The city of Şanlıurfa, modern-day Urfa, is situated in the Taurus Mountains of southern Anatolia, where the east-west highway from Zeugma on the Euphrates to where the Tigris meets with the north-south route from Somaysāṭ.  Known also as Edessa in history, the city is also referred to as Admi in Assyrian cuneiform tablets from the 7th century BCE,  Ōrhāy in Syriac sources, and as ar-Ruhā in Arabic texts.  Its location is important as the city stood along what was once the trade route upon which silk and spices flowed from China and India through to Asia Minor.

Upper Mesopotamia and Syria in the early Christian period, showing Edessa within the Kingdom of Osroene

It was here, for a brief period, that the Kingdom of Osroene, a vassal state of the Parthian and Roman Empires, created their own original mosaics, some of which help us date them precisely because their makers tellingly embedded the year of their creation within their designs. These mosaics, which use the Seleucid calendar, were created from the end of the 2nd century through the middle of the 3rd century CE and have characteristics that make them identifiable to this specific region of modern-day Turkey.

Many of the mosaics which have been discovered in the area of Şanlıurfa originally lined the flooring of archaeological cave tombs in the city.  Some have been discovered during the building of current-day structures, as modern development encroached over the city's historic past.   Unfortunately, and despite their initial recording, a number of known Edessa mosaics were subsequently destroyed. Or, as is the case with this long particular tale, taken halfway around the globe illegally, far from the city and its ancient inhabitants, whose history illustrates once paid deep respect for their dead. 

For some of the lost mosaics, only photos, or drawings, and written academic references remain.  Others, like two fragments from the region's Tripod Mosaic, later surfaced on the Beirut antiquities market, taken there sometime after their initial looting.  

This is the story of just one of the lost Şanlıurfa mosaics. 

In the 1950s and 1960s Judah Benzion Segal, (otherwise known as J. B. Segal) visited and studied at Urfa. A widely respected scholar of Syriac and Aramaic languages, his explorations were funded by the University of London and the Pilgrim Trust Fund.  Before Segal began recording the epigraphic remains he found from ancient Edessa, only three of the four mosaic floors in Şanlıurfa had been documented.  One of those is the Aphtuha mosaic, which is now in the Istanbul Archeology Museums. Two of the others are missing.

In total Segal would go on to document a total of nine mosaic floors in the city's cave tombs dating from the first three centuries CE.  Several of these he visited during his second season in Urfa, in the summer of 1956, which he wrote about shortly thereafter.

The importance of these mosaics tells us much about the rich traditions of Edessa's westward-looking elite during the Osroene Kingdom (132 BCE - 242 CE).  They were a population whose funerary practices were sometimes known to intermingle with tales of Greco-Roman mythology, yet still proudly added touches of their local identity.  Strikingly, in paying homage to their loved ones, they laid their dead to rest with inscriptions in the local dialect, a clue from the past which helps us identify these works of art.

Orpheus Mosaic identified by Segal

One of the mosaics documented by Segal is pictured to the left, in a black and white photograph.  The face of the principal character has been torn away at some earlier point in history.  This mosaic depicts the Greek poet Orpheus, and while the entire mosaic is also known to have been lost, this antiquity is not the subject of this article and serves merely as our story's preamble.

By the 1970s the city of Şanlıurfa's mosaics had suffered greatly. Due in part to neglect, theft, or destruction, it was commonly stated that only a few of the ancient floor mosaics survived in their city of origin, including two preserved in private homes.  A third mosaic and fragments of a fourth had been relocated to Istanbul.  The bulk of the rest of the known mosaics had been destroyed or carted off.  This is the story of one of them.  Or to be clearer, this is as much of the story as I have been able to piece together from a variety of sources, and yet the saga leaves many unanswered questions.

We begin in 1998, when Turkish authorities have reported that they received an unexpected bit of intel into the looting of a previously unknown Odessa mosaic; a clue that wrote the first known passages of the plunder of "Ikinci Orpheus Mozaiği," or what some have called the "Second Orpheus Mosaic."

That year, a customer, perhaps one of the looters, dropped off film to be processed at a photo shop in the city of Şanlıurfa.  The patron then forgot to pick up the film/prints, or perhaps in hindsight, decided that it is wiser to abandon them.  One of the images captured by the eye of the camera detailed an Osroene Kingdom mosaic in situ, its decorative red ocher and black frieze bordering a central element that depicts a scene from the fateful love of Orpheus of Thrace.

The mosaic, like the earlier one documented by Segal, showed the son of Apollo, enchanting a group of ferocious animals with the magical sweetness of his lyre.  As the wild beasts gather around him mesmerized, his playing subdues their innate hostilities. In the upper left, near the head of our instrument strumming Greek hero, is a short, three-lined inscription.

The photo's content also gives a clue that something is afoot.  Resting atop the nearly-perfect mosaic is an incriminating can of Turkish adhesive, a thief's tool.  The metal canister is photographed sitting strategically over the single area of the mosaic's design where some of the tesserae are missing.   Perhaps the can was positioned there to allow the photographer to capture the remaining parts of the mosaic's design or after the looter(s) had set to work.

At some point in the process, the thieves would use the can of sticky fixative, to apply a thin coat over the top of the mosaic's tesserae.  Once applied, the next step would be to cover the entire artwork with cheesecloth or another thin fabric, to give the workmen a flexible, yet sturdy, adhesion to rip up the mosaic from its foundation.  After that, the mosaic would be rolled up and carried off, in the same way as a rolled-up Anatolian carpet.

Seeing the subject of the photo, and probably wanting no problems with the police, the shop owner turned the incriminating evidence to the local authorities.  At some point later, Turkish authorities would postulate that evidence pointed to the mosaic being taken from the Kalkan Neighborhood of Şanlıurfa.  Unfortunately, I have found no reference as to when, or how, this theory was arrived at.  Without further leads, the investigation apparently ground to a halt for a period of time.

The mosaic surfaces one year later, on the art market in the United States

On 9 December 1999, the mosaic appeared with Christie's Auction in New York, identified as Lot 388.  By then its beautiful bordering frieze had been cut away, leaving just the lyre-playing Greek bard surrounded by his animals.  In the span of just one year, the fixative used to uproot the mosaic from its findspot had been carefully eliminated and the mosaic polished up; made ready for purchase on the U.S. ancient art market. 

Christie's listed the mosaic in its 1999 antiquities sale catalog as a Roman Marble Mosaic, Eastern Mediterranean.  It dated the artwork as having been created in 204 A.D.  Not surprisingly for that period, the object's Lot description gave no provenance information whatsoever.


In addition to its striking lack of ownership history, Christie's gets an epic fail for Rand and Becker's translation of the mosaic's epigraphy, and for calculating its age.  At the time of the auction, in 1999, neither man listed as having contributed to the translation had completed their studies.  Michael Rand was studying Jewish liturgical piyyut composed mostly in Hebrew with some Aramaic and Adam Becker was working on a second MA, in Classics.  

The pair translated the small inscription in the field to the left of Orpheus incorrectly as:

"Son of Zagrios, the player, is playing."   

In actuality, we would learn later that this inscription gives the name of the creator and identifies his craft.  The lower inscription was likewise jumbled and incomplete, as was someone's interpretation of the Seleucid calendar.  This error put Christie's dating of the mosaic off by ten years.

Interestingly, the auction house's interpretation failed to document that the inscription was written in Estrangelo, the earliest form of Syriac, and the lingua franca of the peoples of ancient Edessa.  This alone would have given prospective buyers a stronger clue as to the mosaic's origins, something the translators themselves or the auction house either failed to comprehend, didn't value, or intentionally omitted when preparing the write-up for the ancient object.

Meanwhile, Turkish authorities state that they tried to block this sale at Christie's in 1999 on the grounds that the mosaic was illicit and had been removed illegally from the ancient city of Edessa, modern-day Şanlıurfa but I found no opensource evidence to back up this later press statement within searchable records in the public's domain.

In any case, with or without the objection of Turkey, the auction moved forward regardless, despite the auction house's faux pas in art history.  When the hammer dropped, the mosaic had sold for US$ 85,000. 

Given the title of "Orpheus Taming Wild Animals," the mosaic's acquirer was the Dallas Museum of Art. Listed in the DMA's February 2012 catalog with accession number DEACC.1999.305, it was purchased via gifting funds from of David T. Owsley and the Alconda-Owsley Foundation, along with two anonymous donors, in honor of Nancy B. Hamon.  Although Owsley presently lives in New York, he is a prominent collector of antiquities and long-time donor/patron of the DMA whose donations have on occasion been questioned.

From that point onward, the second Orpheus mosaic would remain quietly on public view without garnering much in the way of further attention until 2006, when its inscriptions are correctly translated and published by Professor John F. Healey of the University of Manchester.  It is in the Autumn of that year that the academic penned his article on "A New Syriac Mosaic Inscription" for the October issue of the Journal of Semitic Studies. 

In his article, Healey highlights the correct editio princeps of the inscription written in Estrangelo that Christie's translators had botched.  The first, to the left of Orpheus, gives the name of the creator and identifies his craft stating:

"Bargased, mosaic-maker, laid the mosaic"   

The second, lower inscription reads:

"In the month of Nisan in the year five hundred five I, Pāpā, son of Pāpā, made for myself this chamber of repose, for myself and for my children and for my heirs. Blessed be whoever sees and gives blessing." 

The Seleucid date of 505 translates to 194 CE.  The month of Nīsān translates to March/April.  This dated the mosaic a full ten years earlier than Christie's had recorded it, and more importantly, makes this artwork the earliest dated Syriac mosaic known.  While the article itself may pay-walled, limiting its access, the fully viewable abstract for Healy's paper clearly lists the location of the mosaic as the Dallas Museum of Art.  Given the artifact's newfound importance as the earliest known mosaic written in Estrangelo, it remains to be seen why Turkish authorities did not pick up on this evaluation or the mosaic's location, in order to begin direct overtures with the Dallas museum regarding the suspect object in their collection. Perhaps this was because the article was written in English, or perhaps it was the fact that texts on epigraphic remains were not on anyone's radar.  Whatever the case, this illustrates the disconnect that can occur between academia and law enforcement which when combined can be useful in identifying suspect objects in various known and published collections.
From 28 May-1 June 2007 Turkish archaeologist and Ahi Evran University lecturer Barış Salman attended a Research and Results meeting in Kocaeli, Turkey with his country's Culture and Tourism Ministry - General Directorate of Cultural Heritage and Museums.  There Salman gave a presentation on his findings regarding the life and death of the peoples from the Abgar Royal Period, as interpreted through their mosaics from the Region of Osroene (Şanlıurfa).

In 2008 Salman followed up that meeting with a Turkish-language paper summarizing his presentation at the meeting which at last garnered some attention in Turkey. In his paper, Salman documented that only a small portion of the discovered mosaics from Şanliurfa had been properly recorded or brought to the city's museum or Istanbul over the course of the last century.  The scholar also noted that despite some archival photos and drawings documenting some of the mosaics identified from the 1950s onward, a number of these ancient mosaics had been lost or damaged.  Salman's paper also went on to emphasize that some of Turkey's mosaics had been secreted out of the country and taken abroad.

Salman cited the presence of Abgar Royal Period mosaics at the Louvre Museum in Paris, at the Art Gallery of South Australia in Adelaide, Australia, a Japanese Museum, and in private collections in France.  He also specifically mentioned the "Orpheus Taming Wild Animals" on public view at the Dallas Museum of Art, and cited Professor Healey's 2006 translation of the mosaic.

In 2008 Salman also wrote a second paper,  this time in English, again explaining his concerns about his country's mosaic losses.

It wasn't until American art Historian Maxwell L. Anderson replaced Pitman as the museum's Eugene McDermott Director in early 2012 that anyone at the Texas museum showed much of an interest in exploring the provenance of suspect acquisitions within their collection.

Ushering in a new era, Anderson started his reign by scrupulously emailing General Pasquale Muggeo, then heading up the Carabinieri Command for the Protection of Cultural Heritage, and Maurizio Fiorill, the Avvocato Generale dello Stato for Italy on 27 January 2012 alerting them to the presence of three works sold to the museum in 1998 by antiquities dealer Edoardo Almagià.   Almagià was concerning as his name had already been attached to other tainted objects found in the collections of the Cleveland Museum of Art and Princeton University Art Museum.

The objects Anderson was concerned about were two Etruscan funerary shields from the 6th century BCE and a Volute krater by the Underworld Painter from the 4th century BCE which depicts the man-bull deity Acheloos.  U.S. Customs officials had raided Almagià's New York apartment, in 2006 confiscating photographs, documents, and archaeological material, and the director, perhaps to avoid the scandals that had already befallen other museums, was being proactive about contacting the Italian authorities regarding artifacts within the DMA collection which were proving to be concerning.

By March 2012 (the earliest date I have found via a Wayback Machine save), Dorothy King and the Lootbusters website had drawn attention, to the English-speaking public about the serious problem of mosaic thefts from Turkey, though not specifically the theft of this particular Orpheus mosaic.  King's blog and the Lootbusters' website both pointed to a page that the Turkish Ministry of Culture had published which documented a series of identified mosaics known to be missing from Şanlıurfa Province, including the Orpheus 1, without the face.  The original Turkish ministry link of these identified mosaics and their descriptions can be viewed here.

It is sometime in March that Dorothy King emailed Anderson voicing her concerns about the museum's Orpheus mosaic.*

If the Turkish authorities were, as was later stated, already working on the theft of the second Orpheus mosaic since 1998 when the looted photo was first turned in to the authorities, as well as in 2000 when the mosaic was auctioned at Christie's then why wasn't this mosaic included in this list of missing pieces when Turkey published its detailed alert?

In March of 2012, after Kings contact, Anderson wrote to the Embassy of Turkey stating he did so after having viewed the Turkish Ministry of Culture website with its series of mosaics known to have been illegally removed from Edessa after the 1950s. One of the objects in that list, the Orpheus Mosaic missing his face, which was discussed earlier, raised his concern. 

Speaking of the DMA's acquisition of its own Orpheus mosaic, in relation to the Republic of Turkey ministry web page, Anderson told the press:

"The mosaics included examples that were virtually identical to this one... and with the same Seriac inscription, specifically identifying the site from which the mosaic was obtained. And therefore, it raised questions for me."

Stylistically and iconographically, the DMA's Orpheus is similar to other Edessa mosaics. Specifically, the inscription is similar both in style and content to other Edessa mosaics as is the Estrangelo script which originates in Edessa. Other features typical of the area include the absence of depth to mosaics from that region's design, the light colors used, and the expression and facial features of the subjects. The date of the DMA's mosaic inscription also coincides precisely to the period of mosaic production in Edessa.

In explaining the chain of events with the Turkish authorities, Anderson would tellingly state in the New York Times:

"For whatever reason, they hadn’t found their way to the Christie’s catalog or to us."

Anderson's statement and the timing in which he gave them, again contradicts various Turkish press statements that the authorities in Turkey were aware of the auction at Christie's in 1999 or had paid heed to any of the academic articles written that identified where the mosaic ultimately landed.

By 7 June 2012 outrage over the thefts of Turkish mosaics had started to seriously simmer within the general public in Turkey.  It is around this point in time that Aktüel Archeology Magazine initiated a citizens' campaign to return the Edessa Mosaic located in the Dallas Museum of Art's collection.  In a series of articles and linking blog posts, readers are encouraged to complete an English-language petition to be addressed and sent to the Dallas museum administrators, likely not aware that a discussion with the Turkish authorities had been opened by the museum in the Spring.

Following Anderson's March embassy inquiry the Turkish authorities pulled all the pieces of the puzzle together and an investigation was (re)launched by the Şanlıurfa Chief Public Prosecutor's Office reviewing all of the collected evidence of the theft.  At this point, the previous evidence in the archives of the Attorney General's office, including the 1998 photograph taken by the smugglers before the mosaic's removal, became critical to the success of their investigation and ultimately to the objects forthcoming restitution.

The authorities in Turkey would then forward the museum director their copy of the incriminating photo and six experts, one of which was Barış Salman, weighed-in giving their expert opinions that the mosaic could only have come from within the territory of Turkey.  Faced with overwhelming and irrefutable proof that the mosaic had been looted, negotiations quickly got underway to deaccession the mosaic from the Dallas Museum of Art and to develop a memorandum of understanding covering bilateral cultural cooperation.  As part of this process, the agreement was carefully worded to say that the DMA had acquired the mosaic in good faith in accordance with the laws of the United States of America, and without knowledge of any issue as to its ownership or title.

On 3 December 2012, within this framework of mutual goodwill, the Director-General of Cultural Heritage and Museums for the Republic of Turkey O. Murat Suslu and DMA Director Anderson formally signed their memorandum of understanding in Dallas in the presence of Cemalettin Aydin, the consul general of Turkey in Houston.  With this MOU Turkey took legal possession of the mosaic, while also agreeing to future loan arrangements between the Texas museum and the Republic of Turkey.

On 6 December 2012 the mosaic was flown home to Turkey on Turkish Airlines and was soon exhibited within the Istanbul Archeology Museums Directorate. Later it was transferred, first to the Sanliurfa Museum, and in 2015, to its final home at the Haleplibahçe Mosaic Museum.

"I, Gayyu, daughter of Baršuma, made this grave for myself. Whoever comes here shall not remove my bones. Whoever does it shall not have the other world (afterlife) and Maralahe curses him"

-----------------
*01 May 2020 - This article has been revised for clarity and to include a notation posted by Dorothy King on 30 April 2020 that she had contacted the Dallas Museum of Art Director Anderson in 2012 about concerns regarding the Orpheus mosaic in the museum's collection.
-----------------

Colledge 1994: Colledge, M., "Some Remarks on the Edessa Funerary Mosaics",
La Mosaïque Gréco-Romaine IV, Paris, 1994. s. 189-197. Association Internationale pour l’Étude de la Mosaïque Antique.

Drijvers, H. J. W., Old Syriac (Edessean) Inscriptions, Leiden,1972.

Drijvers, H.J.W and Healey, J.F., The Old Syriac Inscriptions of Edessa and Osrhoene, Leiden-Boston-Köln, 1999.

Güler, S. and Çelik, B., "Edessa Mozaikleri, Şanlıurfa Uygarlığın Doğduğu Şehir", A.C. Kürkçüoğlu, M. Akalın, S. S. Kürkçüoğlu,  E. Güler (ed.) 2002, 182–189.

Healey, J. F., "A New Syriac Mosaic Inscription", Journal of Semitic Studies 51-2, 2006, s. 313-327.

Salman, B., "2005 Yılı Şanlıurfa ve Adana Müzeleri Mozaik Çalışması", 24. Araştırma Sonuçları Toplantısı, C: 1, Çanakkale, 2006, s. 513-532.

Segal, J. B., "New Mosaics from Edessa", Archaeology 12.3, 1959, s. 150-157.

Segal J. B., "New Syriac Inscriptions from Edessa", Bulletin of the School of Oriental and African Studies, 22.1, 1959, s. 23-40.

Segal, J. B., Edessa (Urfa) Kutsal Şehir (Çev. A. Arslan), İstanbul, 2002.