The Ruination of Dura-Europos

Ann Laura Stoler has asked what happens when we shift our scrutiny from the picturesque ruins of empire to the ongoing process of ruination. Examining some of the many and continuing ruinations of the site Dura-Europos on the Syrian Euphrates, this paper compares ancient and modern destructions by the Roman army, Sasanian forces, Yale archaeologists and ‘Islamic State’, and questions the different way these are understood. Through these stories of ruination I interrogate the relationship between archaeologists and processes of site destruction, examining archaeologists not only as those who evaluate loss but as agents of destruction themselves.


Introduction
The history of the site now known as Dura-Europos is known almost entirely from evidence found on the site itself, from the excavations conducted in the 1920s and 1930s, first under Franz Cumont, of the French Academy, and then under the joint scientific direction of Cumont and Michael Rostovtzeff of Yale University (Cumont 1926;Baur and Rostovtzeff 1929). On the edge of the Syrian steppe overlooking the Euphrates valley below and Mesopotamia beyond (Figure 1), traces of human habitation nearby are dated as early as is never only about one type of power, and economic and social imperialism are also evident at Dura in a way that might contradict the strong language of the inscription. For instance, Roman coinage in the numismatic evidence from Dura shows increasing numbers of Roman denarii around this time, from a range of mints, testifying to further connections between Dura and the Roman sphere in the early second century. Indeed, it was Roman and not Parthian coinage which was the normal currency of the site throughout the Arsacid period. Still, as our models increasingly turn to nuanced and complex theories of interaction, we perhaps need this simple reminder that we should not sanitise what could be and often was a relationship of force and domination, a relationship that ruined temples and carved records of its victories into the landscape.

The Ruination of the Roman garrison
'… brave in campaigns, mighty in wars, dead…' (Dura inscription no. 939, Rostovtzeff et al. 1944) These words described Julius Terentius, tribune of the twentieth Palmyrene cohort, on his epitaph (Welles 1941;Nilsson 1942;Rostovtzeff et al. 1944: 176-185, no. 939). We know Julius better than we know most of the Roman soldiers who were stationed at Dura, as he is recorded in papyri from the military archives as well as from a painting, depicting him with his men (Figure 2). His Greek funerary inscription was found in a house near the centre of the city, apparently incomplete-as can be seen, the last section was not carved, but the painted guiding lines are preserved (Figure 3). Whether his wife, Aurelia Arria, or the person she  commissioned to create this did not themselves survive to complete the memorial, we do not know. While it is incomplete, this is a fascinating document: not only because a Roman tribune is being commemorated in a Greek inscription by his wife, but because it is virtually alone at Dura, where we have no tradition of funerary commemoration: no funerary inscriptions, and unlike its more famous Syrian neighbour Palmyra, no funerary portraits. By the third century AD, a Roman military garrison had been installed within the city walls-on its northern side this included the building of a Roman palace, the transformation of houses into blocks of barracks, and the building of an amphitheatre and bath buildings (James 2019). This involved a massive urban transformation and huge building works. For a time, it seems some of the existing urban population held on. We have, for example, a document which details the divorce of a Roman soldier from a local woman (P. Dura 32), and while the military controlled the entire site, patrolling its walls and gates, there was some nominal separation of the communities, with the building of a wall separating the military quarter from the rest of the site. This would not only have displaced many people from their homes, but also placed considerable pressure on resources, as it would have included not only the 1500-2000 soldiers we know were present, but also the larger military community of dependants, servants, veteran households, and others perhaps themselves totalling 3000-6000 people (James 2018: 44-45). Even if we take the "high" estimate of the population of Dura, at 10-15 thousand people, this was a Roman military base which is almost half the size of the existing population of the city was inserted within its walls. In an era where we are ourselves witnessing the mass movements of displaced people, we need to consider more carefully what this would have meant at a lived, human, scale.
It is even more incredible then, in the face of mass upheaval and disruption that there is much evidence of shared practices evident at Dura when spaces are occupied by the Roman army. In the temple of Bel, a painting of Conon and his family (Figure 4), which is probably Arsacid in date (Andrade 2013: 235-236), was joined in the third century by one of our same Julius Terentius, together with his men (Figure 2). These paintings were perhaps made more than a century apart. But they share many things-not least their painted medium and their placement within the sanctuary, in which they coexisted at the time the city was ultimately abandoned, but also aspects of their style and the form of religious dedication, making a sacrifice at an incense altar, and the votive form of the painting itself. Julius Terentius and his men may be fully kitted out as Roman soldiers, right down to the signet ring on Terentius' pinky finger, but they are shown in a way that any Durene would have recognised as indicative not only of shared representations but shared practices of making and of ritual. Certainly, the painting of Terentius and his men could be used to demonstrate imperial connectivity in a traditional sense: here are some Roman soldiers in their gear, with their standards, looking just as we might expect. But if we tell our histories of Roman rule with a broader view, with one that reaches beyond that empire in time and space, we could also see that very same painting as demonstrating religious and artistic practice which was deeply intertwined with local habitus, and local ways of living. The army is demonstrably far from disconnected from local practices in other ways as well. Lucinda Dirven and Matthew McCarty (2014;2020) have recently shown that even the Mithraeum of Dura, which shares so much with Mithraea across the Roman empire, including its darkened character, its benches, and its tauroctony reliefs, also has elements of local practice, including graffiti in the most sacred part of the structure, and a relief which echoes the local practice of dedications by groups of people: in this case, the place of the family group who are normally depicted at Dura is taken by people who are united by their military and Mithraic affiliations. Simon James' work on the military base similarly shows that the Roman military, when they had taken over the north side of the site, probably still opened the gates and allowed access to temples, such as that of Azzanathkona, to continue for civilians, and thus deliberately allowed for the continuation of existing religious practices in parts of the site that had otherwise been seized, wholesale (James 2019).
In the final years of the site, perhaps as late as the early 250s, the situation changed again. Artefact assemblages suggest that much of the civilian population had fled and that many buildings throughout the site, even those which were not architecturally modified into military buildings, were occupied by the Roman military and the accompanying community (Baird 2012c). The threat by the Sasanians was keenly felt by the Roman military, and what had been a walled city became a fortress-literally, in that it became a defensive place. A huge embankment was built against the interior of the city walls to hold back the Sasanian incursion (Figure 5), deliberately and with great effort involving the methodical destruction of buildings and the moving of many metric tonnes of earth, ruining a huge swathe of the site (Rostovtzeff 1934: 10, 312;Rostovtzeff et al. 1936: 179-180, 309). 2 It is this rampart along the interior of the western wall of the site that preserved so much of what we know of it, including many of the parchments and papyri, which survived because of the dry environment created by the deep strata. It also preserved most of the paintings that the site is best known for, in buildings along the western edge of the site. Temples were apparently deliberately desanctified, and a number of them including the mithraeum, the synagogue, the temple of Aphlad, and the Christian building, were all partially destroyed and sealed beneath the rampart (Coqueugniot 2012a). As part of this process, cult reliefs like that of Aphlad (Figure 6) were put out of use by turning them around to face the wall before being sealed beneath the earth. This careful closing of the sanctuary may have been meant as a temporary measure. Aphlad's worshippers, though, would never return to their andron in which this cult relief had been dedicated in the first century. The relief remained there until it was excavated in 1931, by which time of course it was the French rather than the Romans who held sway over this territory (Rostovtzeff 1934: 103-104). The tale of Dura is usually told as one of 'Persian' destruction of a Roman city, but by the time it was taken by Shapur's armies the Roman army had long since ended its urban life.

The Sasanian ruination
'…we approached the deserted city of Dura, situated on the river bank…' (Ammianus Marcellinus 24.1.5) The embankment built by the Romans had ultimately been futile. In the fourth century, when Ammianus Marcellinus accompanied the emperor Julian on his travels, a toponym still survived: Dura. While the name of the site had held on, the only occupants by this time were reportedly the herds of deer hunted and feasted upon by the imperial party. Dura's last gasp, as a military site, was in the 250s. The traditional date of AD 256 for the fall of Dura was based on a single coin found on the person of a soldier who was buried hastily in the siege rampart (Bellinger 1949: 181). 3 While this is circumstantial, there is no evidence for Roman occupation after this date. From the preserved siege works, mines and countermines, as well as evidence within the city, we know some of the specifics of Dura's final, gory, days, including the use of what amounted to poisonous gas (James 2011). By this time the site was no longer a functioning urban environment but entirely a Roman fortress at war. Temples, like the small temple of Bel in block M5, were used for storing ammunition, and the part of the temple of Aphlad that was not now under the Roman rampart was used for munitions storage as well (Rostovtzeff 1934: 101;Leriche 1997: 94;Coqueugniot 2012b: 64). The Sasanians took the city before that ammunition was exhausted. Bodies were left where they had fallen, still wearing their armour, or were hastily covered in abandoned buildings (Baird 2012a: 314). Catapult bolts lay lined up in the street ready to be fired but were never used (James 2007: 41). Among the last of the battles, around AD 256, the Sasanians built a huge siege ramp to breach the city walls, just west of the temple of Aphlad on the exterior of the southwest corner of the site. The interior rampart that had been built to hold back a Sasanian incursion failed, and the Romans, not for the first time but for the last, lost control of Dura. There is some evidence of a Sasanian presence at Dura, potentially immediately after the sack of the city, but this is limited to a handful of coins, and some burials (Bellinger 1949: 10;Saliou and Dandrau 1997: 95). Many buildings apparently stood empty for a time before they collapsed; a house in a part of the site near the river that was excavated by the Franco-Syrian expedition in the 2000s had its floors covered in disarticulated rodent skeletons, apparently owl pellets which covered the floor when birds inhabited the emptied homes (Baird 2012a: 318-319). The thousands of people who had occupied this cosmopolitan city were gone. The coin hoards some of them had probably hoped to retrieve remained in the earth until the picks of the Syrian labourers uncovered them for their Yale archaeologist bosses (Baird 2012a: 316).
All of this, to me, begs the question: at what point does the human devastation of empires become neutralised? Archaeologically, we tend to speak of ' destruction levels' whether these are caused by earthquakes or by armies. W.G. Sebald has written of the way that destructions of the past become lesser as human tragedies, become neutralised, over time. 4 Archaeology has a chance to tell a different story, though, not only one of time capsules from the past, but of the material consequences of empire as it was experienced on a human level: in houses emptied of their most precious things, of bodies still wearing their armour left to rot, of temple gods carefully stored away, of fortunes in coins never retrieved. González-Ruibal (2008: 260) has written of the responsibility of archaeologist to bear witness, and to bring the presence of this evidence to the fore, or in his words, 'put the corpses on the table'. Stoler similarly invites us to reconsider ruins of empire as the human impact of the material consequences of empire, but in Roman archaeology, we are more likely to be telling a story of aqueducts and trade, of 'what the Romans did for us'.

Archaeological ruination
' …a site desolate and forlorn…' (Hopkins 1979: iii) Clark Hopkins (1979: iii), the field director at Dura in the 1930s, opened his book The Discovery of Dura-Europos with an epigram: 'After the siege and victory by the [Sasanian] Persians in AD 256, the record is blank. The mute testimony that remained was of a site desolate and forlorn, where the lonely and level sands covered the bones of the city and stretched away across the desert'.
The imbuing of archaeological interpretations with the historical contexts of their production has increasingly been recognised over recent decades, particularly in relationship to colonial and post-colonial circumstances. 5 Put another way: everything that is known, or is thought to be known, of the archaeological ancient past is as much about archaeology's own past. In the quote above, Hopkins plays on several tropes of archaeology, but perhaps most of all invokes the sense that the ancient past was alone and forgotten, as if its archaeology was lying in wait, waiting for him to discover it: the past of Dura, for Hopkins, was in need of saving.
Of course, this description is in many ways a fiction; the record was not blank (we have textual evidence of a Byzantine hermit at the site), nor was it desolate, as by the Ottoman period a regularly-travelled road ran through the site. It was not even a desert: to one side is the Syrian steppe, and the other the lush agricultural expanse of Mesopotamia. Hopkins, though, was never one to let reality impinge on a romantic Orientalist fantasy. The entanglements of the discipline of archaeology in the destruction of ancient sites are generally little remarked upon in the Roman East, compared to archaeology's self-identification as protector of knowledge through recording and stewardship or its redemptive role through reconstruction. 6 And yet, archaeology has had a persistent role in the creation of regimes of value around particular pasts and sites which is usually ignored as a contributing factor to site destruction. I want to argue here that archaeology could itself be considered a process of ruination.
Archaeology has of course been considered as a form of destruction, but from an epistemological standpoint not a value-laden one: archaeology is a type of anthropology where we kill our informants, as the famous anecdote goes (spoken by a fictionalised archaeologist in Flannery 1982: 275). This alludes to the idea that archaeology is unrepeatable, but I wonder whether we as a discipline have spent enough time questioning whether those informants were ours to kill? Asked differently: whose memories are valued in archaeology, and who is permitted to record or efface those memories? Ruins such as those at Dura-Europos are not found: they are made. As Stoler notes, large-scale ruin-makings are typically state projects, and can involve forced removal of populations and new zones of uninhabitable space, reassigning inhabitable space: a description which we could apply both to the Roman treatment of Dura with the forcible removal of civilians from the garrison, and the archaeological treatment of that same place. 7 While Dura was apparently uninhabited when it was ' discovered', we see the forcible removal of local populations at contemporary archaeological sites in the 1920s and 1930s in Syria, such as, infamously, from the temple of Bel in Palmyra, where the Arab villagers were removed from their homes, (on the expulsion of local peoples from the Temple of Bel in 1929, Anon 1930). Unsurprisingly, these relocated people, perhaps not unlike those moved from the northern side of Dura by the Roman garrison, were not always satisfied with their new dwellings which did not accommodate their ways of life. This displacement at Palmyra was but one of several examples of the reshaping of physical space under Mandatory powers in Syria into ordered colonial spaces: at Palmyra was explicitly undertaken so that archaeological excavations could occur (Neep 2012: 4, 143). Archaeology was both a motivation and tool for imperial powers, and is not easily extricated from the colonial project. It gave a justification for colonial violence including the clearance of inhabitants from particular sites, and for the very presence of imperial powers by providing precedent for European rule of the region. This was also one reason for the particular (but not exclusive) focus on Classical sites.
Just as Hopkins hoped to save Dura from the desolate sands, the West has long claimed itself as the benevolent saviour of antiquities (Figure 7). 8 We see this explicitly here in the Illustrated London News in the 1920s: the British and French Mandates in Palestine and Syria explicitly presumed responsibility for 'saving' and restoring these 'neglected' sites. Archaeological sites, particularly classical ones, became a point of rupture, claimed by the state (first the mandatory government, later the independent Syrian one) to the detriment of local communities. The formal categories of empires have allowed archaeology to become part of the project of the nation state, and indeed as Christopher Jones has recently demonstrated, the nation of Syria has used the ancient past as a means of justification, in the absence of other available discourses-one of the ways this happened was in the linking of Assad family to Zenobia in official Syrian media, enacted for example by juxtaposing Roman monuments with modern building projects on Syrian banknotes (Jones 2018: 40-42). Other empires got their say, too: in the archaeological project at Dura-Europos, the excavation was financed by Rockefeller for the explicit purpose of promoting American cultural hegemony in the Near East (Baird 2018: 4, 9, 12-16). While the famed painted religious buildings of Dura-its synagogue, mithraeum, and Christian building-had been accidentally preserved by the Roman rampart, they were completely erased from the site by archaeologists. The collection of the paintings for museums-the synagogue paintings to Damascus, the Christian and Mithraic paintings to Connecticut-necessitated the wholesale removal of the structures, with their mudbrick walls destroyed in order to remove and preserve the plaster on which the paintings were made. The sculptors of the Mithraic reliefs could not have imagined that one day they would wend their way across oceans. The reliefs at the focus of the mithraeum at Dura were still in situ when this photograph was taken (Figure 8), but their removal, together with that of the paintings, had begun: the braces visible at the top of the photograph, on the arched, painted, plaster were the first stage in taking it away. Eventually, all the mudbrick would be carefully removed from behind the plaster, and the mithraeum boxed up and delivered to New Haven by sea. The erasure of the building was so complete that its place on the site could, by the 2000s, only be located with reference to the relative position of the city walls (Figure 9). At Dura, the  sites of these buildings became empty, except where the new Franco-Syrian expedition re-built some walls of the Christian building and Bel so there would be something for the tourists to see. 9 It is not a part of Western cultural memory that items such as these paintings are displaced-instead, having been saved, they are remade as world heritage so that it becomes ' ours' (Meskell 2018). Local people were made to dig it up, destroy it, pack up the crates and send them away (Figure 10). Those people themselves paid with their bodies and sometimes their lives, and at Dura we know of a number of deaths directly related to the expedition, when people were killed while excavating a city tower which collapsed on them. These occluded histories are not about what is forgotten or neglected, but things that the very formation of the discipline means that we don't see. 10 Just as the Romans transformed the city with the garrison and rampart, the Yale teams moved earth on an industrial scale using the industrial methods of mining and backbreaking labour-methods that are usually blurred, in the background (Figure 11). We as archaeologists are trained  not to see the people in these pictures as humans but as scales (Baird 2011) (Figure 12). We are trained to accept that it was the Yale excavators who ' discovered' the archaeology, when it was never lost to local knowledge, and those American hands were never themselves dirtied with the earth that had covered the objects (Figure 13).

The ruination of the Syrian conflict
'…about 400 groups working there with the aim of finding coins…' (Quote of informant in Brodie and Sabrine 2018: 77) Of course, the Sasanians and the archaeologists have not been the most recent to plunder Dura, have not been the most recent ruin-makers. The site has now been destroyed again by extensive looting since the start of the Syrian conflict, with satellite images showing its walled urban area and surrounding necropolis transformed into a landscape pockmarked with thousands of holes (Figure 14) ( ). Yet, the destruction of Dura by looting was unlikely to have been by Daesh and those commissioned by it alone: local people in difficult circumstances, Assad regime forces, and others probably participated, and the networks used to sell the looted materials also likely pre-dated Daesh's existence. 12 Dura's past (extent of excavation visible in Figure 17) had become an economic resource to be mined from the ground, a source of funding for terrorist groups thanks to the continued existence of international markets in antiquities. 13 Estimates based on computer modelling of the market data on antiquities from Dura have given a mean value of the objects dug up at over eighteen million US dollars (Greenland et al. 2019: 35). The looting of Classical-era sites in Syria seems to have been a deliberate targeting due to the monetary value of Classical objects, 14 which of course relies on their coveted status in Western museum collections. Put another way: Classical archaeology has invented such objects as valuable in their own right and in this way might be seen as complicit in their looting. Archaeology has long been complicit with such work in other ways: the very same archaeologists who excavated Dura in the 1930s were openly buying objects while in Syria to supplement their extraction of objects from the site. Buying objects, at the time, was perfectly compatible with excavating them, and in some ways these were simply two different means to achieve an end-filling the art museum and collections of the Classics Department (Baird 2012b: 38 on correspondence from the Yale University Art Gallery Dura archive which refers extensively to the purchase of lamps and other items while expedition members were travelling in the Middle East). The destruction of the site continues: some of the most recent images I have seen were in a video sent to me by a friend in Syria in late 2017, recording the aerial bombardment of the site museum, which probably occured years before. The museum, a small mudbrick structure rebuilt on ancient footings which housed reproductions and models, had taken on a new life by this time, it seems, as a munitions store (Figure 15), as is evident from the second and tertiary explosions after it was bombed from the air. 15 The bombing seems not to have targeted heritage per se, but incidentally, because Daesh stored their weapons in our sacred places (as museums are often thought of), just as the Romans had transformed some of Dura's temples into storehouses for Roman artillery. Dura-Europos has also turned up as a site named in American Department of Justice Warrants where particular artefacts, described in a press release as being jewellery 'similar to that of Dura-Europos' were named in documents seeking forfeiture of property in 2017 (Figure 16). 16 These are items believed to have already been confiscated by Turkish authorities, but the US sought their forfeiture not because they have been trafficked, but as assets of a terrorist organization (in this case, assets related to a particular individual, Abu Sayyaf, who had engaged US forces). Estimated cash values for each of the objects is given in American dollars: 30 and 50 thousand dollars for these two objects respectively. And yet the very same document quotes the US Attorney for the District of Columbia saying that the warrant 'reflects our determination to locate precious stolen antiquities and preserve the cultural heritage of ancient sites that fell under ISIS's control.' Where once Dura's objects were the target of collection for American universities and museums as cultural objects, now they are the target of the American government, both as monetary devices and symbols of American power over ISIS. The rhetoric and pretence are of the protection of cultural heritage, but that heritage has a price tag and can be seized as booty from Daesh just as the Romans seized the doors of Ammaios' temple.

Conclusions
One of the questions I want to provoke by examining processes of ruination is why certain modern destructions are seen as bad, whereas ancient destructions are seen as preserving archaeology. There is indeed an irony to the hand-wringing of Western archaeologists over contemporary site destruction when the richest archaeological environments are generated precisely by cataclysmic events in the past. It is interesting too, that in the modern satellite imagery which is used to determine the extent of damage to the site of Dura-Europos, many of the holes identified by scholars as ' decades-old looting holes' are actually old archaeological excavations (Casana 2015: Figure 2). In terms of what they physically do to the site, practices called looting and those of scientific excavation are distinguishable only by the words we use to describe them. As we see at Dura, the discovery and 'preservation' of archaeology (like the mithraeum paintings) has its symmetrical counterpart in the destruction by archaeologists. Other ' destructions' have been enabled by archaeologists but blamed on others. For example, when Breasted recorded the very first paintings found at Dura they had been left uncovered by the British military who had discovered them. The eyes and faces of some figures were supposedly scratched by local Bedouin. Breasted excoriates this as barbarism, as 'mutilations' by 'prowling Arabs' (Breasted 1924: 59, 75). 17 Yet, he too left the paintings he recorded exposed (including the painting of Conon and his family reproduced above) and they were then further damaged, by 'strolling Arabs practicing the barbarous iconoclasm which they regard as prescribed by the Koran' (Breasted 1943: 287). We might ask who the real barbarians were. In a similar way, the Syrian Directorate-General of Antiquities and Museums (DGAM), a part of the Assad regime, reports on the destruction of sites (AbdulKarim 2013; DGAM 2014), but presents this damage as something done by terrorists when there is extensive evidence of heritage destruction by the regime itself, and the regime has long been complicit in antiquities smuggling (Dagher 2019: 70, 141). The DGAM professes that Syria's heritage 'belongs to all Syrians regardless of their political views' by releasing statements, printed on letterhead featuring Syria's coat of arms complete with the national flag, concerning the status of antiquities but it places itself, and hence Assad, as the stewards and controllers of that heritage, leveraging it as yet another tool in their arsenal. 18 There is also an irony in that the specific targeting of classical sites for looting or even more spectacular destructions happen precisely because archaeology has been so very successful at turning its objects of study into objects of value (Baird and Kamash 2019). This value is one that, as we have seen, can have a very exact monetary price, but it is also one that sees worth in objects themselves, and not in the contexts from which they have come. What we consider plunder and what we consider collection are also value judgements to do with who holds authority and who, in the discourse of the discipline of archaeology, is permitted to interact with the past and on what terms. The transformation of artefacts into objects of Western value has in some cases made them targets for destruction in these competing discourses. As I have noted, Stoler (2008: 202) has argued that large-scale ruin-making is ' always a state project'. We perhaps need to see that this is equally as true of the Roman destruction of Dura-Europos, with their seizure of the temple doors, as it is of archaeology, with its seizure of the paintings in the mithraeum which was later built on the very same spot. As we construct narratives in our telling of archaeological interpretations or histories of archaeology, we need to consider where our empathy is situated, and why.
There is of course an argument, made for example by Cornelius Holtorf, that we should see all destructions simply as transformations, as the creation of new heritage. In this reading, the moonscape of Dura in the most recent satellite imagery is Syria's new heritage, a landscape of pockmarks where antiquities were mined from the earth, trafficked and sold. Holtorf (2015) has perhaps been the sole voice in questioning the 'loss aversion' of heritage professionals, and seeing destruction of sites such as Palmyra as being a transformation. Whether or not this recent destruction becomes heritagized (Munawar 2019), we should examine not only the destruction as part of the military-industrial complex (e.g. in the bombardment of the site) or capitalism and consumption, in the extraction of objects for sale (González-Ruibal 2008), but that the destruction and ruination as something done by archaeology as a colonial instrument, bringing with it colonial modes of forgetfulness or occlusion of certain pasts and certain presents.
Writing the history of archaeology and reflecting on archaeology's colonial underpinnings does not give us the moral high ground: archaeology's own past is still present in the very notion of archaeological expeditions, in the foreign schools, and our continuing work on the region (Stoler 2016: 25). We need to recognise the continued way this operates and the regimes of value it creates: we need to see that when the reopening of the Damascus Museum is celebrated (as many archaeologists did publicly on social media when it reopened in October 2018), we are not just celebrating access to cultural heritage, we are lending scholarly credence to the legitimation of the Assad regime through that heritage. 19 Ruination, as I hope I have shown, is a political project: it is something that happens to people, and is a relation of force. 20 We need to recognise the ruinations not only of sites but also of lives: archaeology, just like Roman expansion, was a form of colonial violence which took lives, used people's bodies, transformed landscapes, occupied cities, displaced people, captured objects, and selected which histories were worth telling. And yet, one of the things we can recognise in archaeology is that destruction does not mean things are gone. In studying the past, it is we who choose what to resurrect. This means that we have a responsibility to think about our place and motivations as (often) foreign archaeologists in the Middle East, to bear witness to destruction, to reflect on whose stories we tell, and how we tell them.

Notes
1 Securely dated parchments from civic archives which evidence Arsacid hegemony at the site between time of Trajanic incursion and the late-second century, e.g. P. Dura 22 (AD 133/134), P. Dura 23 (AD 134), P. Dura 24 (AD 150/60), etc. 2 The sequence of destruction included knocking down the easternmost part of each city block along the western edge of the site and backfilling the upstanding structures with mudbrick packing and rubble to create a rampart. Several phases of work seem to have occurred, and were recorded in the site notebooks and plans for some blocks (most clearly in blocks L7 and N8). 3 The date relies on the end of the coin series at Dura-it is worth noting that civic mints in Syria cease production around the same time, in AD 256-257: see Butcher (2012). 4 Archaeology tends to treat ' destruction layers' as just that: neutral and abstracted rather human-wrought; we might consider this as an erasure of human pain and suffering. Compare discussion of Sebald in (Remmler 2005: 59-60). 5 E.g. the critiques of Dunbabin on Greek colonies in the Western Mediterranean (De Angelis 1998). 6 Reconstruction might be literal, in the rebuilding of archaeological monuments, or more general in capacity-building projects such as those funded (e.g. in the UK) by the British Council's Cultural Protection Fund. 7 'Large-scale ruin making takes resources and planning that may involve forced removal of populations and new zones of uninhabitable space, reassigning inhabitable space, and dictating how people are supposed to live in them. As such, these ruin-making endeavours are typically state projects, ones that are often strategic, nation-building, and politically charged… Ruins draw on residual pasts to make claims on futures…' (Stoler 2008: 202). 8 Organisations such as UNESCO 'salvage' rescue operations 'that European empires claimed as their benevolent task' (Stoler 2008: 198). 9 On rebuilding at Dura, see e.g. the work of Bessac (2005), and on the initial motivations of the Franco-Syrian expedition for stabilisation and reconstruction (Leriche et al. 1986). 10 Occluded histories are not about what is forgotten or neglected, but 'part of what such geopolitical formations produce.' (Stoler 2016: 10). 11 For other examples of satellite imagery of looting at Dura, (Luo et al. 2018: 16-17;Luo et al. 2019: 9-10). Analysis of Dura destruction images also in Digital Globe imagery acquired by American Association for the Advancement of Science and published in 2014, https://www.aaas.org/resources/ancient-history-modern-destruction-assessing-status-syria-s-tentative-world-heritage-sites-7. 12 On the ' culprits' of looting being 'members of the local populace, opposition fighters and Syrian regime soldiers', (Al-Azm 2017: 94); The DGAM (Oct 31, 2016) blamed local peoples for the looting, even naming those from the village of Dweir specifically: http:// www.dgam.gov.sy/index.php?d=314&id=2093. On the range of actors in looting at Dura, see Hardy 2015a on the relationship between ISIS' approach to antiquities and that of the Assad regime, Al Mohamad Forthcoming. 13 On the continuity of the illegal trade throughout the past decades in the face of UNESCO and other efforts to stop this trade via policy (Brodie 2015). Coverage during the looting was provided by a variety of Facebook groups, including e.g. Le patrimoine archéologique syrien en danger. 14 Danti et al (2014), note this as a pattern: 'Recent looting at Jebel Khalid matches a pattern of targeting Seleucid, Roman, and Byzantine sites in regions controlled by SARG and Islamist extremist forces, especially Islamic State (viz. Apamea, Dura Europos).' 15 Pers comm. December 2017; video archived at EAMENA, http://eamena.arch.ox.ac.uk. Interpretation of secondary and tertiary explosions as evidence for munitions storage, Simon James, pers comm. The date of the video cannot be verified; satellite imagery indicates the footage might be of damage incurred in a previous year (the roofing, at least, seems to have disappeared from this building by 2014, as is visible on satellite photos and in DGAM photos of damage to the site, http://www.dgam.gov.sy/index. php?d=314&id=1550). ASOR also tracked the destruction of sites in Syria and Iraq, based on research conducted by the 'Safeguarding the Heritage of the Near East Initiative', funded by the US Department of State, and reporting with weekly and monthly reports: http://www.asor.org/chi/reports/weekly-monthly (Dura appears in these as a site of major looting first in Weekly report 4, by Michael Danti, issued on September 2, 2014) (on this project see also Danti 2015;Danti et al. 2017); as pointed out by ASOR, the material in the reports is useful for monitoring but comes from a range of sources and much is unverifiable. Reports at Dura (or likely incidents related to Dura) in Incident Reports SHI 14-0078; SHI 14-0098; SHI14-105; SHI 15-061; SHI 15-0161; SHI 17-0081. The Eamena (Endangered Archaeology in the Middle East and North Africa) project at Leicester, Durham and Oxford has also catalogued the destruction at Dura: https://database.eamena.org/. Others are collated in UNESCO archive reports in the Observatory of Syrian Cultural Heritage: https://en.unesco.org/syrian-observatory/. On the reasons for targeting cultural property during conflict, see e.g. Brosché et al. 2017; on the complications and potential need for looting by some, (Hardy 2015b). Videos likely to have been taken at Dura also appeared in 2015 Buzzfeed reports on antiquities looting (Giglio and al-Awad 2015). 16 On the sale of antiquities as a reappropriation of resources from the state by ISIS, see Jones (2018: 52). A small number of objects which likely came from Dura or nearby have been located, e.g. objects confiscated by authorities in Turkey, and in some cases the US has sought to seize these objects as assets of a terrorist organisation. Some were destined for sale in the US, and the attempt of the US to seize these objects engages with them as objects primarily of monetary, rather than cultural, value. E.g. The same language was used in the popular press when it was the paintings were publicised, reporting that the 'best painted' face had been scratched and defaced (e.g. AP, Laurence Daily Journal June 9).