Remarks on Samarra and the Archaeology of Large Cities
Northedge, Alastair, Antiquity
In 1995 Roland Fletcher published a work entitled The Limits of Settlement Growth, in which he looked at the possibilities of settlement size in ancient cities. While much of his analysis does not concern us, the question of the world's largest ancient cities is an interesting one. He distinguished between 'compact' and 'dispersed' cities, where in the latter case occupation of low density was spread over a large area, with perhaps a ceremonial centre. A size limit of 100[km.sup.2] was postulated for pre-modern cities, before the introduction of modern communications. The examples of 'compact' cities used were, in China, Chang'an under the T'ang dynasty, Loyang, Heifeng and Hangchow under the Sung, Peking under the Yuan and Manchu dynasties, and Nanking under the Ming, in Japan, Edo under the Tokugawa, Kyoto and Nara, in India, Vijayanagar and Delhi, in the Mediterranean, Rome, Constantinople and Carthage, and in the Near East and Islamic world, the complex of Seleucia on the Tigris and Ctesiphon, Baghdad, Samarra, Cordoba, Cairo and Isfahan under the Safavids.
The size limit for a 'compact' city would certainly be less than he supposed, for a 'compact' city may have included low density elements, and, if the information about the plan is not too detailed, one can be mistaken. For example, Chang'an is known to have included fields and villages within its walls. There is also the issue, for flat open sites, of which areas were occupied when, for the centres and occupied areas evidently tended to move over time. What may appear to be a very large city complex, may not in fact be so, as not all sections would have been inhabited at the same time. Delhi and Ctesiphon are good examples of this.
Most importantly, there is a question of stability. It seems likely that the very largest cities of ancient times, comparable in size to modern cities of the nineteenth and twentieth centuries, were unstable and could not be expected to last more than a generation or two. Those cities which retained their dimensions over many centuries, such as Rome, Constantinople and Cairo, were relatively small, whereas Cordoba is quite a good example of a city which remained stable within its walls, but during the height of the Umayyad dynasty of Caliphs in Andalusia in the tenth century expanded to a vast size for a limited period of less than a century. No doubt the existence of 'mushroom' cities of the greatest dimensions was stimulated by unusual economic opportunity, and the import of foodstuffs, whose supply was naturally of doubtful stability.
However, the striking point to be drawn from the presentation is how little is known about the largest cities of antiquity, and how small an information base there is to work on. While it is naturally not possible to excavate much of a large city, the sites of nearly all these cities are covered by their modern successors, or surrounding cultivation, and even the dimensions cited are in question. In the case of Imperial Rome, naturally, centuries of work assembling the texts, excavations and surface features have led to a good state of knowledge. At the other extreme, in the case of Baghdad under the Abbasid caliphate, not one feature of the massive early city of the eighth and ninth centuries AD has been archaeologically recovered, and the size of the city is simply a guess, based on the texts, although Lassner calculates that the area must have been of the order of 70[km.sup.2] (Lassner 1970: 157; see also Adams 1965: 89).
Samarra is the only exception known to this author, where the question can be studied in any detail, for the city was built in a steppic environment, and only a small town survived into modern times, leaving the main area uncovered and uncultivated. The main point to be made in this paper is the obvious one that when one does have access to detailed information about such a large city, the actual development can be seen to be complex, and in many ways specific to its case. …