London and New York, 1978)
  • W. Ball and J. C. Gardin: Archaeological Gazetteer of Afghanistan/Catalogue des sites archéologiques d’Afghanistan, 2 vols (Paris, 1982)
  • J. Kalter: ‘Die Sammlungen des Linden-Museums aus Afghanistan und der Nachbargebieten’, Afghanistan Journal, 9/3 (1982), pp. 76–85
  • N. Dupree: ‘National Museum of Afghanistan’, Art Museums of the World, i (Westport, 1987), 26–30
  • Afghanistan: Une histoire millénaire (exh. cat., ed. M. C. M.-C. Bianchini; Barcelona, Cent. Cult. Fund. Caixa Pensions; Paris, Mus. Guimet; 2001–2)
  • F. B. Flood: ‘Between Cult and Culture: Bamiyan, Islamic Iconoclasm, and the Museum’, Art Bulletin, 84/4 (2002), pp. 641–59
  • C. H. Bleaney and M. Á. Gallego: ‘Museums’, Afghanistan: A Bibliography, Handbook of Oriental Studies: Central Asia, 13 (Leiden, 2006), pp. 126–7
  • F. Tissot: Catalogue of the National Museum of Afghanistan, 1931–1985, Arts, Museums and Monuments Series (Paris, 2006)
  • Afghanistan, les trésors retrouvés: Collections du Musée national de Kaboul (exh. cat. by P. Cambon and J. F. Jarrige; Paris, Mus. Guimet, 2006–7)