How Much Was Taken? The Loss Figures for Cyprus and Where They Come From

Almost every article on looted Cypriot antiquities quotes a number, and almost all of them trace back to one sentence written in 1999. That sentence, a Cyprus Police estimate of "more than 60,000 ancient artefacts" removed since the Turkish invasion of 1974, has been reprinted for a quarter of a century, sometimes with its original wording and sometimes with church art folded into it. A smaller set of figures sits beside it: the number of churches affected, a count of Cypriot lots at two auction houses, percentages drawn from museum catalogues, police valuations of seized hoards, and the counted objects of individual seizures and returns. This entry takes each figure in turn: where it was first printed, what it counts, whether a method was ever published, and what it leaves out. Throughout, the figures for church art (icons, frescoes, mosaics and liturgical objects, almost all medieval or later) are kept apart from those for archaeology, because the published record keeps them apart at the start and the press often does not.

The figures at a glance

The table divides into two kinds of number. The counts at the bottom (lots, catalogue entries, objects handed back) can be checked against a list. The estimates at the top cannot, because no list or method has been published for them.

The 60,000 estimate and how its wording changed

The figure was first put into print by Sophocles Hadjisavvas of the Department of Antiquities, at the symposium on illicit antiquities held at the McDonald Institute in Cambridge in 1999 and published in 2001 [2]. The A. G. Leventis Foundation quoted the passage in full: "Tomb looting in Cyprus has a tradition of over 3000 years but since 1964 has intensified, particularly after the Turkish invasion of northern Cyprus in 1974. Since then it is estimated that more than 60,000 ancient artefacts have been illegally transferred to different countries of the world, together with perhaps another 16,000 icons and mosaics stolen from churches" [1, p. 29]. A Department text by Hadjisavvas distributed through the Press and Information Office gives the source and keeps the same structure: "It is estimated by the Cyprus Police that since the Turkish invasion of 1974 more that 60,000 ancient artefacts have been illegally transferred to different countries of the world", and "to this number one should add approximately 16,000 icons and mosaics stolen from at least 500 churches" [3].

Three points follow from the original wording. The 60,000 is a police estimate, and no method, inventory or earlier source for it has been published. It counts "ancient artefacts", and the church art is a second number added on top. It describes objects "transferred to different countries", which is a claim about removal, and says nothing about how many were sold or where.

The wording then drifted. In 2015 Hadjisavvas restated it as "Some 60,000 ancient and medieval artifacts were exported and sold on the international black market", citing his own 2001 page [4, p. 130]. That sentence adds medieval objects to the 60,000 and adds a sale on the black market, neither of which the 1999 version contains. Fehlmann, attacking the figure in 2014, rendered it as "60,000 classical and pre-classical artefacts" said to have left the north [5, p. 161]. Writers who quote the 60,000 today may therefore be quoting an antiquities figure, a figure for antiquities and church art together, or a figure for sales, depending on which restatement they read.

Church art and the count of churches

The church figures are a separate series and should be read as such. The 16,000 of 1999 counted "icons and mosaics stolen from churches" [1, p. 29]. In June 2024 Michalis Gavrielidis, head of the police office for antiquities trafficking, gave "over 16,000 Christian icons, mosaics and murals" taken since 1974, in the same report that announced the last return from the Dikmen estate [9]. Both are figures for church art alone, and the second gives no archaeological count beside it.

For buildings, Hadjisavvas's Department text speaks of icons stolen from "at least 500 churches" [3], and his 2015 paper says that "some 500 churches have been ill-treated, while others have been deliberately destroyed", citing Chotzakoglou's survey of 2008 [4, p. 130]. The figure counts buildings converted, damaged or demolished, not objects removed. A higher figure of 550 or more circulates in the press, but no source for it has been traced, and it is not used here. The losses of church art, and the cases of Kanakaria, Lysi and the Munich seizure of 1997, are the subject of looting in occupied northern Cyprus.

Fehlmann's auction count and his "few thousand"

The only published attempt to test the 60,000 against data is Marc Fehlmann's, written from Eastern Mediterranean University in occupied Famagusta. He is openly hostile to the Republic's figures, which he calls "Greek Cypriot propaganda" [5, p. 163], so his framing is his own; his counts are another matter. Students on the university's Cultural Heritage Management course of 2009 to 2010 "analysed all antiquities sales from Sotheby's and Christie's between 1988 and 2008" [5, p. 158]. Across those twenty years the two houses offered 52,482 antiquities. Of these, "only 1,842 items (4 per cent) could be identified as being of Cypriot origin", 377 had a history of ownership before 1970, and 15 could be traced to the Cesnola collection, de-accessioned by the Metropolitan Museum in the 1920s [5, p. 161]. A second test used three published private collections in Cyprus, the Severis, Giabra Pierides and Zintilis catalogues, about 1,500 objects in all: 3% had an ownership history before 1974, 2% a find spot in the north and 19% a find spot in the south [5, p. 163].

From this he concluded that 60,000 was "implausible", because so large a volume "would have caused the collapse of prices for Cypriot antiquities", and that "a more likely estimate of the numbers of antiquities that have left Northern Cyprus illegally since 1974 is a few thousand, although this does not minimise nor eliminate the problem of looting" [5, p. 161].

His counts are solid for what they cover. The "few thousand" is not derived from them. Two houses at the top of the market are the venues least likely to accept undocumented northern material, and loot would more probably pass through regional salerooms, dealers, private sales or collectors on the island, where Fehlmann himself supposed that "the majority of potential collectors of Cypriot antiquities are based" [5, p. 163]. His study shows that 60,000 Cypriot antiquities did not pass through Christie's and Sotheby's; it cannot show how many were removed. The price argument also assumes that the 60,000 went to the open market, which the 1999 wording does not say.

Hardy's 67% and what it measures

Samuel Hardy's study of the 1963 to 1974 conflict is often cited for a higher figure. He applied a method adapted from work on Cycladic figures: objects "without information (or innocuous explanations of the lack of information) are probably illicit antiquities" [6, p. 463]. For a peacetime baseline he used two Louvre catalogues of Cypriot antiquities, 319 objects in all, and found that "214 (67.08%) of the 319 objects were neither scientifically excavated nor found through survey or by accident. The majority of the public collection had probably illicit origins" [6, p. 466]. For four private collections formed during the conflict, 387 of 390 objects "appear to have been looted" [6, p. 466].

The 67% measures the share of a museum catalogue with no recorded excavation or chance find. It is a proportion of 319 objects in one museum's pre-conflict holdings, and it does not count anything removed after 1974. A missing find record means the digging was undocumented; it does not show that the object left Cyprus unlawfully. Hardy himself notes that the colonial system of partage ceded at least a third of an expedition's finds to the excavators [6, p. 466], and objects from divisions and licensed sales left the island lawfully whatever their find record. Hardy calls his own approach a "crude categorical method" [6, p. 459]. His results matter for this entry in a different way: they show that a large share of Cypriot antiquities in old public collections came out of the ground without record, long before 1974, which is one reason why a loss figure dated from the invasion cannot be checked against what is on the market.

Police valuations, seizures and counted returns

The figures that can be checked are small. In January 2010 police in the south broke up a ring working in the Limassol and Paphos districts. The Cyprus Mail reported nine people arrested and remanded, "over 100 artefacts, including statuettes and urns" found so far, and antiquities "worth about €11 million" [7]. A report in the Washington Examiner the same day gave ten Cypriots detained and the stash "estimated to fetch 11 million Euros, or $15.5 million" [8]. The Director of Antiquities said at the time that "from what we have seen rare antiquities are not included" [7]. Fehlmann called the €11 million "obviously nonsense" and argued that it would amount to "several truckloads" of Cypriot material [5, p. 158]. Spread over a hundred or so objects, the figure comes to around €100,000 each; in our own unpublished research on 2,168 Cypriot vessels sold at auction, the median price was about $820. A police valuation is an estimate made for a prosecution, and it is reported here as the police's.

Seizures in the occupied north in the early 1980s were counted piece by piece. The Turkish Cypriot weekly Olay reported in 1982 that 217 antiquities had been confiscated in 1975 from the shop of the mayor of Morphou, some marked with the initials and numbers of a registered Greek Cypriot collection, and that on 1 January 1982 "18 pieces of antiquity were confiscated while being smuggled out", sold for DM 1,000 to a German buyer [10]. Its other reported seizures from tourists were of similar size.

Returns are also counted. The final return under the Munich agreement with the estate of Aydın Dikmen, who was never convicted of any offence, was "24 ecclesiastical relics and 36 prehistoric and other antiquities", sixty objects in all [9]. That return is the clearest single proof that one network handled both trades: more than half of what it gave back was not church art. A return of 36 objects from Australia in 2023, Chalcolithic to Roman in date, was reported with no find spot and no mention of the north [14].

The lawful export beside the losses

The loss figures are usually printed without the export figures for the same island. Luigi Palma di Cesnola alone removed a collection "originally numbering over 35,000 items in all, some 6,000 of which remain" in the Metropolitan Museum [11, p. 116]; his methods and his last shipment were disputed even at the time, and the collection is a measure of scale rather than an example of good practice. The Swedish Cyprus Expedition of 1927 to 1931 took "two thirds of the Expedition's outstanding archaeological production, circa 12,000 of the 18,000 objects" to Stockholm under the division agreed with the colonial government [12, p. 36]. The Cyprus Museum itself sold antiquities for decades, as described in how the Cyprus Museum sold antiquities, and licensed dealers exported under export licences and labels until 1996. Between them, Cesnola and the Swedish expedition took close to 50,000 objects off the island before 1932. The routes are set out in the legal export of Cypriot antiquities. This is why Fehlmann's 377 lots with a pre-1970 history, and the 15 traced to Cesnola, are a small and visible part of a much larger body of lawfully exported material on the market.

Evidence that the figures undercount

The case against reading the small counts as the whole story is strong, and some of it comes from Fehlmann. On the page after his "few thousand" he wrote: "It is a fact that apparently well-informed gangs regularly plunder archaeological remains in the North of the island and discover ancient sites that are not known to the Department of Antiquities. It is also a fact that these gangs and their middlemen are rarely caught and that the damage they cause cannot be determined" [5, pp. 161-162]. Hadjisavvas reported that he "personally observed recent large-scale looting" in the necropolis of Agia Irene (Ayia Irini) in the north-west [4, p. 132]. Digging that is never discovered produces no seizure, no return and no auction lot, so every counted figure is a floor.

Hardy, writing of the 1963 to 1974 conflict, judged that the collections he studied were "only the traceable few from collections of thousands of antiquities on the island", and that "the extraction of tens of thousands (or hundreds of thousands) of antiquities, within the decade of civil war... must have involved large-scale participation" [6, p. 466]. Those years fall before 1974, but the point applies to every period: tomb robbing on an island of this density of cemeteries can remove more objects than any count records.

The dated record of antiquities incidents in the north clusters in 1975 to 1982, and few are reported for the 1990s. That gap may reflect an absence of reporting as much as an absence of digging, since the Department of Antiquities has had no access to sites in the occupied north since 1974. The low counts are evidence that bulk archaeological looting has left few traces in the major salerooms and the seizure record; they are not proof that it did not happen.

Reading a figure in the press today

A loss figure in a news report can be checked in four steps. The first is to find whose number it is: the 60,000 is a police estimate first printed in 1999, with no method published, and it should be quoted as such. The second is to see what it counts. The original separated "ancient artefacts" from "icons and mosaics", and a report that gives one number for both, or that sets the 16,000 church figure beside a return of prehistoric pottery, has merged two trades. Reuters in July 2024 listed "bronze age antiquities in Australia" among objects taken from the north after 1974 [13], although the report of that return gave them no northern find spot [14]. The third is whether the figure means removal, sale or value: a police valuation is not a price, and the 2010 valuation is about a hundred times the median auction price per object. The fourth is the export record for the same object type, because a Bronze Age jug of a type found at a northern cemetery may have left Cyprus through the division of 1930, the Cesnola sales or the Cyprus Museum's own saleroom decades before 1974.

What survives these checks is narrower than the headline and still serious. Church art was taken from the occupied north in the thousands, from about 500 churches. Antiquities were taken from registered collections, museum stores and unrecorded sites in the north, and in organised cases in the south, in numbers no one has counted. The best-documented counts run to hundreds of objects, and the only published estimates, 60,000 and "a few thousand", rest on no stated method. For a Cypriot antiquity offered for sale today, the question that matters is the documented history of that object, which the provenance and provenience of a piece can sometimes answer and an island-wide estimate cannot.