Treasures of the Great Silk Road

by Edgar Knobloch

  In 1924, the village of Balkh was framed by the old citadel mound in the north, by the ruins of two caravanserais in the east, by the ruined madrasa in the south, and by a Jewish quarter in the west. By 1947, this village has disappeared. A reconstruction was undertaken, based on a modern layout with a circular park in the centre, between the Green Mosque and the madrasa, and with several large avenues radiating from it. The intention was to rob Mazar-i Sharif of its present position and make Balkh the capital of the province again. This plan misfired, and Balkh remained what it is, a sleepy little town or large village, inhabited mostly by Uzbeks, in which the only life is to be found around the bazaar street.

  In the environs of Balkh, as in those of the Bukhara and Samarkand, there was a wall, 36 miles long, which surrounded the town and neighbouring villages. It ceased to exist in the ninth century, when both the shahristan (inner city) and the suburbs received their own walls. Under Chingiz-Khan, Balkh was completely destroyed following a rising of its inhabitants, and was still lying in ruins in the fourteenth century, when Ibn Battuta passed here.

  We crossed the river Oxus into the land of Khorassan and after a day and a half’s march through a sandy uninhabited waste we reached Balkh. It is in utter ruin and uninhabited, but anyone seeing it would think it inhabited on account of the solidity of its construction.1

  However, it was restored some time later, but did not regain its former importance. This is how Clavijo saw it, in 1404:

  This city is very large and it is surrounded by a broad rampart of earth which along the top measures thirty paces across. The retaining wall flanking the rampart is now breached in many places, but inside this last the city proper is enclosed by two walls one within the other, and these protect the settlement. The area between the outer earthen rampart and the first inner wall is not occupied by any houses and no one lives there, the ground being divided up into fields where cotton is grown. In the space between the second and the innermost wall there are houses, but still this part is not very closely crowded. The innermost circle of the city however is densely populated; and unlike the other towns which we have come to in these parts, the two inner walls of Balkh are extremely strong and as yet well preserved. They treated us with much honour in Balkh, providing us amply with provisions and excellent wine.2

  The walls of the city were examined meticulously by two French expeditions. Their findings confirmed that the Bala Hissar, the citadel with the fort, was the most ancient part of the city. (See 66.) South of it was a suburb with its own walls, which in one period extended to the east. Later, this eastern extension was abandoned and the city extended to the west. It is not certain how far back the history of the citadel goes. Pottery shards dating from the pre-Kushan period were found here, and some authorities are inclined to believe that the earliest fortifications could be older still. If this is right, then Balkh could indeed have been the legendary Bactra, capital of the Graeco-Bactrian kingdom of the third to second century BC.

  The funeral mosque of Khoja Abu Nasr Parsa, also known as the Green Mosque is the most conspicuous monument in present-day Balkh. The khoja, a theological lecturer in Herat in Shahrukh’s time, died in Balkh in 1460. (See 54.) The shrine – which is not a mausoleum for it contains no tomb – is a typical Timurid building comprising a large iwan flanked by two truncated minarets, with an octagonal structure under a ribbed dome on a high drum behind it. It was built near the saint’s tomb at the end of the fifteenth century.

  The colours of the facade are confined to white and dark and light blue, reinforced by discreet touches of black. It is the absence of purple and other warm tints which produces the silvery effect. This effect is continued by the dome, whose fat round ribs are covered with tiny bricks glazed with greenish turquoise… The building as a whole is unsubstantial and romantic. An unknown force seems to be squeezing it upwards. The result is fantasy, and in some lights, an unearthly beauty.3

  Byron was certainly a sensitive observer.

  The dome has been recently restored, but the tilework on the iwan and on the drum is original and rather damaged. The sides and the back of the octagon are bare. There is a band of decorated mukarnas forming the transition between the dome and the drum. On the drum are two lines of Kufic inscription in white tiles framed by black on a blue background. Interwoven in these bands are two small bands of highly stylised script in white tiles only. The portal screen of the iwan is framed by two vertical bands of elongated Kufic, each terminated at the bottom with a floral medallion. The corners of the screen are formed by two ‘corkscrew’ pillars (Byron’s expression) covered with alternating geometrical and floral ornaments. The bottom ends of these pillars are vase-shaped and richly decorated with floral mosaic faience in white and blue. A wide band of Kufic in the same colours and a band of mukarnas form the transition between the ‘corkscrew’ and the vase. The minarets, of which only the lower parts remain, stood immediately behind the pillars. They were decorated with geometrical ornaments in panels and bands, and with bands of two different varieties of script: a Kufic similar to that on the drum and above the vases, and a ‘cursive’ Kufic, or rather a special kind of Naskhi similar to that already seen on the eastern iwan at Gazurgah in Herat.

  The iwan niche has a geometrical decoration on the inside of the arch as well as on the entrance wall. Floral ornaments adorn the spandrels of the arch. Above the entrance is a wide rectangular band of script, again in ‘cursive’ Kufic, framing a pattern of square panels composed of a stylised angular script-like design around an octagonal medallion.

  Five-sided bays on two floors flank the iwan on both sides; the lower ones have been restored. The upper ones are covered with a semi-dome and are richly decorated with arched floral medallions and spherical floral ornaments in blue and gold.

  Inside, where women are not admitted, the mihrab niche is embellished with mosaic faience with geometrical and floral motifs. The squinches and the stalactite niches are adorned with inscriptions and floral ornaments. The dome, carried by a triple band of mukarnas, is decorated with ribs reproducing those on the outside and painted with repetitive floral motifs.

  Some 40 miles from Balkh, on the Akcha road, the mausoleum Baba Hatem is a monument from the Ghaznavid period. A cube-shaped, domed building, it has above the entrance an inscription frieze in floral Kufic, in incised terracotta and another inscription in decorative Kufic on the drum. Inside, the dome rests on squinches or quarter-vault pendentives. An interesting decorative element are stylised ibexes.4

  In the fields south-west of Balkh stands the oldest existing building in the area, and one of the oldest known monuments of Islam. It is the No Gumbad (Nine Domes) mosque, also known as the Masjid Hadji Piyade. It is rectangular on the inside, 60ft by 51ft, and almost square on the outside, 66ft by 64ft. As the name indicates, it has nine domes, 13ft in diameter, supported by massive columns measuring 5ft in diameter.5 (See colour plate 18.) The domes have collapsed, and the fallen masonry hides the lower parts of the columns, thereby preventing any measurement of their height. Some of the arches supporting the domes still exist. The mihrab was in the centre of the south-western wall. The material used was partly baked bricks and partly mud bricks (pakhsa), covered with alabaster stucco. The decoration on the arches and capitals is mostly of incised alabaster, and shows stylised floral motifs divided into geometrical fields and separated either by meanders or by simple straight or circular bands. There are also geometrical ornaments consisting of circles and half circles, octagons and four-leaf figures.

  The dating of this monument poses serious problems. Pugachenkova, who based her dating mainly on architectural analysis, found similarities with certain monuments in Iran and Turkestan dating from the tenth century. Nevertheless, it bears no resemblance to the early Arab type of mosque, and points to pre-Islamic models, for example to Sasanian palaces and temples. On the other hand, the decoration, which is ‘of exceptionally high quality’, points to ninth-century origins.

  The architecture reflects
aspects of three different forms: the iwan, consisting of three barrel-vaulted and arcaded naves; the apadana, which was a square with four columns in the centre supporting four cross-beams and thus dividing the whole into nine equal parts; and the Soghdian kushk, which was a square building consisting of nine cells covered with cupolas and divided by walls. Thus the architecture represents a mixture of foreign models imported from the West (Mesopotamia and Iran) and of traditional local ones. The fact that local models were used to facilitate the adaptation of the foreign ones, would point to a period fairly soon after the conquest of Islam, and yet later than the Samarra monuments. It would seem likely, therefore, that the building originated sometime in the first half of the ninth century. At any rate, the No Gumbad mosque belongs to the transitional period of medieval architecture, when it was still linked with pre-Islamic traditions but when new features were already beginning to emerge, foreshadowing the formation of a new architectural style, which would reach its peak between the tenth and twelfth centuries.

  Half way between Balkh and Mazar-i Sharif is the mosque of Takht-i Pol whose three domed halls alongside each other indicate Indian influence. The halls have slightly pointed arches with mukarnas decoration. There is an iwan in front of the central hall; this hall is higher than the lateral ones and contains the mihrab. The interior is richly decorated with paintings in a good state of preservation, showing a mixture of late Indian and Islamic traditions.

  Mazar-i Sharif, the present provincial capital, is a modern town centred around the shrine of Hazret Ali, one of the holiest places of the Shi’a. Tens of thousands of pilgrims flock here, in particular for the New Year celebrations (Now Ruz, 21 or 22 March) to pray to Ali, the son-in-law of the Prophet and the fourth caliph, who was murdered in Kufa in 661. (See colour plate 23.) According to legend, his followers put his corpse on a white she-camel and let her loose. She wandered as far as the neighbourhood of Balkh, where, some four centuries later, the body of Ali was found intact. A shrine was built over the tomb in 1136, but was destroyed by Chingiz-Khan. The tomb was rediscovered in 1480, and Husayn Baykara erected a new sumptuous mausoleum over it. The building then underwent numerous restorations and reconstructions, so that hardly anything remains of the original Timurid structure.

  The main building, with an imposing iwan, stands in a large courtyard. There are three gateways into the yard, on the north, east and south sides; the main gateway, flanked by minarets, is in the south. The west side is taken up by a large mosque. Originally, there was a gateway similar to the others, the dome of which had been incorporated into the mosque. The pattern of the pavement in the yard consists of outlines of innumerable mihrabs, which are used by worshippers at prayer time. Non-believers may enter the yard but not the sanctuary.

  The building has two pointed domes indicating an inner and outer sactuary. Inside, there is first a carpeted and decorated anteroom in which, by the door to the inner chamber, stands a huge bronze cauldron similar to that in the mosque in Herat, but with no decoration. The tomb of Hazret Ali is in the inner chamber, surrounded by a railing and covered with embroidered cloth. The walls and ceiling are decorated with painted floral ornaments dating from the second half of the nineteenth century. The entire outer façade is covered with glazed tiles, mostly turquoise and blue, but the ornaments are not of any special interest. On the outside, a number of individual shrines are built against the walls of the sanctuary. Several tombs on the west side belong to the members of Amir Dost Muhammed’s family.

  By far the most exciting site in the region of Kunduz is Ay Khanum (Moon Lady), which was discovered in 1963 by French archaeologists at the confluence of the Amu Darya and the Kokcha. It is the eastern-most genuine Greek city, a proof of the Hellenistic period in Bactria sought by archaeologists for decades. The site, which can be dated to between the fourth and the second centuries BC, is situated on the left bank of the Amu Darya, and comprises an upper town with a huge citadel in its northern part and a lower town based on a regular grid of broad straight streets with residential and administrative areas, including a palace with a peristylar courtyard, a Greek-style gymnasium, a large private house, an Oriental temple and other buildings.6 It had gates and ramparts with rectangular towers more than 66ft wide. Water was brought by canals from the river.

  The ceramics and terracotta roof tiles resemble those found in other Hellenistic cities from the same period in Western Asia and on the Greek mainland. Persian-style column bases and Corinthian capitals provide evidence of the co-existence of the two civilisations. Two important Greek inscriptions in stone have been found here, which confirm the overall Greekness of the site.7

  The city was destroyed by fire, most probably when the area was invaded by the nomads sometime at the end of the second century BC. It has never been reoccupied, the most plausible reason being that the destruction of the irrigation system rendered the region uninhabitable.

  The site of Ay Khanum at the banks of the Oxus (see below), has been systematically looted and pillaged for years. Even heavy earth-moving equipment has been brought in by looters. The destruction of the city thus being nearly complete, a reconstruction in 3D has been attempted by a Franco-Japanese team inspired by the Japanese film The Illusive Alexandrias, shown in 2003.8

  In 1946, at Khisht-Tepe, some 56 miles from Kunduz on the bank of the Amu Darya, the frontier guards from the district of Kala-i Zal found an earthenware vase full of coins when digging foundations for a new stable. There were 628 Bactrian coins of various sizes and denominations with the names, in Greek, of twenty-two rulers. This find, which became known as the ‘hoard of Kunduz’ provided some useful information for the dating and history of the last decades of the Bactrian kingdom. It is now accepted that these coins date from the second century BC; the latest of them come from the last two decades of that century, from the period immediately preceding the invasion of the Yue-che and the establishment of the Kushan Empire.

  Some 9 miles north of Pul-i Khumri, within sight of the main road to Mazar-i Sharif, Surkh Kotal is the site of a Kushan temple from the second century AD that provided ‘the first definite evidence of an indigenous Bactrian art, possible inspiration for the later Gandharan style.9 (See 60.) It was discovered in the course of road building in 1951, and was excavated systematically by the French archaeological expedition between 1952 and 1963. The excavations revealed a hilltop complex consisting of the main temple, a secondary temple and a monumental staircase leading down the hill. The main temple stood on a brick platform faced with stone revetment decorated with pilasters. It had strong thick walls, and contained a square central room surrounded by corridors on three sides, with three entrances facing east. In the centre was a stone platform with a huge column in each corner. The bases of these columns are still in situ. The temple building stood in the centre of a large paved courtyard with portico and numerous niches in the walls, where large painted statues once stood. From the outside, the whole complex looked like a fortress, with solid walls, a series of towers and narrow entrances. The secondary temple beyond the outer wall on the north side contained a square fire altar. The 23ft-wide staircase starts from a huge terrace on the east side of the main temple. As the archaeologists followed it downhill, they discovered that it consisted of five flights and five terraces.

  In 1957, a large limestone slab, covered with Greek letters in an unknown language was unearthed on the fourth terrace. It has been subsequently found that the twenty-five-line inscription, ‘probably the most important single specimen from Surkh Kotal’10 is in the Eastern Iranian or Bactrian language. It has not yet been fully deciphered, but mentions King Kanishka and refers to the repairs of the sanctuary. There seems to have been a fire during the reign of Kanishka’s immediate successor, and it can be assumed that these restorations and repairs were undertaken soon after the fire. This would place the inscription – and the fire – in the second half of the second century AD. The inscription also includes the word ‘Bagolango’, Old Iranian for ‘temple’ or �
�sanctuary’. It is easily recognisable in the present-day name of Baghlan. In the Middle Ages, Baghlan was the name of the whole district, thus meaning ‘district of the sanctuary’. The original temple of Kanishka extended down to the third terrace; the fourth was the work of the restorers and the fifth was probably added between the third and the fifth century AD after the Sasanians had succeeded the Kushans. Although Surkh Kotal was built in the peak period of Buddhism, it was certainly not a Buddhist temple. It is assumed that it was a dynastic temple dedicated by King Kanishka to his own divinity. A huge headless statue was found here that could represent Kanishka himself. It is possible that the statue, which is now in the Museum of Kabul, was a cardinal deity in the temple. On the other hand, a stone platform like the one in the central room was typical of the Iranian fire temples and the secondary temple did contain a fire altar. There is little doubt, therefore, that the cult of fire played a leading role in the rites practised here.

  Like Ay Khanum, Surkh Kotal is not a Graeco-Buddhist site (or, to use another terminology, Indo-Greek). It is a mixture of Greek and Persian elements, thereby evidencing that the traditions of Achaemenid Persia were still strong several centuries after its disappearance.

  To mention only a few examples of the Kushan mixture of Greek and Persian in Surkh Kotal, we may select the following: they wrote their own language in Greek letters; they worshiped their own King Deity with fire; they built a fire altar inside a turreted fortress but surrounded the courtyard with a Greek portico; for decoration they used Iranian motifs such as stepped battlements and arrowheads together with typically Greek motifs such as Corinthian capitals and garlands borne by Amorini; they clothed their figurines in oriental and Central Asian dress and Greek drapery.11