The End of the Temples: An Archaeological Problem

AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem BRYAN WARD-PERKINS In this article I wish to explore a simple question – “Why does archaeological evidence not play a more central role in the increasingly sophisticated and nuanced literature on the end of Roman paganism?”1 Historians are well aware of the danger that our Christian texts may over-dramatize the end of the traditional cults, with accounts of saints like Martin in Gaul, John Chrysostom in Syria, and Shenoute in Egypt rampaging through cities and countryside forcibly destroying temples and smashing idols.2 They are also aware that the legislation in the Theodosian Code, ordering the closure and sometimes the demolition of the temples, is both contradictory and impossible to judge as to its effectiveness. There are even occasions where a textual reference to the ‘destruction’ of a temple can be tested against the material record, and be found to be greatly exaggerated. Justinian, for instance, is said by Procopius to have “pulled down the temples” at Philae; but the surviving remains of the buildings prove beyond doubt that their structures were hardly touched.3 Why then does the archaeological evidence, which is both ideologically neutral (unlike any of our texts) and scattered evenly across the empire, not play a greater part in assessing the reliability of texts, and the extent to which their messages can be generalized?4 There should, in theory, be plenty of archaeological data for the problem we are investigating, by contrast, for instance, with the earlier pagan assault on Christianity. This attack had primarily been aimed against people rather than buildings, and such assaults are extremely difficult to document archaeologically – Christian martyrdom, for instance, only shows up clearly in the material record because Christians, after the triumph of the Church, were able to build great basilicas over _____________ 1 2 3 4 I am very grateful for help in the preparation of this article from Jitse Dijkstra and Georgios Deligiannakis. I would also like to thank Johannes Hahn and his colleagues for inviting me to a colloquium in Münster that was both highly instructive and very enjoyable. For instance, CASEAU 2004, 120: “All these stories are to be taken with extreme caution, however. The word destruction might cover mere symbolic gestures such as knocking over a statue from its pedestal, as well as the complete dismantling of the building.” Prok. bell. Pers. 1.19.37 (trans. DEWING 1914, 188–189); DIJKSTRA 2004, 152–154 (with reference to the earlier literature). See, for example, the writings of Béatrice Caseau, who is very knowledgeable about the available archaeological evidence and uses it intelligently, but who none the less relies much more heavily on texts: CASEAU 2001 and 2004. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR 188 Bryan Ward-Perkins the graves of their martyrs. Pre-Constantinian Christianity was a religion of rooms adapted for worship within ordinary houses, like those excavated at Dura Europos, and of tiny monuments, like the first shrine of St. Peter on the Vatican Hill.5 These humble structures are found archaeologically only in exceptional circumstances – the frescoes of the house-church at Dura Europos were preserved because most of the house was sealed under a defensive earthwork, and the shrine on the Vatican would never have been identified as an important Christian memorial, if Constantine had not placed his high altar directly over it. Unsurprisingly, given the humble nature of early Christian monuments, the pagan persecutions were aimed much less at these material objects than they were at Christians themselves. Paganism, on the other hand, had enjoyed centuries – in Egypt, indeed, millennia – of prosperity and of establishment as the official religion of the state. As a result, the Roman empire was filled with an extraordinary wealth of temples, and of statues and reliefs of the gods. Furthermore these statues were often of stone, which is extremely durable, and the temples (except in the most remote northern provinces) were always of solid stone or brick, and sometimes on a massive scale (as at Baalbek or at Egyptian Thebes). The fate of some religions is difficult to trace archaeologically – pre-Constantinian Christianity is one (except from graves with inscriptions); Germanic paganism is another. Egyptian and Graeco-Roman paganisms, however, are almost perfect for the archaeologist – their temples and statues are amongst the most impressive and solid remains of their age. The Christian challenge to paganism was very explicitly a challenge to these imposing pagan buildings and sacred images, rather than a persecution of individuals, and the triumph of Christianity was marked by the destruction or removal of cult objects, above all statues, and by the destruction or, at least, closure of the temples. These are changes that should, in theory, show up in the archaeological record. Archaeology can only tell us indirectly about the history of living people, but it is very good at telling us the history of things, and archaeologists are highly fortunate when those things are as solid and difficult to miss as the remains of ancient paganism. Why then is the archaeological contribution to our knowledge of the end of Roman paganism so patchy and so marginal to the story as it is currently told? Of course, there are some pagan objects, even from the Roman period, that do not show up in the archaeological record. For instance, the chopping down of a sacred grove can never be detected in the surviving material record – unless ancillary buildings and altars leave signs of its fate. Similarly, cult objects in wood very seldom survive, and it is hard to imagine their removal or destruction leaving any tangible trace. The destruction of images in precious metal is equally invisible, since their remains will have ended up in the melting-pot. When Constantine _____________ 5 KRAELING 1967; TOYNBEE – WARD PERKINS 1956, 135–194. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem 189 stripped pagan temples of their gold and silver cult statues, Eusebius tells us that the metal was then purified “by smelting with fire” – in other words, it was melted down and added to the imperial coffers.6 This action probably did leave a significant material trace, in the issuing of a new and abundant gold coinage under Constantine; but sadly we will never be able to trace the ounces of gold in these solidi back to particular desecrated pagan shrines. However, the pagan temples leave impressive archaeological remains, and pagan statues of marble or stone, whether shattered or intact, should also survive well in the soil.7 Why, then, have we learned comparatively little from the archaeological record about what happened to these in the face of a triumphant Christianity? In part, we must certainly blame the priorities of archaeologists themselves. As is well known, until recently (particularly in the Mediterranean region), there was simply no interest in documenting how classical buildings and classical objects fell out of use. ‘Excavations’ were really massive clearance exercises down to the first impressive classical remains, and even quite substantial things that stood in the way could be swept aside without adequate record. The Palatine in Rome, for instance, was cleared down to imperial levels in the late nineteenth and earlier twentieth centuries. Somewhere on this hill was the important early-medieval oratory, and later monastery, of S. Cesario in Palatio; but, besides two fragmentary frescoes and a few scattered architectural elements, nothing was found of it, and its precise location remains in doubt.8 Ecclesiastical buildings in Rome, even from the early Middle Ages, were always solid structures, and the monastery must also have had its own graveyard – that S. Cesario and its monks disappeared without trace under the pick-axes of the excavators of the Palatine tells us a great deal about pre-modern excavation methods and priorities. Unfortunately, huge numbers of temples, probably by far the majority of those known today, were uncovered early in the history of archaeology – because they were prime targets for archaeologists interested in fine sculpture and impressive architecture. At Luna in Italy, for instance, both of the temples so far identified were the subject of nineteenth-century excavation, and were also the initial focus of attention when the systematic uncovering of the city began in the 1970s, before the archaeologists working there had mastered the techniques of stratigraphic excavation. Effectively, the remains of these two buildings were simply cleared of the overburden of soil.9 If there ever were traces in the archaeological record of when and how these temples closed, they have gone for ever; the only chance now of shedding any light on the chronology and the process of abandonment would be to find and identify architectural elements from them, reused in later datable buildings. _____________ 6 7 8 9 Eus. v.C. 3.54. Though there is, of course, the certainty that many a marble statue was burnt for lime. AUGENTI 1996, 50–55 and 64. FROVA 1973, cols. 2–19. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR 190 Bryan Ward-Perkins Sadly, a similar story can be told of hundreds of temples, including some about whose fate we would very much like to know more. When John Wood began excavations at Ephesus in 1869, on behalf of the British Museum, he started, not unnaturally, on the site which he (correctly) identified as that of the great temple of Diana of the Ephesians. What Wood found was that the temple had been very effectively demolished – remarkably little of it remained in situ and the haul of sculpture to take back to his pay-masters was extremely disappointing. He also identified the foundations of what he believed to be a church built over the site.10 We would dearly love to know whether this temple – to the many-breasted goddess who earned the opprobrium of St Paul – was the victim of early and systematic Christian demolition, and perhaps of the construction of a triumphalist church on the site. But, because of the nature of Wood’s clearance, all such evidence has gone, and it is not even certain that his ‘church’ was really a Christian building.11 Neglect by archaeologists of what we might term ‘abandonment history’, has certainly also played a major part in depriving us of information about the end of pagan statuary. In a recent book, Eberhard Sauer, has drawn attention to this, outlining the evidence for deliberate attacks on some pagan sculpture, and pointing out correctly that archaeologists have traditionally been very much keener to stick arms and heads back on statues, than to worry about how they ever became detached.12 Although a very great deal of evidence will have been lost, for instance to the restorers of the eighteenth century, Sauer is also certainly correct that it is not too late to submit some long-excavated finds and standing monuments to this kind of analysis – though, of course, it will often be impossible to tell whether a headless Venus lost her head to a deliberate blow, or to accidental damage when subsidence or an earthquake toppled her from her plinth. We should, however, note that – in order to place the evidence for deliberate damage to statues in its true perspective – we would also need to investigate seriously the evidence that other pagan sculpture was never deliberately harmed, and consider the reasons behind the preservation within the cities of Late Antiquity of some undamaged pagan imagery.13 Again, as with the broken statues, archaeologists in the past have tended to fall under the thrall of the objects themselves, and have spent too little time wondering how and why they managed to survive. Sauer offers, as one example of the kind of research that could fruitfully be done, the temples of Egypt, many of which (but not all) were carefully worked over, at tremendous effort, to destroy elements of their sacred sculpture (fig. 1). _____________ 10 11 12 13 WOOD 1877; the evidence for the church is presented on the plan at p. 262. See FOSS 1979, 86–87. SAUER 2003, esp. pp. 17–18. As argued by LEPELLEY 1993, 13. Lepelley compares the saving of pagan statues from destruction in Late Antiquity, to the work of Alexandre Lenoir salvaging royal tombs and other monuments of French history during the Revolution – which is an interesting parallel and thoughtprovoking, but perhaps a little over-romantic. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem 191 The phenomenon is well known, but, as far as I am aware, has never been studied systematically. There appear to be characteristics common to all the sites where this defacing occurred – above all, an attack on the carved human figures, but no interest in the sacred hieroglyphic texts and only occasional attacks on the representations of animals (despite the fact that these are generally sacred, and often, indeed, gods in their non-human form). On the relief illustrated here, for instance, the sacred vulture is spared, but the enemies of Egypt (in the grasp of the pharaoh), who are entirely neutral in religious terms, have been carefully defaced. All of this points to a considerable degree of ignorance about the ancient religion when the work was carried out. On the other hand, there are also interesting differences between what happened at various temples. At Medinet Habu (the mortuary temple of Rameses III near Thebes) the defacing can perhaps be associated with the building of a Christian church within the inner courtyard of the temple, but elsewhere, as at Dendera, no church was ever established within the building. Yet at Dendera the work of defacing was even more systematic – often the whole figure was attacked, whereas at Medinet Habu effort was concentrated on the face and hands.14 If a lack of scholarly interest in how temples and cult statues were abandoned were the only cause of problems, we would at least have a clear path forwards, with the potential for greatly improving the archaeological data forthwith. Unfortunately, this is not the only difficulty – there are also substantial innate problems with the archaeological record. Archaeology is very good at illustrating broad trends, the end of paganism and the Christianization of the empire being no exception: in the archaeological record the temples are adapted to new uses or torn down for their building materials, and a rash of new churches appears. But, if we seek tight chronology and the human motivation behind this change, we will generally seek them in vain. For reasons that will become clear, there is little chance that archaeological evidence will ever produce a significantly refined chronology of when temples were closed; or provide conclusive evidence as to whether the process was essentially violent, or pacific. Although we would like to know whether paganism went out with a bang, or died with a whimper, I doubt whether even the best archaeological work will ever be able to tell us. All too often, even when perfectly excavated, the remains of a temple are not sufficiently well preserved to give us reliable evidence of its abandonment history. In the case of cities that survived into the sixth and seventh centuries, as most Roman cities did, the abandoned temples became a major source of cheap stone, and were, over time, systematically taken apart, down to the level of their foundations, or even below them. In the process, any trace of their immediate post-abandonment history will have been destroyed, and, with it, any accurate indication of when and how they were abandoned. For instance, in the 1980s, _____________ 14 SAUER (2003, figs. 47–54) illustrates several examples at Dendera. His fig. 50 (at p. 94) is particularly interesting: all the human figures are defaced, but the god Horus, in the form of a falcon, is left intact, down to the detail of his prominent erect penis. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR 192 Bryan Ward-Perkins careful excavation revealed the location and plan of the Capitolium of Verona, facing out over the forum of the city.15 A statue-base datable to 379/82 happens to survive from this same city, which records the re-erection of a statue “in the crowded forum, after it had lain for a long time in the Capitolium”, which suggests that the temple was already in decay by the end of the 370s.16 It would be very nice if the archaeological evidence could confirm or deny this impression. However, what were found in the 1980s were badly robbed-out foundations that could give no hint as to the building’s earlier fate. Unfortunately, the chronological precision that we need, in order to chart the demise of paganism, is generally too tight for even well-preserved archaeological evidence to be of a great deal of use – it makes a considerable difference to our full understanding of the phenomenon, whether a temple was abandoned in 350, in 400, or in 450. If closed in 350, the temple had gone even before the emperor Julian’s ‘pagan revival’; if in 400, it may have been forcibly closed as a result of Theodosius I’s anti-pagan legislation; but if in 450, it may have been quietly left to decay, because no-one any longer cared for it. Even the best excavation can only very rarely offer such chronological precision. Charting archaeologically the abandonment of a building’s original function requires clear and datable traces of the change of use to survive in the soil, whereas all too often the evidence is either absent or ambiguous. Temples, in common with all classical public buildings, had solid stone and marble floors and might well have been kept clean, even if their religious use had ended and they had been adapted to a new function. Only when the roof fell in, or an accumulation of rubbish was allowed to build up – or, better still, when another solid building was built over the site – will any firm indication of abandonment have accrued. In researching this paper, I had a look at a number of Oxford churches that have recently been converted to secular use (two have become the libraries of colleges, and one a café) – in all of these cases the structural alterations have been minimal, and it would be very hard, or impossible, to detect the change of use archaeologically, in any normal circumstances. At Aphrodisias in Caria, the site of the great temple of Aphrodite was converted during Late Antiquity into the cathedral of the city, dedicated to the Mother of God. The initial clearance of the site by archaeologists provided no date for this transformation; however, more recent (and much more patient) work has suggested, on the basis of pottery evidence, a date in the second half of the fifth century for the construction of the church.17 This is a major step forward in our understanding of the abandonment history of the temple. But it cannot tell us whether the bishop of Aphrodisias took over a semi-ruinous building, a structure that was already being used for some other function, or a temple which was in good condition and still revered by many within the city. _____________ 15 CAVALIERI MANASSE 1987, 15–17. 16 CIL V 3332. 17 SMITH – RATTÉ 1995, 44–46. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem 193 Our best hope for archaeological precision, both as regards chronology and motivation, is undoubtedly violent destruction, particularly by fire, since this can leave material evidence that is both unequivocal and datable. Such evidence should exist, since several texts from the East record the burning of temples (whose wooden roofs made them vulnerable to arson).18 Most famously, this was how the great temple of Zeus Marnas at Gaza was eventually destroyed, according to the account given in the Life of Bishop Porphyry.19 There are indeed a few archaeologically attested cases of temples destroyed by fire, at roughly the right date for us to attribute the destruction, with reasonable confidence, to Christians. For instance, the floor of a rural sanctuary of Minerva at Breno (in Italy’s Val Camonica), excavated in 1986, was found to be covered in a layer of charcoal and ash (datable to after the fourth century), which also contained a marble statue of the goddess, missing its head and arms.20 Other complicated scenarios could be devised to explain this damage, but the simplest explanation is that this was deliberate destruction by Christians. Similarly, at Cyrene (in modern Libya) the vast marble cult statue of the temple of Zeus was broken into small pieces, and there are marks of burning on the limestone walls of the cella. Here there is no evidence to date the damage, since the temple was crudely excavated in 1861 and 1926. As a result, the case is not as strong as that at Breno – but Christian activity is, none the less, the most likely explanation.21 What is surprising, however, given the textual references, is that there are not more such cases. It is, indeed, tempting to deduce from this that temples were very seldom burnt down. However, although I suspect this is true – particularly in towns, where it can never have been safe to start major fires – I do not think it can yet be demonstrated archaeologically with any confidence. Too many temples were, after their closure, robbed down to their floor-level of their stone-work, for us to say with any confidence how they were initially destroyed – whether through neglect, or through violent damage, even if the latter was by fire. In the case of Gaza’s Marneion, for instance, our text tells us that, subsequent to the fire, the remains of the temple were demolished and a church built on the site.22 Activity on this scale is quite likely to have removed all archaeological trace of the original fire, however dramatic and destructive. It may be only at rural shrines, such as that at Breno, which were then entirely abandoned, that we will find good, undisturbed evidence of the initial cause of abandonment. _____________ 18 The evidence is presented by CASEAU 2001, 93. 19 Marc. Diac. v. Porph. 69. For the text and its context: HAHN 2004, 202ff. 20 CANTINO WATAGHIN 1997, 133–135, describing also a similar find at Industria. As far as I am aware, the Breno site has never been fully published, but a good illustrated interim report by ROSSI (1986) appeared in Soprintendenza Archeologica della Lombardia, Notiziario 1986, 65– 67. 21 GOODCHILD – REYNOLDS 1958, 39–40. 22 Marc. Diac. v. Porph. 75–79. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR 194 Bryan Ward-Perkins As Richard Bayliss has recently pointed out, the archaeological evidence for Christian damage to temples is all too seldom clear-cut, and too often open to wishful thinking – as I also discovered when I tracked a number of supposed cases back to their original publications.23 Most disappointing was the evidence from the Mithraeum under the church of Santa Prisca in Rome, whose excavators believed they had found clear signs of its violent destruction by Christians, and which is cited by Sauer as a particularly good example of passionate religious iconoclasm.24 The cult-niche, with its stucco figures, was certainly badly damaged when discovered, and bits of the relief were found scattered around the room; many of the frescoes too were badly damaged. But a careful examination of the published photographs of the latter did not suggest to me that they had been savagely and systematically attacked with axes, as their excavators claimed; rather, the plaster looks to have been in a generally very poor state when uncovered, and to have decayed randomly across the wall. Even some frescoed heads, which should have been the first target of iconoclasts, were well preserved when excavated, including the haloed head of Mithras himself (which, we are told in the published report, was destroyed, not by fourth-century Christians, but by a botched attempt at restoration in 1953).25 As for the stucco figures in the niche – stucco is a fragile medium, and, while they might have been deliberately damaged, it also seems possible that they had decayed and fallen apart. The head of Mithras, although detached from its original setting, was found in very good condition – a Christian iconoclast could easily have crushed it under foot. A rather different example – this time where the problem of interpretation is primarily due to a lack of firm dating evidence – is the Gallo-Roman sanctuary and villa excavated at Saint-Aubin-sur-Mer on the coast of Normandy. Here the limestone statue of a seated Celtic goddess, broken into pieces, was discovered _____________ 23 BAYLISS 2004, 23–25. Bayliss discusses the thorny question of whether one can tell Christian destruction from that wrought by earthquakes, foreign invaders, or neglect. For earthquakes and temples, see ROTHAUS 2000, 39–44 and 60–61 – but see also the review (and subsequent debate) in the on-line Bryn Mawr Classical Review 2002.02.25. The case of Corinth is interesting: the excavations were good by earlier twentieth-century Mediterranean standards; but, even in this case, we cannot confidently say what brought the pagan buildings down. 24 VERMASEREN – VAN ESSEN 1965, 129–131, 151 and 241f. SAUER 2003, 134–136. 25 See VERMASEREN – VAN ESSEN 1965, plates 53–65. The excavators too were puzzled at the survival of some of the sacred representations, though this did not deflect them from their overall interpretation: “After the destruction of the cult-niche and some of the monuments (though not others, the real meaning of which was not understood), the whole building was filled with rubbish” (VERMASEREN – VAN ESSEN 1965, 241–242). SAUER (2003, 135–136) argues – at some length – that it was precisely because of their ‘passionate hate’ that the Christian iconoclasts wielded their axes so inaccurately! The head of Mithras, which, according to Sauer, was a particular focus of attack, is at the right-hand end of his fig. 63 (on p. 135). VERMASEREN and VAN ESSEN (1965, 150) recount its true (and more banal) fate: “Of the head [of Mithras], only the outline has been preserved, with part of the halo; the rest of the face was lifted off in 1953 by the Istituto Centrale del Restauro, but did not respond to treatment and is now destroyed.” AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem 195 dumped into a half-filled well (fig. 2).26 Unfortunately, this find was made more than sixty years ago in very difficult circumstances – during a rescue-excavation in 1942–1944, necessitated by the building of the Wehrmacht’s Atlantic Wall. Violent Christian destruction (perhaps in the late fourth century) is an entirely plausible explanation for the find; but there are some problems with this interpretation. The statue was certainly not destroyed at the same time as the temple, which, oddly enough, had been demolished long before the well in which the goddess was found had even been dug. Furthermore, no datable finds from the fill of the well were recorded, so when exactly it was filled, and the statue dumped, cannot now be known for certain. The event may have occurred as the result of violent Christian destruction; but it is just possible that it happened much earlier, or much later, long after paganism was a contested issue. The statue may have been discarded, less as an act of hot-blooded iconoclasm, than as part of a process of ‘tidying up’ by local inhabitants, now somewhat embarrassed by their outmoded pagan statuary. Should we imagine a snatch-squad of monks, descending on the site and smashing the statue to pieces, before hurling it into an abandoned well, or should we envisage a Gallo-Roman owner, quietly ordering a superfluous object to be broken up and dumped into a convenient hole? It can certainly be difficult to tell, from archaeological evidence alone, quite how hot-blooded an act of destruction was. In the case of buildings, our sources tell us that some temples were demolished by zealous Christians, in order to end their pagan use – but, archaeologically, it would be very difficult to identify destruction of this kind, from the ‘calm’ demolition of an already abandoned structure, in order to recycle its stone.27 In both cases, the building materials will have been carted off for reuse. In the destruction of some sculpture, there does seem to be evidence of gratuitous violence, suggesting that this was fuelled by religious fervour. Sauer illustrates and discusses two marble statues, of Serapis and Cybele, from Sarsina in Italy, which were broken into hundreds of pieces (over three hundred in the case of the Serapis).28 It is hard to imagine how this could have happened, unless religious enthusiasm powered the sledge hammer. On the other hand, even systematic destruction need not always be attributed to hatred or enthusiasm. At some point in Late Antiquity, probably in the fifth century, the shrine of the imperial cult at Aphrodisias, the Sebasteion, was transformed into an area of shops. When this happened, the sculpture that decorated the monument, though much of it was set high above the ground, was systematically worked over. The figures of emperors, and of personified provinces, were left untouched, except for the removal of the genitals on nude figures; but representations of the pagan gods and god_____________ 26 EBLÉ 1948, 365–383. The statue was published by BÉQUIGNON 1949, 83–97. 27 The recorded cases of systematic destruction are collected by CASEAU 2001, 92. See also BAYLISS 2004, 22f. 28 SAUER 2003, 139–142. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR 196 Bryan Ward-Perkins desses were carefully chiselled out.29 In a case like this, should we imagine an angry bishop ordering the violent destruction of pagan images, or the civic authorities calmly and systematically up-dating a monument for new times? As so often, both are serious possibilities, leaving it open to the present-day observer to read into the evidence the scenario they prefer. Such are the problems with the archaeological evidence; but such also are the fascinating possibilities that it raises. Bibliography Augenti, A., Il Palatino nel Medioevo: Archeologia e topografia (secoli VI–XII), Rome 1996. Bayliss, R., Provincial Cilicia and the Archaeology of Temple Conversion, Oxford 2004. Béquignon, Y., Statue gauloise découverte à Saint-Aubin-sur-Mer (Calvados), MMAI 43 (1949), 83–97. Cantino Wataghin, G., La conversion de l’espace: quelques remarques sur l’établissement materiel chrétien aux IVe–Ve siècles, d’après l’example de l’Italie du Nord, in: Rouche, M. (ed.), Clovis: histoire et mémoire. Le baptême de Clovis, l’événement, Paris 1997, 127–139. Caseau, B., ȇȡȝıȞı‫ה‬ȟȝտȚȡ‫ה‬ȣ. La désacralisation des espaces et des objets religieux païens durant l’Antiquité tardive, in: M. Kaplan (ed.), Le sacré et son inscription dans l’espace à Byzance et en Occident (Byzantina Sorbonensia 18), Paris 2001, 61–123. Caseau, B., The Fate of Rural Temples in Late Antiquity and the Christianisation of the Countryside, in: Bowden, W. – Lavan, L. – Machado, C. (eds.), Recent Research on the Late Antique Countryside (Late Antique Archaeology 2), Leiden – Boston 2004, 105–144. Cavalieri Manasse, G. (ed.), Il Veneto nell’età romana, vol. 2: Note di urbanistica e di archeologia del territorio, Verona 1987. Dewing, H.B., Procopius, Works, vol. 1 (Loeb Classical Library), London – Cambridge, MA 1914. Dijkstra, J.H.F., A Cult of Isis after Justinian? ZPE 146 (2004), 137–154. Eblé, E., Découvertes à Saint-Aubin-sur-Mer, Gallia 6 (1948), 365–383. Foss, C., Ephesus after Antiquity: A Late Antique, Byzantine and Turkish City, Cambridge 1979. Frova, A. (ed.), Scavi di Luni. Relazione preliminare delle campagne di scavo 1970–1971, Rome 1973. Goodchild, R.G. – Reynolds, J.M., The Temple of Zeus at Cyrene, PBSR 13 (1958), 30– 62. _____________ 29 SMITH 1987, 97–98. While the iconoclasts failed to identify a few of the obscurer gods (e.g. Hygieia), and hence spared them, they were certainly better-informed than the people who chiselled out the reliefs at Medinet Habu and Dendera: all the emperors and provinces were spared, and almost all the gods were destroyed. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem 197 Hahn, J., Gewalt und religiöser Konflikt. Studien zu den Auseinandersetzungen zwischen Christen, Heiden und Juden im Osten des Römischen Reiches (von Konstantin bis Theodosius II.) (Klio-Beih. n. F. 8), Berlin 2004. Kraeling, C.H., The Christian Building (= vol. 8, part 2 of The Excavations at DuraEuropos, Final Report), New Haven 1967. Lepelley, C., Le musée des statues divines. La volonté de sauvegarder le patrimoine artistique païen à l’époque théodosienne, CArch 41 (1993), 5–15. Rossi, F., [Interim report on excavations at Breno], Notiziario de Soprintendenza Archeologica della Lombardia 1986, 65–67. Rothaus, R.M., Corinth, the First City of Greece: An Urban History of Late Antique Cult and Religion (Religions in the Graeco-Roman World 139), Leiden 2000. Sauer, E., The Archaeology of Religious Hatred in the Roman and Early Medieval World, Stroud 2003. Smith, R.R.R., The Imperial Reliefs from the Sebasteion at Aphrodisias, JRS 77 (1987), 88–138. Smith, R.R.R. – Ratté, C., Archaeological Research at Aphrodisias in Caria 1993, AJA 99 (1995), 33–58. Toynbee, J. – Ward Perkins, J., The Shrine of St. Peter and the Vatican Excavations, London – New York – Toronto 1956. Vermaseren, M.J. – van Essen, C.C., The Excavations in the Mithraeum of the Church of Santa Prisca in Rome, Leiden 1965. Wood, J.T., Discoveries at Ephesus, Including the Site and Remains of the Great Temple of Diana, London 1877. AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR 198 Bryan Ward-Perkins Fig. 1: Medinet Habu, mortuary temple of Rameses III: defaced relief AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR The End of the Temples: An Archaeological Problem 199 Fig.2: Saint-Aubin-sur-Mer: broken statue of Celtic goddess dumped into a well AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR AUTHOR’S COPY | AUTORENEXEMPLAR