Begram Web

With regard to the finds as a possible royal ‘treasure’ hoarded over time, the architectural evidence uncovered from Begram fails to support either that Begram had been a royal residence of the Kushans or that the site was ancient Kapisi or any other Kushan royal capital. The nature of the physical structures discovered at Begram would seem to suggest that the site was something considerably less than royal. Of the rooms excavated in what has been termed the ‘palace’ within the ‘New Royal City’, the two largest were Rooms 10 and 13, those used to store the objects which made up the ‘treasure’. Yet these two largest rooms were of the less than regal size of approximately 6 x 8.5 m. and 6 x 10 m., respectively, and all the other rooms excavated were rather smaller (Fig.23). Although only part of the site at Begram was fully uncovered and thus one cannot conclusively determine what the complete site looked like, the excavated area certainly lacked rooms of the size and grandeur -- no royal statuary, architectural features, or reliefs -- one would expect for the summer ‘palace’ of a Kushan emperor and his court.

Much of the problem about the dating of the ivory and bone carvings is exacerbated by the scant and often confusing information obtained through Ghirshman’s excavations at Begram. There seems to be little controversy about the era of the construction of the buildings in which the ivory and bone objects were discovered, that is 1st century CE. The problem centers around when this second -- at least according to Ghirshman -- period of occupation came to an end. Ghirshman’s hypothesis was that an imminent Sasanian invasion caused the abandonment of the second city of Begram (Ghirshman 1946:99). The sealing of the two rooms in which the ivory and bone objects were found suggested to Ghirshman that the inhabitants sought to protect them from this attack, perhaps for later retrieval. Deep layers of ash discovered near the enclosure wall indicated to Ghirshman that the city was sacked in that invasion. To secure a date for the invasion, Ghirshman noted that the most recent coin type found in the second city of Begram was that of Vasudeva, the last king of the Kaniska era, whom he places in 220-230 CE. Ghirshman’s invasion theory was, in his eyes, validated by an inscription of Shapur I discovered on the Ka’bah Zardusht at Naqsh-i-Rustam, in which the conquest of Kushan territories in parts of modern Afghanistan and Pakistan is mentioned. Ghirshman placed this invasion around 241 CE (Ghirshman 1946:100). But if Ghirshman was correct about the destruction of Begram II by a Sasanian invasion, Vasudeva should not have been the last ruler of this period unless coins remained in currency longer than their ruler [5] . The most recent chronology of Kaniska, places him in the early 2nd century CE (Sims-Williams and Cribb 1995/96:106), which would place Vasudeva at the end of the 2nd or early 3rd century CE rather than in the mid-3rd where Ghirshman placed him.

The problem with dating layers on the basis of coins is further illustrated by the numismatic finds uncovered in Rooms 10 and 13. During the campaigns of 1937 to 1941, approximately four hundred fifty coins were discovered in the ancient city of Begram, ranging in dates from the Graeco-Bactrian kings to the various rulers of the Kushan Dynasty [6] . Relatively few coins were uncovered in the two rooms which harbored the Begram ‘treasure’, most of these of the early kings of the Kushan dynasty. But a lack of precise dating information about most of the coins, plus the incompleteness of the excavation record concerning the exact location and position of each, have meant that even those coins which have been identified raise far more questions than they answer.

For example, such mixed messages are sent from the coins found in Room 13. A total of twenty coins were uncovered there, with nine of them having been identified. At a depth of 2.10 meters was discovered a coin of Kujula Kadphises, founder of the Kushan dynasty sometime in the early to mid-first century CE. This was relatively close to the depth at which most of the ‘treasure’ objects from Room 13 were discovered (2.30 m.). But at a depth of 2.50 meters, lower than both the Kujula Kadphises coin and the ‘treasure’ objects, a coin was found from one of Kujula’s successors, Vima Kadphises, who probably ruled some time in the mid to late first century CE. Also found at a depth of 2.50 meters was a copper coin of Gondhophares, an Indo-Greek king and contemporary of Kujula Kadphises. To further complicate matters, six coins found at 1.80 meters were from the reign of Vasudeva I, whose later Kushan rule can most probably be placed in the second half of the second century CE. Thus, within less than three-quarters of a meter, we have a variety of coins which range from the early to mid- 1st century to the late 2nd century CE [7] .

Ordinarily, one might try to draw conclusions from the specific depths at which given objects are found, but in the case of coins in general and these coins in particular, such layer-dating is problematic. In the first instance, even if stratigraphic layers presented a linear progression of dates, there are no specific dates affixed to the reigns of Kaniska and the other Kushan kings, so that identification of a coin as from one of their reigns does not provide precise enough information from which one could transfer a date to other objects found in the same layer. Moreover, in the case of the Begram site, the ‘layers’ in which the coins were found were themselves far from precisely defined. In addition, the locations of the coins and their relations to one another were not described in detail in the excavation reports, and the distances between and among the coins which were noted are so small that it is difficult confidently to ascribe the significance of separate ‘layers’ to them. For example, the coin of Kujula Kadphises was found at a depth of 2.10 meters, only 0.20 meters shallower than most of the ‘treasure’ objects. But a coin of Vima Kadphises, Kujula’s son and successor, was located deeper than that of the earlier king, just the opposite of the situation which would exist if the ‘layers’ corresponded in a linear fashion to the successive eras.

This anomaly of coins found seemingly in reverse chronological order highlights the general problem of trying to tie the coins to the same ‘layer’ as objects found nearby. Because of their small size, coins are particularly subject over time to random redistribution as a result of earth movements both natural and man-made. Coins filter down more easily than larger objects, and because of their light weight they are also more easily churned upwards. Thus, when the distances among coins of different eras are as small as they are in the case of the Begram ‘treasure’ rooms, it is perilous to assume that the depths at which they were initially located were the same as either the depths at which they were found or the depths at which nearby objects were initially located.

Further, while they are still in currency, coins are not normally stored in a fixed location, as are pieces of furniture or other domestic goods; their very nature as currency makes them unlikely to be stored away in sealed rooms. Thus, if the sealed rooms of the Begram ‘treasure’ were a hoard, as some scholars suggest, there would be no reason for local, still current coins to be included, and indeed no scholar has argued that the coins were a part of the hoard. Similarly, if the sealed rooms were a merchant’s store rooms, there would be no obvious reason for a merchant to store away coins with the merchandise. In either case, then, it seems unlikely that the coins found in the rooms were intentionally placed there at the same time as were the ‘treasure’ objects. And finally, even if one could imagine that as part of a hoard of diverse objects someone would have included a number of out-of-date coins as artifacts, the fact that they were out-of-date when stored would render them misleading in any attempt to use them to date other stored items.

Given the wide range of uncertainties attached to any dating inference which might be drawn from the Begram ‘treasure’ coins, most scholars have chosen not to rely on the coins as a basis for making deductions about dating the finds. This writer is aware that four of the earlier coins -- the Kaniska in Room 10 and Kujula Kadphises and Gondophares in Room 13 -- were found relatively close to the depth of the ‘treasure’ objects, and that from this fact an argument could be fashioned tying the finds to the eras of those rulers. Then, depending on the proposed dates for the beginning of the Kushan era -- early to mid 1st century CE -- one could argue for the ‘treasure’ to be dated similarly [8] . However, the coins alone simply do not provide any particular leverage for one proposed date over another.

[5] Kuwayama (1974:59), in an article on the dating of Begram III, noted that excavators in Afghanistan such as P. Bernard have found that a few coins of rulers are sometimes uncovered from layers that are not contemporary with their reign. However, it is unclear what significance, if any, these coin finds have. And since it is very hard to tell from Ghirshman even what stratigraphic layers exist at Begram, this information is not helpful with regard to dating Begram.

[6] See Ghirshman, 1946, pp. 85-87. Thousands of coins were found on the plain of Begram by Charles Masson and others. However, since none of them was found in an archaeological context they are only useful for a general discussion on the chronology of the region and not for dating the ancient city.

[7] In Room 10, a coin depicting Kaniska on its obverse side was found at approximately the same depth (2.60 m) as the objects of the ‘treasure’. Seven other unidentified coins were found. If even one of these is determined to be from an era later than Kaniska’s rule, no conclusion at all could be drawn from the coins about the dates of the other objects found nearby.

[8] This would support this thesis’ wider argument for a 1st century date for the finds.

Begram Ivory and Bone Carvings