The Decline of Urban Life, I: Spolia in Defensive Architecture


Such large quantities of spolia are available in our area because, probably from the 6th century, there was a decline of urban life; and although the process probably began from a more prosperous base than it did in the West, and in cities of greater magnificence, its tell-tale features are to be seen throughout Turkey, Syria and North Africa. (Of course, not all antique structures or fortresses were in cities, and many such artificial emplacements also declined, releasing spolia, as conditions changed[1].) Taking as a suitable measure the twenty cities of Byzantine Asia named in a list compiled by the tenth-century Emperor Constantine Porphyrogenitus, Clive Foss[2] has concluded that in Turkey there was a significant decline in standards of urban life, of population and of commerce; and he gives summaries of the situation at centres such as Ephesus, Sardis, Miletus, Priene, Pergamon, Smyrna and Magnesia. Similarly, there are indicators of decline in Byzantine North Africa, such as much smaller new churches[3], although the extent of city-dwelling is a matter of fierce debate[4].


As in the West, three factors determined mediaeval use of (and perhaps attitudes to) earlier buildings:


1.       The range of building types was now too large for reduced needs; so that as activities declined, so the buildings they supported (from the bath and the agora to the colonnaded street and the temple) became surplus to requirements, probably in the course of just a few generations. Fountains lost their water; and the pipes became furred, as at Perge, where this still in evidence at the acropolis fountain and in the baths to the north of the south gate – surely signs of long-time neglect;

2.       The longevity of antique walls and fortresses provided a constant reminder of the greater glories of the past. Strongly built, they long survive more fragile structures such as temples, and many continue in use when temples decay or, like Baalbec or Didyma, are converted into forts.[5] Mediaeval fortress builders would have been well aware of previous achievements in the genre, the more so because so many of them reuse the materials on site. To take examples from just one area - the Morea - the fortresses of Corone, Zarnata, Old Navarino, Methone, Arkadia, Argos, Patras are all in this category[6].

3.       Finally, the inability of reduced populations to defend existing enceintes required the construction of smaller ones. One factor promoting the use of spolia is the dearth of quarrying. It is probable that quarry production declined drastically (but perhaps never ceased) in Turkey during the Byzantine period[7], just as it did further West - so the re-use of earlier architecture and sculpture from now-unneeded buildings was an obvious course of action to take. Again, no proof has emerged yet of the re-opening of quarries in the 9th century in Proconnesus, Attica, Phrygia, or in the Islands of the Aegean. Sodini has suggested[8] that quarries slowed down and perhaps stopped between proto- and medio-byzantine periods, with a decrease in colonnaded streets, and basilican churches replaced by small Greek-cross employing few columns, so that diversity is replaced by Empire-wide homogeneity. Perhaps, therefore, the very diversity of spolia in a sea of uniformity is a factor in their popularity, especially when we add Sodini’s observation of  a progressive return to antiquarian sculpture, especially in 14th century Constantinople.


What happened to existing sets of city walls in the East when urban life declined? Apparently, just what happened in the West. From records of refurbishment in the Late Empire, with proclamations of renovatio which we have already surveyed, it is clear that the concept of the city as an entity protecting its citizens continued but, given a lower population, there seem to be no cities in Turkey (with the exception of Constantinople) in which the originally designed or upgraded circumference of walls could be defended. At Ephesus, for example, large areas of the city were in ruin by the time of Diocletian, at the end of the third century AD; and although some monuments were repaired under the Tetrarchs, with a spurt of work in the fourth and early fifth centuries, and some prestigious Christian ones built (such as the Church of St. John), the city soon had to respond to the Persian invasions. Ephesus built a new wall, using spolia, encompassing only a proportion of the classical city (abandoning, for example, the Agora), and incorporated within that wall suitable monuments, such as the theatre. We find older materials being regularly re-used for refurbishment - such as in the Baths of Scolastica, or in the great marble streets at Ephesus, the flanking colonnades of which are assembled from a curious mixture of columns of different materials and girth, all of which must have been robbed from earlier, now derelict buildings.



The Decline of Urban Life, II: The Stripping of Cities


In spite of building walls with spolia, so small was the population in our area that we must picture even greater quantities of building material than still survive lying for centuries in deserted antique cities - which was also the case in the early mediaeval West. And the desire for sturdy construction caused castle builders to go in search of convenient building blocks and other members, which they found in great quantities by stripping decayed cities. A large number of these in Turkey, and several of the important ones in North Africa, are situated on a river and/or near the sea, so transport was no problem[9]- even, perhaps, from Greece to Venice.[10] Post-mediaeval maps and charts, whilst not an infallible guide, indicate that many of the antique centres were calling-points, or points de repère, for trade, as they were earlier for pilgrims.. Beaufort's survey of the south coast for the Admiralty in 1810-12 notes[11] that, In a naval point of view, the resources of the coast of Karamania must be considered important; and he goes on to list the quality of the harbours, water, wood, and food supplies, none of which had diminished by his day. Much still remains on such sites, but it is certain that large numbers of buildings have completely gone - and now reside in the walls of mediaeval fortresses. As we shall see in a survey of spolia defences, the walls of Nicaea  and those of the lower fortress at Seljuk, amongst others,  provide conspicuous examples of this process.


There are four reasons for stripping cities of their materials:


1.       The re-use of fallen antiquities in houses, or to make walls for sheep-pens and the like. Thus the frons scenae of the theatre at Alabanda is still a pig-sty with the farm attached; and at Uzancaburc village walls are still constructed from column drums, just as they were of Doric capitals at Assos. Another good place to see this would be Aizanoi, where the people still live amongst and still make use of the ruins, because the site has not been tidied up by the archaeologists: it is still a lived-in antique city, and we can still walk around it with Fellows’ description in our hands, and make full sense of what we read.[12]

2.       As well as building materials, mediaeval searchers also needed metal, especially lead and iron, for daily needs as well as for weapons and especially missiles - but not for cramping mediaeval walls. Both these metals were used in Antiquity as cramps (sometimes iron covered in lead) to hold together the blocks of large antique walls. There was perhaps plenty of metal to be found in exposed walls. At Nyssa, for instance, the retaining walls of the theatre were half-buried long ago by a land-slip; and although the upper sections, dirty through long exposure to the air, have been robbed of cramping materials, nobody bothered to dig into the earth to rob the lower sections which, recently uncovered, are now fully in view, pristine-clean and undamaged;

3.       The thirst for antique columns to be used in contemporary buildings is important all over the ex-Roman world because of its effect on the dismantling of antique structures, Matching sets seem to have been as prized in  the East as in the West, and the idea has a long pedigree: for example, S. Agnese fuori le Mura, Rome, where great care has been taken, and the best columns go nearest the altar; and Old S. Peter's[13]. In the East, we find Justinian importing large and impressive columns from various parts of the Mediterranean for his building projects in Constantinople. There are few mosques of the Seljuk period or later which do not boast several antique columns or pieces of architectural decoration, and it sometimes seems that remarking on matching sets of columns might indicate that make-and-mend sets were commonplace[14]. The consequent destruction of any building of which columns formed a part is easily illustrated: Bernardo Michelozzi and Bonsignore Bonsignori went to Turkey in 1497/8, and wrote interesting letters back to their home city of Florence[15]. They visited one of the great temples, that of Hadrian at Cyzicus, and were amazed by its dimensions, with 70-foot columns and Corinthian capitals nine feet in height. Others had been impressed before them: Cyzicus’ convenient location near the Sea of Marmora meant that Justinian took spoils for Hagia Sophia (completed 548), and the Turks (much later) for the Suleymaniye Mosque, building between 1550 and 1557. Material was also disappearing in the 15th century as well: for Ciriaco, who visited and drew the site in the middle of the century, counted more columns than Michelozzi. We do not know whether columns were still being carted off for use (and, if so, to where); but Bonsignore does refer to the Turkish propensity for refashioning the drums (10 of which made up one column) into cannon balls (this is dealt with in detail below), and Ciriaco refers to the destruction of the temple in his day by lime-burners.

4.       Finally, and particularly relevant to our purpose, enormous quantities of spolia disappeared into defensive walls, frequently for display, but sometimes as mere infill. An example of "hidden" antiquities from Greece would be the protection afforded the pedimental sculptures of the Temple of Zeus and, for Turkey, the preservation of the reliefs of the Great Altar of Zeus, at Pergamum. Because of the monumental nature of the original constructions, the walls built with blocks from older monuments are often very thick, and sometimes even double-skinned (as at Saintes or Bordeaux in France); hence they have been a prime agent in preserving both bas-reliefs and architectural members;


Everywhere, the later the (post-antique) building, the less likely it is to find a set of matching columns, let alone capitals and bases. Such "marble starvation" is perfectly obvious in religious buildings, where various devices (differential heights for impost blocks or bases; adding sections to damaged columns; reserving the good material for the presbyterium) must needs be employed. (It may very well have happened in the East precisely because of the profligate use of column shafts in fortress walls.) Here are two contrasting Turkish examples, one early, one late:

1.       In an interesting case in AD 400,  Bishop Antoninus of Ephesus was charged with theft, and actually deposed, for going against the strictures of Constantine the Great and taking columns from the adjacent church to decorate his dining-room, and marble from the baptistery for his bath[16] (public baths may still have been functioning in Lycia in the 6th century or even later[17], as they probably were in Antioch, and displaying art as well[18]). Parallel examples can be demonstrated from the ongoing excavations at Aphrodisias;

2.       The second example is the Great Mosque at Adana (a city not far from a whole range of antique sites) which was built beginning in 1513, and opened for worship in 1541. There are re-used classical columns in the partly-covered courtard and also inside the mosque. Five large, matched grey-white columns grace the interior, but there were insufficient sets of smaller columns to use in the courtyard. The builders have disguised this fact by adopting two procedures (both of them seen further West as well, and in profusion): they have alternated granite with marble columns for the front rank, choosing complete columns for the purpose; then, for the posterior rank, they have carefully pieced together column fragments, sometimes from similar but not identical marbles. Indeed, closer study shows that even the front-rank marble and granite columns, although similar, are by no means identical in material, height or girth. From such painstaking work we may certainly deduce the scarcity of suites of columns, but also suspect that it reflects the prestige in which antiquities were frequently held.


Such marble starvation occurred even in towns such as Seljuk, and it was apparently not satiable by looking around nearby Ephesus, probably because most useful and beautiful material had already been picked over by the fortress builders, and used in enormous quantities in the road-facing defensive walls; so that anything further was difficult to retrieve. Evidence for this can be found in the Isa Bey Mosque, just below the lower fortress, and built in 1375. The courtyard is adorned half-and-half with granite and veined grey-white marble columns, with different bases and capitals - some of the latter antique, others in Moorish stalactite design. And although the granite columns are placed together, they match neither for height nor girth. Inside the mosque, the enormous granite columns (for which the builders have provided new stalactite capitals) can only have come from some very large Roman building at Ephesus.




[1] J. D. GRAINGER, The cities of Seleukid Syria, Oxford 1990, p.199, emphasizes the artificiality of the cities of Seleukos I which, after the Islamic takeover, rapidly slid down into ruin and desertion and decay. They were valued for their fortifications at times, but not as cities;

[2] C.   FOSS,  Late antique and Byzantine Ankara, in Dumbarton  Oaks Papers 31, 1977, 27-87; see p. .27;

[3] CHRISTIE, The archaeology of Byzantine Italy, cit., p. 265;

[4] A. WALMSLEY, Byzantine Palestine & Arabia: urban prosperity in late antiquity, in N. CHRISTIE & S.T. LOSEBY, editors, Towns in transition: urban evolution in late antiquity, Aldershot 1996, pp.126-58, sees prosperity and substantial growth (trade, pilgrims) until the 6th century;

[5] W. MUELLER-WIENER, Mittelalterliche Befestigungen im suedlichen Jonien, in Istanbuler Mitteilungen, 11 (1961), pp. 5-122: see 36 fig 8 for the placement of the castle on the temple;

[6] K. ANDREWS, Castles of the Morea, Amsterdam 1978 for details;

[7] M. WAELKENS, editor, Pierre éternelle du Nil au Rhin: carrières et préfabrication, exhibition, Brussels, 1990, 53-72. M. WAELKENS, Technique de carrière, préfaçonnage et ateliers dans les civilizations classiques (monde grec et romain). Cf. N. ASGARI, Objets de marbre finis, semi-finis et inachevés du Proconnèse, pp. 106-26: p. 109: production slowed down in the Byzantine period, and resumed again only in the 18th-19th centuries;

[8] J.-P. SODINI, La sculpture médio-byzantine: le marbre en ersatz et tel qu’en lui-même  in MANGO & DAGRON editors, Constantinople and its Hinterland (27th Spring Symposion Byz Studs Oxford 1993), Aldershot 1995, pp. 289-311;

[9] cf. Ferrara, in UGGERI, Il reimpiego dei marmi antichi, cit., p. 612;

[10] MAYER, ALVAREZ & RODA, Los materiales lapideos, cit.: in the discussion, R-H Bautier cites a notorial act from the early 14th century at Venice, when Marco Polo (cousin of the traveller) concludes a contract promising to transport a Greek temple - colonnes, plaques et ornements sculptés, de Nègrepont à Venise – but gives no reference;

[11] MS of BEAUFORT, Karamania, in PRO ADM7/847, pp.34ff;

[12] FELLOWS, Journal, cit., p. 47: at Assos, in one place I saw thirty Doric capitals placed up in a line for a fence – and he draws the sally port in the walls, and praises them highly; pp.137ff. for Aizanoi;

[13] J. CHRISTERN, Der Aufriss von Alt-St-Peter, in Roemische Quartalschrift fuer christenliche Altertumskunde und Kirchengeschichte, 62 (1967), pp. 133-83; see pp. 172-5 for the aesthetics of the layout of the nave and some of the aisle columns;

[14] RAMUSIO, Delle Navigationi, cit., vol. II, 78ff, Viaggio d’un mercante che fu nella Persia: the quibla of the main main mosque at Tauris has square vaults, lequali sono sostenate da colonne di marmo, ch’e di tanta finezza, e cosi lucente, ch’assomiglia al cristallo fino, e sono tutte d’una medesima longhezza, e grossezza;

[15] E. BORSOOK, The travels of Bernardo Michelozzi and Bonsignore Bonsignori in the Levant (1497-98), in Journal of the Warburg and Courtauld Institutes,  36 (1973), pp. 145-97;

[16] C. FOSS, Ephesus after Antiquity: a late antique, Byzantine and Turkish City, Cambridge 1979, p. 521;

[17] A. FARRINGTON, The Roman Baths of Lycia: an architectural study, Ankara 1995, p. 119.

[18] F. YEGUL, Baths and bathing in classical antiquity, Cambridge MA & London 1992, pp. 314-349 for late antiquity and early Byzantium;