You are here:Home/Research/ Byzantine Studies/ Project Grants/ Archaeological Survey of Boğsak Island and Its Environs

Archaeological Survey of Boğsak Island and Its Environs

Günder Varinlioğlu, University of Pennsylvania, Project Grant 2009–2010


Boğsak Island in Isauria, a small islet (ca. 7 ha) lying one kilometer offshore, is remarkable for its sizable Christian settlement with complex infrastructure and architecture. While the rugged, mountainous terrain of Isauria (Roman Rough Cilicia) made land access and transportation extremely difficult, its equally rough shores were part of the Mediterranean maritime networks, especially during the prosperous and well-connected world of late antiquity. The Isaurian coastline has many anchorages, bays, and islands that were provisioned by coastal plains, the fertile Calycadnus delta, and the agriculturally marginal, yet successfully exploited mountainous hinterland. While other islands in this region (Kösrelik, Dana-Pityussa, Güvercin) were also occupied and/or inhabited intermittently in late antiquity, only at Boğsak did a dense and extensive settlement develop, despite the absence of safe natural harbors, fresh water, arable land, or useful natural resources that could support the population. Boğsak was part of a network of urban and rural settlements in the territory of Seleucia ad Calycadnum, the capital city of Isauria. It was situated ten kilometers (4 ½ nautical miles or about one hour sailing distance) southwest of the port of Holmoi, which serviced the capital and the pilgrimage center of St. Thekla at its outskirts. This small rocky outcrop, rising steeply about fifty meters above sea level (fig. 1), controlled the southern approach to a large bay protected from harsh northern winds.

Boğsak Island and Bay are attested in texts and maps only occasionally: Stadiasmus Maris Magni, from the second half of the third century CE, includes the earliest reference to the island, while the bay appears in the fourteenth to sixteenth-century portulans and maps. By the end of the fifteenth century, Portulan Rizo indicates that the formerly inhabited island offered neither anchorage, nor any other amenities. A Greek inscription found on the island by G. E. Bean and T. B. Mitford in the 1960s and read by G. Dagron and D. Feissel in 1987, gives crucial information about the settlement: Αὔξη Ἀστερέ[α] ὐκωδωμουμένη ὡς πώλης (Long live Asteria, built like a city). This short acclamation suggests that Asteria was the name of the settlement on the island and possibly also on land. This settlement did not have the administrative status of a polis, yet possessed urban aspirations.

Despite their good state of preservation, the archaeological remains on Boğsak Island have remained largely outside the scope of scholarly interest. In 2010, with the permit of the Ministry of Culture and Tourism in Turkey, I launched an archaeological and architectural survey of Boğsak Island and its environs. This project aims to understand the formation, expansion, and end of Isaurian rural settlements during the first millennium CE in topographically and agriculturally marginal landscapes such as Boğsak Island. Although this is a new survey project, it stems from my dissertation entitled “Rural Landscape and Built Environment at the End of Antiquity: Limestone Villages of Southeastern Isauria (University of Pennsylvania, 2008), in which I explored late antique villages in the mountainous hinterland of Seleucia ad Calycadnum. In the Eastern Mediterranean, the fourth through seventh centuries are characterized by the exploitation of, and settlement in, semi-arid rural territories. However, small islands with a tight symbiotic relationship with the land have rarely become the subject of scholarly investigations.

Fieldwork Campaign in 2010

The first fieldwork campaign of the Boğsak Archaeological Survey in 2010 focused on the preliminary investigation of architectural remains and the physical characteristics of the terrain, as well as the sampling of surface finds on Boğsak Island. Our small team consisted of three recent graduates from the Department of Archaeology at Mimar Sinan University, Istanbul, and two students of Geomatics Engineering at İstanbul Technical University, supervised by Dr. Caner Güney from the same institution.  The campaign’s objective was to investigate the character and distribution of archaeological and architectural remains, as well as the challenges presented by the terrain and vegetation, in order to develop a survey methodology suitable for this specific site. In other words, the first campaign was intended to serve mainly as a reconnaissance and preparation for future seasons.

At the beginning of the fieldwork, a low-resolution satellite image downloaded from Google Earth was the only available visual aid; plans, large-scale topographic maps, and aerial photographs of the island did not exist. On a very difficult terrain characterized by rugged rocky shores, steep slopes, and thick vegetation, and in the hot and humid July weather, a small but dedicated and hardworking team walked across the island and worked on the preliminary identification and documentation of the main neighborhoods, cemeteries, water and drainage systems, and churches. Thus, although we were unable to access all the sections of this small island, we developed an understanding of the overall layout of the settlement. After the completion of the discovery phase, we started mapping the architectural remains in the sections where bedrock prevented the growth of a dense plant cover, since we did not have the means or the time to carry out extensive vegetation clearance. Therefore, we limited our work to the northern and northwestern coastline, and to the northern part of the island’s ridge.

Of the five churches we located during the survey, we have mapped only the main contours of three churches (Churches I, II, V). For mapping, we utilized two Leica GNSS (Global Navigation Satellite System) units, one serving as the base station placed on the summit of the island, the other acting as the rover, operated in Real Time Kinematic mode to measure the terrestrial coordinates of the points marked on the sketch-plans, which we prepared during the first few weeks of the campaign. Total Station was used only sparingly because the vegetation cover severely impeded the visibility between the station and the reflector placed over the points to be measured. Instead, the rover unit, which does not require a visual connection with the base station, was the quicker, and therefore the preferred mapping method.

Finally, we carried out an aerial photographic (but not photogrammetric) documentation of the island to guide the mapping of the architectural remains in the upcoming seasons and also to give us an idea about the architectural remains in areas that we were unable to reach. We used a helium-inflated blimp carrying a remote controlled DSLR camera. The blimp was navigated by means of ropes, which were controlled by two team members, one walking on the island, the other stationed on the boat around the island.

Although our main focus was the discovery and mapping of architectural remains, we also sampled surface finds in select parts of the island. Following the stipulations of the research permit, we collected very small quantities of diagnostic sherds and glass, metal (mainly nails), and unique pieces of architectural revetment which we deposited to the storage of the Silifke Museum for preservation and analysis by specialists. Among these, four bronze coins were later studied by Prof. Oğuz Tekin from Istanbul University, who dated one coin to the fourth and the others to the seventh century CE. 

Preliminary Conclusions

The first campaign of the Boǧsak Archaeological Survey revealed that the settlement on the island was much more extensive than earlier scholars had noted. In fact, almost the entire surface was built up with structures constructed with mortared masonry of local limestone. Only the northern and western coastlines, where we identified maritime installations and stairs, served as the main points of access. The densest part of the settlement seems to be located on the northeastern slope, currently under heavy vegetation cover, where we found structures (probably houses) built on terraces. Their rectangular and arched windows overlooked Liman Kalesi and the harbor at Holmoi to the north (fig. 2). These one- or two-story buildings were covered with tiled pitched roofs. Their basements were equipped with deep vaulted cisterns. Similar remains on the eastern slope could be observed from a distance and on aerial photographs, but could not be accessed.  Encircling the summit on the north and south are cemeteries consisting mainly of graves carved into the bedrock. Finally, among the five churches found on the island, four were three-aisled basilicas with column capitals datable to the last quarter of the fifth century. Church V (fig. 3), a small square building with a centralized cross-shaped interior surmounted by a masonry dome, has previously been called a “martyrion” by Prof. Ayşe Aydın, based on its formal characteristics. This structure also stands out with its opus sectile pavement made of several types of imported marble that looters recently dug out and damaged. 

The 2010 campaign of the Boǧsak Archaeological Survey served as an introduction to a settlement that was previously very little known in the scholarship. The exploitation of the semiarid countryside in late antiquity is well studied in the Eastern Mediterranean, but islands have rarely become research subjects. The coastal settlements in the region have already been severely damaged or entirely destroyed since the 1970s, while surviving remains are under imminent threat by the tourism industry and modern development, which pays little attention to the preservation of the architectural heritage, let alone the less conspicuous and unattractive small finds without touristic value. The first campaign of the Boğsak Archaeological Survey has alerted local officials and the Department of Antiquities and Museums to the need to protect this unstudied and neglected island. Future campaigns will continue to document the exceptionally rich architectural and archaeological remains on Boğsak Island and its environs.


We would like to thank the archaeologists at the Silifke Museum who have greatly facilitated our survey: İlhame Öztürk (director), Özgür Topbaş (our representative), and Ulaş Demir. The first campaign of the project was made possible with the generous support of Dumbarton Oaks, Suna-İnan Kıraç Research Institute on Mediterranean Civilizations, and DÖSİMM of the Ministry of Culture and Tourism in Turkey.