Examples of Cities Stripped to Build Defensive Walls


The usual reason for stripping cities is convenience: nearby monuments are piled up on existing structures: for example, the Byzantine fort at the top of the theatre at Miletus (a radical shrinking of the city area), or the fortress built over the gymnasium at Stratonikeia (where not enough is known about the wall circuit to form a judgment). In some instances, brutality verging on violence is done to noble antiquities, conspicuously at Limyra, where the Byzantine defences march over the corner of the Hellenistic Ptolomeion (a cubic base with an Ionic tholos above), and reuse some of its splendid cannellated columns and upper structure therein (together with seats from the theatre nearby). If ever there were an opportunity for conspicuous display, this was it; and it was ignored - so, clearly, not all Byzantine wall builders were attuned to the aesthetics of ancient architecture, or at least, wished to capitalise on them. Frequently, however, earthquakes conveniently did the initial dismantling, both in Turkey and in North Africa. Similarly, an earthquake may have been responsible for the abandonment of the fortress built on the ruins of Baalbec[1].


Excellent examples of how antique cities were stripped to provide for mediaeval fortresses and walls are Nicaea and Seljuk, Ankara and Korykos. But there are plenty of others, such as around Byblos (Lebanon), where the splendour of the mediaeval fortifications relies on their antique materials.





The north facade of the castle at Byblos (Giblet) followed the line of the Roman road by the acropolis, and incorporated a double colonnade, no doubt the easy source of the column shafts used en boutisse in its walls. There is no doubt about the extent of reuse here because, as Longchamps remarks, the Crusaders built atop Phoenician, Persian, Graeco-Roman and Fatimid walls.[2] The fortifications to the town make great use of column shafts, which project some three or four inches from the plane of the wall, almost like the bosses left on some classical Greek structures, to milk yet more decorative effect from the shadow of its profile thrown by the sun against the backing wall. The Crusader Castle over the north entrance employs the same techniques, whilst the keep is built of enormous conglomerate blocks in the lower courses, and all the blocks have all their sides drafted back in the hellenistic fashion – a most impressive effect. Similarly the south-facing glacis, looking onto the moat, employs at least 44 shafts in an area of 600 square feet, in a diapered effect that is repeated in the next curtain to the east, as well as in the north-east corner tower, which has an enormous block at its base. The builders have left several centimetres of each column protruding. These columns perhaps came from the large and impressive Roman nymphaeum immediately to the north, for which only the capitals and bases survive.


Outside the castle is a line of mediaeval walls, of spolia limestone blocks for the towers, and petit appareil for the curtains; the northernmost corner tower has carefully fitted spolia blocks, some very large (8 x 3 x 2 feet), and column shafts in groups of four. The harbour was flanked by twin towers, the northernmost one of which still stands to its full height, with columns in the foundations. Inside the harbour is the town wall, now only some ten feet in height, but in its existing 200 yards length are column-shafts, in either two or three rows, some 300 or more on just this one stretch. Any suggestion that these are not decorative is scotched by the fact that in many cases the limestone course blocks have themselves been cut back, a quarter of a circle each around a shaft, to allow the shaft to punctuate the coursing. This was necessary because of the enormous size of some of the columns: there are pink granite shafts some two feet in diameter. To the west of the fortress, the excavators have collected together some 200 and more column shafts, the ends of which are very worn, and presumably extracted from the now collapsed (sea) walls.



Nicaea and Nicomedia


Nicaea, once an important city,[3] now a backwater, is still surrounded by its two sets of walls, five kilometres in length, and punctured by imposing gates at the cardinal points (except toward the lake, where the gate has disappeared). It was during the classical and mediaeval periods an essential stage on the route from Constantinople into Bithynia[4], as well as sometimes the seat of the Emperor; hence the comparison with the capital is the more pressing. This is reinforced by Nicomedia, of which little now remains, but which had spolia walls similar to those of Nicaea. Nicomedia declined as Constantinople grew, only sixty miles away. An earthquake in 358 destroyed the city, which never recovered, the more so since Justinian abolished the postal service from Chalcedon to Dacibyza and had his courriers go across the sea of Marmara to Helenopolis and Nicaea. What little digging was done (and almost no publication) showed baths revetted with marble, a colonnaded square paved with marble, and colonnaded streets meeting at right angles, together with massive public buildings. Its walls were built by Diocletian, just like those of Nicaea in alternating brick bands and rubble, and with ashlar spoils from the Hellenistic walls. The Lascarids refaced some of the earlier walls and towers rather than, as at Nicaea, building their own outer enceinte. Foss quotes Odo of Deuil, on the Second Crusade that By its lofty ruins overgrown with thorns and brambles, Nicomedia first showed us its ancient glory and the inertia of the present rulers – presumably the ruins were impressive because his contemporaries only occupied acropoleis, as he points out. [5]


It is not by chance that, at Nicaea, it is the walls and towers between towers 69 and 73 that get the gorgeous marble revetement, in order to impress anyone approaching from the Istanbul road, whereas immediately east of Istanbul-Kapi the walls swing quite sharply away from the gate. If proof were needed of conscious beautification, there is also tower 94, erected on the south side of the southwest  sea  gate.  Schneider  dates these  particular beautifications around 727,[6] which might also be the date of the insertion of dark marble blocks as tie-bars, especially on the long eastern section. The first wall was erected in 268 by Claudius Gothicus, the second (separated from it by a fosse) by the Lascarids, who may not have been the first to restore the earlier wall. With the exception of the stretch facing the lake (west), the walls of Nicea, although they cannot compete with those of Constantinople for height or length, both sets, the lower outer walls and the higher main set - are marvellously complete, gates and all. Ogier de Busbecq found Nicaea a mournful place because of its relative depopulation, the more so because of the Turks digging spolia for use in Constantinople, and battering with their hammers a cuirassed statue they found.[7] Nor was such a jaundiced opinion solely that of Western, classically educated sophisticates, as the quotation from Jalal al-Din Rumi, the 13th century dervish, in the Mamaqib al-Arifim, at the beginning of this paper, makes clear.[8] This was a common theme of visitors to Turkey in later centuries.


Nicaea’s walls are well decorated with reliefs, with large heads, and also with column-shafts, both as horizontal wall-ties and as decoration: in the central one of the three square towers of the north wall, some 37 columns are used to decorate the upper storeys and, at the same time, to act as floor-joists. The East Gate has reliefs, still visible, which impressed Kinnear, as did the reliefs and heads on the North (Constantinople) Gate. The south gate, for Bursa, has marble blocks and an inscription [9]. The walls and towers on the north side are noticeably of creamy-white marble (in contrast with the much darker blocks used elsewhere: it is far from fanciful to perceive the desire to create an effect on the side of the city facing Constantinople, since we find exactly the same attention given to marble display in the more important parts of other citadels, such as Seljuk. Thus the antique monuments of Nicaea have been reused in the construction of the first set of mediaeval walls, in a direct echo of the marble prestige of the finest of Constantinople's set pieces, the Porta Aurea itself.



Seljuk and Ephesus


At Ephesus, as might be expected, the Byzantine walls took in much less ground than Lysimachus’ Hellenistic defences, and made great use of spolia from adjacent monuments, some of which might have been conveniently demolished by earthquake. At some later date, the (surely small) population moved about 2km to the north, to the settlement now called Seljuk. This was still strictly within the purlieu of Ephesus even if outside the walls, the most conspicuous monument being a Byzantine fortress containing the 7th century Basilica of St. John. The entrance to this fortress, perhaps of the mid-seventh century with a mid-eight century rebuild,[10] is liberally decorated with spolia, as are walls adjacent to the basilica with columnae caelatae from the archaic and late classical builds of the Temple of Diana, in what is arguably an evocation of the grandeur of the past, while sculptures from the same location have been found in the fortress walls' backfill[11]. On even higher ground is the citadel. At Ephesus and Seljuk, then, the newer settlement is built with spolia from the old. Fellows[12] notes that the town of Seljuk is entirely composed of materials from Ephesus, and these old castle and mosque walls have become in their turn our quarry for relics of antiquity. For Foss, the walls of Seljuk are seventh century, like those of Pergamum and Sardis, which he  ascribes to time of Constans II (641-668).


At Seljuk, the lie of the land dictates that spolia increase as we move down the long hill from the upper fortress toward the Gate of Persecutions – and with immense quantities of spolia on the eastern-facing (or road) side. Many monuments must have been demolished to provide such a quantity of blocks: the theatre, large as it undoubtedly was, would have been an insufficient quarry for even a fraction of the span from the north to the south gates - a distance of some three kilometres, with regular towers as well. A pride in the past and its productions is obvious, not only in the splendour of the lower citadel and its curtain wall to the upper citadel (supposedly built in the 7th to 8th centuries, against Arab incursions), but especially with the Gate of Persecutions to the lower citadel, so called from a mis-reading of its re-used antique bas-reliefs,[13] which were carted off for the 6th Earl of Bedford to Woburn, in 1819. This is built entirely with re-used creamy-white marble blocks, and decorative friezes were incorporated to beautify it. Chandler, who visited the site in 1764, remarks on the theatre or stadium seats buttressing the Gate,[14] and Pitton de Tournefort admires the Lower Citadel precisely because of the beauty of its spolia, including the reliefs. [15] That the intention is to impress is confirmed by viewing the inner skin of the wall, which is only rubble and brick.


The upper fortress, overlooking the site of the Temple of Artemis, is built largely with rubble and brick toward the west (toward the sea) and the north - except for the use of squared blocks in the 45-degree revetments between the towers, presumably to guard against mining). But a considerable amount of antique material is to be seen in the east gate, facing the road, the arch of which is supported on antique blocks, and all along the adjacent walls. The antique-block revetments continue round onto the south wall, facing Seljuk itself, which also incorporates some antique and Byzantine blocks - including part of a figural relief (perhaps from a sarcophagus), a triglyph, and a frieze. May we conclude from this disparity that, just as at Costantinople and Nicaea, the gleaming marble walls were to used as a distant advert to the traveller? The disparity between Upper and Lower Fortress certainly struck Charles Tompson[16], who found in the upper citadel several curious Fragments of antique Marble being carelessly intermix’d in the Walls amongst other less valuable Materials, but then towards the south, the Remains of another Citadel of greater Antiquity, the Works whereof were cover’d with the finest Marble – and he then admires the bas-reliefs in the Gate of Persecutions and ( like many travellers) mistakes the Isa Bey Mosque for the church of S. John.


If the upper fortress has sparse antiquities, this is not the case with the eastern curtain wall joining the upper to the lower fortress (the latter containing the Church of S. John): recently dug out cleared of debris, this contains thousands of antique blocks, several of them with inscriptions, some presumably from large public buildings, and stands to an average height of 3.5 metres. Lawrence[17] suggests this work might date to the 8th century. Again, no doubt some of the material came from monumental tomb structures from outside the walls of Ephesus - Ephesus now being conspicuously lacking in such tombs, whereas at other sites in Turkey (Hierapolis, Assos, Patara, Eleiussa Sebaste), they are plentiful.





Here the city walls are of various periods, topped off with Seljuk merlins and proud inscriptions. South of the Gate of Hadrian in Antalya, a splendid triumphal arch, are laid some fine large blocks; adjacent to the gate, however, the builders have thrown together large and small blocks and levelled them off more-or-less every six courses or so. The result is a mess, contrasting with the towers flanking the gate, which are presumably 2ndC BC (the date of the foundation of the city) and, like the surviving stretch of wall to the north of the gate, of impeccable courses of large blocks, as presumably representing the most important (landward) aspect  of the city. At the bottom of Kurtulus Sokak is a mediaeval rebuild, with column shafts, decorated corbel blocks framing an inscription, and the use of fine sheets of marble veneer for decorative effect – probably part of the Seljuk refurbishment. Whilst such shafts could have come from Antalya itself, it is likely that many came from nearby Perge: this still boasts a splendid colonnaded street, but with conspicuous gaps (and many more bases on site than columns to go with them); and the enormous south baths have few columns left, and only insignificant scraps of marble veneer. By contrast, most of the granite columns of the agora are still in place: are we therefore permitted to conclude that the (Seljuk?) robbers preferred marble, and left granite alone? But marble may have been in short supply even at Perge, witness the construction of the episcopal basilica there using granite - not marble - columns, which probably came 150m from the palaestra of the North Baths. This is odd, and matches the odd feature of the spectacular colonnaded street, namely that the western side is almost entirely marble, but the eastern side is granite: is this make-and-mend after earthquake damage?




[1] cf. H. KOHL et al, Baalbek, II, Berlin & Leipzig 1925, pp. 3-11 for an outline of Baalbec in the Middle Ages;

[2] P. DESCHAMPS, La Défense du Comte de Tripoli et de la Principauté  d’Antioche, (Les Chateaux des croisés en Terre Sainte, II), 2 vols, Paris 1973, p. 214;

[3] Background in A. BRYER Nicaea, Byzantine city, in History Today, XXI.1 (1971), pp. 22-31; for a study of the walls, see A. M. SCHNEIDER, Die Stadtmauer von Iznik (Nicaea) Istanbuler Forschungen 9, Berlin 1938; and also FOSS, Byzantine Fortifications, pp. 79ff.. For a the use of spolia in Islamic Iznik, cf. K. OTTO-DORN, Das islamische Iznik (Ist. Forsch. 13), Berlin 1941. An equally short account appears as The city walls of Nicaea, Antiquity 12, 1938, pp. 437-43, where he suggests (p.439) that the cladding of some towers and curtains with blocks of fine marble occurred after the Arab incursions and depradations of 727, probably also the ramparts were covered with marble and then crenellated - but he doesn't say why. A longer treatment is A. M. SCHNEIDER & W. KARNAPP, Die Stadtmauer von Iznik (Istanbuler Forschungen 9), Berlin 1938;

[4] J. LEFORT, Les communications entre Constantinople et la Bithynie, in MANGO & DAGRON editors, Constantinople, pp. 207-218, Cf. p. 217: Tous les chemins menaient donc à Nicée, qui est aujourd’hui bien à l’écart, mais qui etait au Ive siècle et resta pendant tout le moyen âge le principal noeud des communications avec l’Asie – and Justinian restored the palace at Nicaea, and a bath;

[5] C. FOSS, Nicomedia and Constantinople, in MANGO & DAGRON, Constantinople, pp. 181-190;

[6] SCHNEIDER, Die Stadtmauer von Iznik, cit., plates 35-6, 42; and pp. 36-43;

[7] DE BUSBECQ, Travels, cit., p. 59.

[8] Quoted by VRYONIS, Nomadization, cit., p. 71;

[9] KINNEAR, Journey, cit., pp. 29-30: The outer port is apparently the work of a later age than the other, and consists of three blocks of white marble finely carved, which in all probability belonged to some temple or church, since the ground is strewn with similar materials;

[10] Arguments in FOSS, Byzantine fortifications, cit., p. 132;

[11] M. BUYUKKOLANCI, Fragmente der Bauplastik des Artemisions von Ephesos: Funde den Grabungen der Johanneskirche in Selcuk, Jahresheft des Oest. Archaeologische Instituts in Wien 62, 1993, pp. 95-104;

[12] FELLOWS, Journal, cit., p.206;

[13] General account in MUELLER-WIENER, Mittelalterliche Befestigungen, cit., pp. 89ff;

[14] R. CHANDLER, Travels in Asia Minor, 1764-1765, London 1971, p. 76;

[15] Relation d'un voyage du Levant, 2 vols, Paris 1717; cf. vol II, p 513;

[16] C. TOMPSON, Travels through Turkey in Asia, the Holy Land, Arabia, Egypt, 3rd edition, 2 vols, London 1767, I.71;

[17] A. W. LAWRENCE, A skeletal history of Byzantine fortification, in Annual of the British School at Athens, 78 (1983), p. 203;