Treasures of the Great Silk Road

by Edgar Knobloch

  Professor Schlumberger preferred to call this distinctive culture Graeco-Iranian.

  Samangan, on the road between Tashkurghan and Pul-i Khumri, is a modern town, but some miles from it is the site of Takht-i Rustam or Haibak, formerly a Buddhist monastery consisting of a number of caves and a stupa dating from the fourth to the fifth century AD. It was visited by several travellers in the nineteenth century, but was first systematically explored by Foucher in 1923, and then by Mizuno in 1959 and 1960.12

  The monastery stands at the bottom of the hill and has five caves. In the first of these is an anteroom and a large circular room with a huge lotus in full bloom carved in the ceiling. This is a unique form of decoration. The only comparable one, but much smaller, is in one of the monasteries in Taxila. Two niches, one above the other, are in the back wall, but there is no trace of statues. The second cave is a long double corridor with a vaulted ceiling, with two entrances and individual cells for the monks. The third cave, similar to the first, also has an anteroom and a square room with an unadorned ceiling. There is one large niche for a statue, and squinch arches in the corners. The fourth cave, which contains four rooms, was utilitarian. There were benches along three walls in the middle room and a square water tank in the centre. It was most probably used as a bathhouse, especially as the fifth cave, next to it, has numerous features indicating that it was a lavatory. On top of the hill, across from this complex, is a stupa that again has a unique feature. (See 59.) It looks as if it were sunk into a large pit. In fact it stands in an open-topped cave hewn out of solid rock. The top of the stupa shows out of the cave. On it is a harmika, a kind of square balcony, also hewn out of the rock. (The harmika was originally a support for the pole holding the chatra, or umbrella.) Inside the harmika is a round and domed room that housed the reliquary. It is possible to enter the cave at the base of the stupa through a tunnel, and to walk around it. The passage around the stupa is approximately 6ft wide. The stupa, of polished limestone, is 92ft across and 26ft high.

  The last great archaeological discovery before the curtain dropped on Afghanistan was no doubt V. Sarianidi’s excavations at Tilla-Tepe, not far from the village of Shibargan, in the years 1977–78. The aptly named Tilla-Tepe (Gold Mound) lies near the site of Yemshi-Tepe, a Bactrian city of the Kushan period, where remains of a temple, dating probably from the late second millennium BC, were first explored. The six tombs excavated at Tilla-Tepe dating from the first century BC to the first century AD, yielded almost 20,000 gold objects, ornaments, jewels etc, as well as important information about the population of that period. Certain aspects of the funerary rites point to Scythian origins: women’s jewels prove that women enjoyed a priviledged position in the society, a phenomenon typical of the nomads. The artistic quality of the jewels is inferior to the Hellenistic ones of Greek Bactria, and may have been influenced by Parthia. On the other hand, the treatment of animal motifs reflects the Scytho-Sarmatian traditions and an affinity with the Siberian ‘animal style’.

  There are artefacts pointing to Graeco-Bactrian origin, others may be classified as Graeco-Roman, some are of local Bactrian style, some show a mixture of Graeco-Roman and Siberian Altaic influences; the animal style of Scytho-Sarmatian origin can be found in some. The sixth group is that of objects showing local Eastern Persian or Old Bactrian traditions pointing back to the Bronze Age. Here, according to Sarianidi:

  …nomad art of the steppes made an important contribution by virtue of its dynamic though conventional lines and its naivete of expression… It may be suggested that nomad art served to catalyze the interaction of the two old – Greek and Bactrian – art trends… 13

  The chronological gap between the Oxus Treasure and Tilla-Tepe has now been filled by the newly found temple treasure at Takht-i Sanghi (Southern Tajikistan), (cf. Litvinski, B. and Pichikian, I., ‘Archaeological discoveries in South Tajikistan’, Bulletin of the USSR Academy of Sciences, 1980). (See p. 149.)

  In 1966 the hoard of Tepe Fullol was found in the province of Badakhshan. It was dated to the Bronze Age, and consisted of five gold and twelve silver vessels from different periods and locations. The motifs and the techniques used were a mixture of Indian, Iranian, Central Asian and Mesopotamian elements. On this evidence it seems probable that most elements could be dated to c.2500 BC.

  Another two important hoards of coins were found in 1948 and 1992 respectively, at the site of Mir Zakah in the Gardez region.

  According to an unconfirmed report, a Russian bomb has accidentally uncovered another treasure hoard, in the province of Pakhtia. In 1997, diggers found large quantities of jewels, gold and silver coins as well as several statues there.

  * * *

  NOTES ON CHAPTER XVIII

  Full details of abbreviations and publications are in the Bibliography

  1 Gibb, Ibn Battuta, p.175.

  2 Le Strange, Clavijo, p.198.

  3 Byron, The Road to Oxiana, p.256.

  4 Melikian-Shirvani, ‘Baba Hatem, un chef-d’oevre d’époque ghaznevide en Afghanistan’, in VIth International Congress on Iranian Art and Archaeology, Tehran, 1972.

  5 Pugachenkova, G.A., No Gumbad in Balkh.

  6 Duprée, L., Afghanistan, p.289.

  7 Duprée, L., Afghanistan, p.291.

  8 Lecuyot, G., Ishizawa, O., La restitution virtuelle d’Aï Khanoum, Archéologie, 420/05.

  9 Duprée, L., Afghanistan, p.292.

  10 Duprée, L., Afghanistan, p.292.

  11 Duprée, N.H., The Road to Balkh, pp.18–19.

  12 Foucher, A., ‘Notes sur les antiquites bouddhiques de Haibak’, JA, 1924; Mizuno, S., Haibak and Kashmir-Smast, 1962.

  13 Sarianidi, V., The Golden Hoard of Bactria, p.55.

  XIX

  THE SOUTH

  The period of glory of Ghazna (Ghazni is a modern spelling) was comparatively short, spanning less than two centuries. In the middle of the tenth century it was a petty town in eastern Afghanistan, theoretically belonging to the Samanids, but in practice control from distant Bukhara was tenuous. Little is known of its history in earlier times, except that after the demise of the Hephthalites (White Huns) some time in the seventh century, it was ruled, together with Kabul, by the Turk-Shahis, a dynasty that introduced a strong Indian influence. The Turk-Shahis were subsequently replaced in the middle of the ninth century by the Hindushahi rajas, a dynasty of Indian origin, and most probably of Brahmin faith. The Arab raids from Sistan, which became frequent from the end of the seventh century onwards, were aimed more at exacting tribute, plunder and slaves than at establishing a permanent military occupation, and in general the local rulers were able to preseve their authority. Connections with India were numerous and there is much evidence of Indian religious and cultural influence until as late as the tenth century.

  Samanid control of Ghazna was established when, in 961, their commander-in-chief in Khorassan, Alptigin (Tigin – also spelled tegin or tagin – is a Turkish military title) wrested the town from the last Hindu Shahi ruler and was himself appointed governor. By then the power of the Samanids was already in decline and the administration and defence of their outlying provinces was entrusted, to an ever-increasing extent, to Turkish mercenaries and slaves whom the rulers believed to be more reliable than the local landed aristocracy. In this they followed the example of the caliphs of Baghdad, and so via the Samanids the concept of a slave army passed from Baghdad to the Ghaznavids.1

  Alptigin was followed by a series of other Turkish slave governors until the office fell in 977 to Sabuktagin (Sebüktigin), who held it for twenty years, ruling virtually autonomously.

  His eldest son, Mahmud, concentrated the administration of all the provinces in his own hands, and when the Samanid dynasty was extinguished in 999 he found himself undisputed master of Ghazna, Balkh, Bost and Herat, as well as Termez. The rest of Khorassan, namely Merv and Nishapur, was also under his control, but there he had to fend off the claims of the Samanid successors, the Turkish Karakhanids. In 1008, when the Karakhanid
rulers were defeated decisively, he ruled over an empire that stretched from western Iran to the gates of India. Over the next years he pursued a systematic policy of bringing under his control the outlying dynasties that had enjoyed loose tributary status under the Samanids. In this way he extended his territories to include Sistan, Khorezm and the principalities on the upper Oxus. Even before he had consolidated his holdings in Khorassan, he began to invade India, against which he conducted at least seventeen campaigns. He added north-west India and the Punjab to his empire and enriched his treasury by looting wealthy Hindu temples. Probably more important, his mullahs converted many Hindus to Islam, thus beginning a process that plagues the subcontinent to this day.

  The Samanid court in Bukhara served as a model to the Ghaznavid sultans not only in their organisation of administrative offices, but also as a centre of arts and Islamic learning. The Samanids patronised writers, scholars and in particular Persian poets. It was certainly the ambition of Mahmud of Ghazna to emulate his former masters in this respect. Unfortunately, Ghazna was still more like a military camp at this time, and hardly a place to attract a great number of illustrious men. In the course of his campaigns Mahmud brought entire libraries back to Ghazna and used his power quite ruthlessly to bring scholars in – by force, if necessary.

  The situation was similar in architecture and the arts. There can be little doubt that artisans and craftsmen were imported from the conquered lands just like the scholars and men of letters. Gradually, during the reign of Mahmud and Masud, ‘a certain style of building developed which used marble and carved decoration grafted on to the more sober traditional Persian technique of brick construction and moulded brick decoration’.2 Little has survived of the buildings of this period, the main reasons being, as listed by Bosworth:

  …the effects of an extreme climate; natural catastrophes like earthquakes and floods; the ravages of war; the use of comparatively perishable materials like sun-dried brick, for stone and even fired brick were infrequently used; indifferent workmanship; the theft of building materials by the local population: all these combined to make much building work, however splendid and imposing at the time it was put up, impermanent and short-lived.3

  Unfortunately, ‘no adequate description has come down to us of Ghazna at the time when it was rebuilt and adorned by Mahmud…’.

  After the defeat of Sultan Masud by the Seljuqs in 1040, Khorassan was lost to the Ghaznavids, thereby changing the cultural orientation of the remaining empire quite considerably. It acquired a ‘predominantly Indian outlook’,4 which its heartland had no doubt had before the arrival of Alptigin and Sabuktagin. Overshadowed by the growing importance of Lahore, the city lived on for another century, until the Ghorid sultan Ala-ud-Din surnamed Jahan-suz (World Incendiary) took Ghazna by storm in 1149 to avenge his brother’s death at the hands of Bahramshah the Ghaznavid. He then both sacked and burnt the city, which never recovered from the calamity. The tomb of the great Mahmud nevertheless appears to have been spared, unless it was restored, for Ibn Battuta saw it there in the fourteenth century.

  The mausoleum of Sultan Mahmud is in the village of Rauza, about half a mile outside the town, on the right-hand side of the road when coming from Kabul. It is a modern building constructed on the site of a palace in a garden that was once called Bagh-i Firuzi (Emerald Garden), and was one of Mahmud’s favourite resting places.

  The tomb inside was in all probability erected by Mahmud’s son Masud, and can be dated to the first half of the eleventh century. It is a marble-faced sarcophagus with a triangular prismatic slab on top. The slab is decorated on one side with a carved two-line inscription band in superb calligraphy, and a smaller band running along the base. On the other side is a three-lobed medallion with six lines of Naskhi script. In the middle of each side of the sarcophagus is an ornamental medallion framed with a flat Kufic inscription. The tomb shows the first traces of Indian influence in Islamic art. The bulbous arches in the medallions, as well as the naturalism of its floral decoration, already show certain tendencies that crystalised under the late Ghaznavids into what became known as Indo-Muslim art.

  Next to it, in the same village, is an undecorated Timurid building, the mausoleum of Sultan Abdul Razzak, built at the beginning of the sixteenth century and recently restored by an Italian archaeological expedition. Abdul Razzak was a one-time rival and challenger of Babur. The mausoleum is an interesting example of a central domed chamber in the middle of a square, flanked by four massive iwans, which are linked with fortress-like bastions and round towers.

  Excavations of the palace of Masud III nearby are the work of Italian archaeologists. A large courtyard has been uncovered, measuring 42 x 20m, which was paved with marble, with an iwan in the middle of each side, and surrounded by a number of buildings including a mosque, an audience hall, administrative offices and royal apartments. The main entrance was on the north side.

  The layout of the palace is similar to that of Lashkar-i Bazar, but the decoration is much richer. The four-iwan court was surrounded by buttresses, which probably supported arched niches. The carved marble panels show three-lobed arches decorated with palmettes, and scrolls with animals and birds; above these was a Kufic inscription in Persian, and higher still more scrolls and inscriptions in stucco and terracotta. Out of 510 epigraphic panels, 44 are (or rather were) left in situ, painted blue on a bright red ground. The whole inscription was c.250m long, an eulogy of the Ghaznavid dynasty modelled on the famous epic Shahname by Ferdausi and extolling also the beauties of the palace. It is one of the earliest surviving secular inscriptions glorifying both the building and its patron – a form which in the eleventh century may have been common throughout the Islamic world.

  Tepe Sardar is a mound hiding the remnants of a large stupa on a square base where a huge statue of a reclining Buddha was found. The building material was diaper masonry of the Taxila type. Rows of votive and commemorative stupas surrounded the main one.

  The medieval city of Ghazna once stood on the desolate plain between the village and the modern town. (See colour plate 24.) In the middle of the plain, accessible by a dusty road running westwards from the main road towards the town, stand two decorated towers, the most famous monuments of Ghazna and two of the very few surviving examples of Ghaznavid architecture and decoration. A sketch made in 1836 shows that originally they were more than twice as high. Above the lower octagonal part rose a circular shaft, which still existed when some early photographs were taken.

  The larger tower was built by Masud III, and should be dated between 1099 and 1114, but the smaller tower, with a much simpler decoration, is not older but younger and must be attributed to one of the last Ghaznavids, Bahramshah (1118–52). (See 68 and 69.) A detailed epigraphic analysis of the tower of Bahramshah was made by J. Sourdel-Thomine.5 This places it firmly under the reign of that last member of the dynasty who still resided in Ghazna. Architecturally, it belongs to the end of the eleventh century and should be compared, for example, with the minaret of Uzgen (Uzkend) in Kyrgyzstan. The decoration lacks the richness of the arabesque and floral ornament of the older tower, but its geometrical patterns in brick are, like the script, too sophisticated for the building to belong to an earlier period. Views still differ as to whether the towers were built as minarets or some kind of ‘towers of victory’.

  Two prehistoric sites in the Kandahar province are of interest. Mundigak, north-west of the city, and Deh Morasi Ghundai, south-west of it. Both were probably, in their earliest period, some 4–5,000 years ago, peasant settlements the economy of which was a mixture of farming and animal husbandry. But whereas Deh Morasi Ghundai seems to have remained a small semi-sedentary village growing wheat and barley and breeding sheep, goats and cattle until the end of its existence some time after 1500 BC, Mundigak developed into an urban centre comparable to the cities of the Indus valley civilisation. In fact, it is highly probable that both sites had connections with that civilisation. Mundigak ceased to be the urban centre in south
-east Afghanistan when Kandahar became urban in pre-Achaemenid times.

  The site of Deh Morasi is some 20ft high and 165ft long; Mundigak is 66ft high and 495ft long.6 Seven successive layers of habitation were uncovered at Mundigak, of which the fourth already shows signs of a fully-fledged town, with a granary and a massive complex, probably a temple and a palace. It seems to have been destroyed at least twice, in its fourth and fifth layer, but has been rebuilt each time. The sixth and seventh layer, which correspond to the periods around and after 1500 BC, appear to have been occupied only sporadically by nomads or semi-nomads.

  The old city of Kandahar – Shahr-i Kuhna – lies about 3 miles west of the modern town, and is dominated by the tall mound of the citadel, the Kasr-i Naranj.

  The site was excavated by the British Institute of Afghan Studies in 1974 and 1975. In the first year a trench was cut in the city wall, which revealed layers of five periods, of which the first roughly corresponds to that of Mundigak VI and may be as early as the first half of the first millennium BC. The pottery finds in its later two seem to belong to the Achaemenid period, and it is possible that the pakhsa wall of Period II defended the Achaemenid city of Harakuwatis.

  Period III is characterised by unbaked bricks and was, most probably, Greek, while the defences of Period IV, which consisted of a solid brick wall, are believed to be Kushan. The layer of Period V is Islamic.7

  The excavations of the second year concentrated on the high ground within the walled enclosure and had as one of their objectives to seek the solution of the problem of Alexandria of the Arachosians, or Alexandropolis, which might have been the Greek city of Old Kandahar. The layers revealed corresponded to the periods II to V as identified in the earlier dig. Although no evidence of the Alexander period was recovered, the finds of coins, pottery and some architectural remains would suggest that a Greek colony, or quarter, was added to the pre-existing settlement. This evidence of Greek presence at Kandahar certainly strengthens the hypothesis that Alexandria of the Arachosians was located at Shahr-i Kuhna.8