Navigation – Plan du site

AccueilNumérosXXVIIEarly Byzantine Structure at Gere...

Early Byzantine Structure at Gerenkuyu Mevkii of Yalı–Bodrum

İnci Türkoğlu
p. 109-125

Résumé

Preparations are underway for a structure in the Gerenkuyu Locality of Yalı neighbourhood in Bodrum district to be restored. The building was first identified as a church of Late Antiquity and registered as cultural property. It features a figural floor mosaic with a pair of leopards jumping onto a crater. Later publications identified it as a bathhouse of the Early Byzantine period. However, during the cleaning work carried out in late 2018 as preparation for a restoration project, new details have been uncovered bringing out the necessity of revising the structure’s function. The author’s proposal for the structure’s function is a countryside residence of a squire from the Early Byzantine period.

Haut de page

Texte intégral

The images used in this article are contributed by various people: Images dated 2018: Maylasa Mimarlık and the author; images dated 2012: Prof. Dr. A. Şevki Duymaz; and images dated 2010: Mehmet Uyargil. The plans are worked by the author on the original AutoCAD files of Architect Atilla Düz (MSc) and Maylasa Mimarlık Müh. Proje İnşaat Turizm İhracat İthalat Ltd. Şti. The author would like to express her gratitude to all those who contributed.

1. Introduction

1A structure in the middle of fields about 500 m from the coastline of the Barbaros Bay in the Yalı neighbourhood of the Bodrum district in the Muğla province is at the focus of preparations for a restoration project by the Bodrum Municipality. The building standing by itself amidst the fields suffered from the Kos earthquake in July 2017 and its northern half collapsed. In preparation for the restoration project, the structure was cleaned, uncovering some more details, which have been lost for a while. The structure is located at 36,9985N and 27,5137E; Sheet 019A-01A-2A [23], Insula 176, Lot 42 [1045].

2The building was registered as a monument on 26 April 2000 Decision no. 9358 by the Izmir Second Regional Board for the Preservation of Natural and Cultural Properties based on the inventory dated 10 November 1999. This inventory recorded the structure as a church from Late Antiquity (Fig. 1). V. Ruggieri’s monumental book on Byzantine Caria lists the monument’s location as Kamp Yeri (Ruggieri 2005: 79). Ruggieri dates the mosaic to the sixth century CE but ascribes no function to the building and places it to the post-Byzantine era, which is the Turkish period. In 2006, a bilingual article by Collins and Zäh presented the building in detail and based on similar examples for the fresco, mosaics and apsidal layout, it was proposed as a bathhouse from the reign of Justinian I (527-565 CE). The third volume of the cultural inventory of the Muğla province, prepared under the direction of A. Diler and published in 2013, gives a ‘revised’ inventory form and states that this is a bathhouse alongside giving Collins and Zäh as reference (Diler 2013: 1360). The same volume also gives a new version of the original registration inventory form citing the building as a bathhouse (Diler 2013: 1358). In the ‘Introduction’ section of volume III.1 (Diler 2013: 53-54) the monument at Gerenkuyu is described as a bathhouse from the sixth century CE.

Fig 1: Original registration inventory form dated 10 November 1999

Fig 1: Original registration inventory form dated 10 November 1999

2. The Building

3The rectangular structure is situated on a terrain descending slightly southwards and extends roughly in the east-west direction. It measures 14.30/14.40 m in the east-west and 7.80/7.99 m in the north-south directions. On the west façade is a small apse-like protrusion. No remains of other structures around are noted but a continuous heap of rubble extending out from the southern wall of the building might belong to a terrace wall of about 1 m height. Furthermore, small heaps of stones were piled up in the fields around by the local farmers. The structure was built with rubble, broken stones and lime mortar. On the corners are relatively larger and coarsely dressed blocks. Between the stones are smaller ones. There is an insignificant amount of brick use.

4North Façade: Currently, the north façade is in ruins except for a small part in the west (Fig. 2). Originally it housed the only doorway leading into the interior of the building and there was a small loophole window in the eastern half (Fig. 3). The pieces of the doorway, fallen down in the earthquake together with the rest of the façade, were removed and placed aside in the course of cleaning work in early December 2018. The jambs, lintel and threshold are of limestone. The jambs consists of multiple pieces; the western jamb is in three pieces while the eastern jamb is in four pieces. In the middle of both jambs is a horizontal piece giving a decorative effect. The façade bears traces of plaster in patches. Over the lintel was a shallow relieving arch of stone.

Fig 2: General view from the north, before cleaning work (November 2018)

Fig 2: General view from the north, before cleaning work (November 2018)

Fig 3: General view of the east and north façades (2010)

Fig 3: General view of the east and north façades (2010)

5East Façade: This blind façade has a loophole about 150 cm above the current ground level. The smaller square hole further down is not a window but rather one of the several holes where beams should have been placed. About the middle are traces of plaster (Fig. 4).

Fig 4: General view of the east and south façades (2018)

Fig 4: General view of the east and south façades (2018)

6South Façade: Accentuated with three windows in its western half, the south façade also has a loophole window and two holes for beams in its eastern half. Along the entire façade, a ledge of 10 cm height and 6 cm depth extends at the level of the window sills and it was later filled with small stones, few pieces of bricks and lime mortar with brick powder (Fig. 4).

7West Façade: This façade has an apse-like protrusion in the middle. In the middle of the upper part of this protrusion is a loophole window and on the top levels of the straight wall sections is another loophole window on either side. In addition, a hole for a beam is noted by the north edge, about mid-height (Fig. 5). However, a fig tree grown in the recent years conceals the protrusion from sight and has reached the level of a threat for the building. A section of 50-60 cm in height, running along the top of all the façades seems to have been a later construction, which may indicate a repair. This level corresponds to the vaults inside (Figs. 4, 5).

Fig 5: West façade (2010)

Fig 5: West façade (2010)

8Superstructure: The structure is covered with a flat roof on the exterior. The barrel vaults of stone and lime mortar inside are covered and levelled with a mixture of mud mortar probably containing local geren clay and stones and coated with a “concrete” layer (Başak and Bektaş, 1983: 121) (Fig. 6).

Fig 6: Detail of the west vault and the mud-stone filling and ‘concrete’-like flat roofing (2018)

Fig 6: Detail of the west vault and the mud-stone filling and ‘concrete’-like flat roofing (2018)

2.1. Interior

9The interior of the structure is arranged as three wings: north (Units 01, 02 and 02A) and south wings (Units 03, 04 and 05), which are adjacent to each other and extending in the east-west direction and bounded by the western wing (Units 06 and 07) (Plans 1&2). Each of these three wings are covered with a barrel vault.

10

11

12North Wing (Units 01, 02 and 02A): The only doorway leading to the interior is located about the middle of the north façade, somewhat offset to the east, and it opens directly onto the eastern end of Unit U01 (Fig. 3). Earlier photos show that it was arranged as a low-arched opening on the inside. Originally covered with a barrel vault in the east-west direction (Fig. 7), the entire vault of the north wing and a large part of its exterior wall fell in the earthquake of July 2017 (Fig. 2). Therefore, the northern half of the building was buried under debris until the cleaning work in early December 2018. As it is known from previous accounts and photographic documentation, the floor of U01 is paved with polychrome mosaics, which will be presented below (Fig. 8).

Fig 7: General view of the north wing, looking west (2010)

Fig 7: General view of the north wing, looking west (2010)

Fig 8: Mosaic pavement in U01 (2012)

Fig 8: Mosaic pavement in U01 (2012)

13Unit 02 to the east of U01 was originally accessed via a doorway at the southern end of the wall separating the two units (Figs. 7, 9a-b). The northern and southern walls of the unit were accentuated with a large but shallow niche (W. 270 x H. 150 cm); the extant southern niche bears traces of painted plaster but it is not possible for the time being to state whether these traces actually belonged to a fresco or not (Fig. 9b). The wall separating U02 from U03 was originally pierced by a round-arched doorway at its eastern end but it was later blocked with similar masonry; and, it is difficult to discern it from the north whereas it is easily noted from U03. At the top level of this doorway is a hole for a wall-beam. A second wall-beam hole is found on the extant southern niche. At the springing level of the barrel vault, now fallen, is a series of square holes for tie-beams. The existence of two supporting arches for the vault is known from previous photographs; one between U01 and U02 and the other between U01 and the west wing (Fig. 7).

Fig 9: Units 02 and 02A a: in 2010; b: after cleaning work in December 2018

Fig 9: Units 02 and 02A a: in 2010; b: after cleaning work in December 2018

14The cleaning work carried out in early December 2018 revealed traces of the floorings of the units, except U01. As a result of the cleaning work, Unit 02A became discernible and a stone blockage with mud mortar was uncovered before the non-extant northern niche and the entranceway to U02. This pavement is not attested in U02A. Between units U02 and U02A the foundation of a wall was uncovered but nothing indicating a doorway could be attested (Fig. 9b).

15A doorway right opposite the main doorway leads from U01 into U04 in the south wing. The round arch of this doorway looks somewhat pointed at top suggesting a somewhat amateurish workmanship. This doorway flares out into U04 (Fig. 10).

Fig 10: Doorway leading into U04, showing springing of the supporting arch and holes for tie beams (2018)

Fig 10: Doorway leading into U04, showing springing of the supporting arch and holes for tie beams (2018)

16South Wing (Units U03, U04 and U05): The south wing adjoins the north wing on its south side and has the same length. It was partitioned into three units. In the west, it is separated from the west wing with a wall pierced by a doorway in its southern end. Units 03 and 04 are separated from each other by two protrusions from the walls; U04 and U05 are separated by two wall sections with a passageway in between. All three units are covered with a single continuous barrel vault with no supporting arches but only traces of grey plaster.

17The easternmost U03 is arranged almost like an iwan opening onto U04 via two protrusions. At the eastern end of its north wall is a doorway, later blocked, originally opening into U02A; at the centre of its east wall is a loophole window; at the western end of its south wall is another loophole window. Each of these walls also have a hole for a wall-beam (Fig. 11).

Fig 11: Unit 03, looking east (2018)

Fig 11: Unit 03, looking east (2018)

18Protrusions from the north and south walls: Units 03 and 04 are thought to have the same floor level (Figs. 12a-b-c). The protrusions separating them from each other were built with two piles of coarsely dressed blocks bounding the niche recess and smaller stones with mortar in between, creating the semi-circular niche. Although they seem to be added later on, a close examination indicates that the walls were slightly caved to place the blocks and there is no plaster layer between the niches and the body walls (Fig. 13). The interiors of the niches were probably faced with marble plaques in seven or nine facets; however, currently, the “marble” plaques are covered with a layer of petrified black substance (Fig. 12c). The floor level of the north niche is clearly attested from the platform surviving but the south niche’s floor has been lost. Another interesting feature of these protrusions is their height reaching 200-220 cm from the current walking level as inferred from the traces on the walls. They were also surmounted with an arch. No traces of any water supply or chimney are attested to ascribe any function to them.

Fig 12: After cleaning work (December 2018)

Fig 12: After cleaning work (December 2018)

a: Southern protrusion niche; b: Northern protrusion niche; c: Detail of revetment in the northern protrusion niche

Fig 13: Joining of the partitioning walls (2018)

Fig 13: Joining of the partitioning walls (2018)

19Unit 04 in the middle of the south wing has a window opening out on its south wall. No traces of any flooring have been attested (Fig. 14).

Fig 14: General view of the south wing looking west: protrusion niches in the front, U04, partitioning walls, U05, doorway opening into U06; after cleaning work (December 2018)

Fig 14: General view of the south wing looking west: protrusion niches in the front, U04, partitioning walls, U05, doorway opening into U06; after cleaning work (December 2018)

20Unit 05 has interesting features. On its north wall is a shallow rectangular niche topped with a round arch. However, the window opening outside in its south wall is arranged as a semi-circular niche on the wall (Fig. 15) and the intrados of the window arch bears traces of a fresco (Collins and Zäh 2006: 295). The floor level of U05 is clearly lower than that of U04. The bottom level of the window’s niche is also higher than the current floor level of the room. The two blocks with a hole in them should have been brought there from elsewhere much later. In February 2019, the building was damaged by illicit diggers once again and in U05, a pit like a well with a diameter of about 1 m and depth of 2 m was dug in the floor before the window; this “well” cut through a blockage of rubble and lime mortar; there was no indication for any hypocaust or other installation under the floor (Fig. 16).

Fig 15: Niche-window arrangement in the south wall of U05 (2018)

Fig 15: Niche-window arrangement in the south wall of U05 (2018)

Fig 16: The “well” dug in U05 by illicit diggers in February 2019

Fig 16: The “well” dug in U05 by illicit diggers in February 2019

21The walls separating U04 from U05 rise about 2 m with a doorway left in between. These walls were not meant to carry any load and their connection to the main body walls is similar to that of the protrusions with niches separating U03 from U04 (Fig. 14).

22West Wing (Units U06 and U07): The west wing adjoins the north and south wings on their west and is covered with a north-south oriented barrel vault. The north half of the vault collapsed and serious cracks appeared in the 2017 earthquake. Currently it looks as it may collapse any moment. This wing was also divided into two by a stone wall rising about 2 m (Fig. 7), which collapsed in the earthquake as well (Fig. 17).

Fig 17: View of the apse-like protrusion and U07, after cleaning work (December 2018)

Fig 17: View of the apse-like protrusion and U07, after cleaning work (December 2018)

23Unit 06 opens out with a round-arched window in the south wall and a wall-beam hole over it. The doorway at the south end of its east wall leads into U05. On its west wall is an apse-like protrusion, slightly bigger than a semicircle, protruding out (Fig. 18). Cleaning work brought to light a flooring of mortar with dense brick powder (horasan).

Fig 18: Unit 06, looking south, after cleaning work (December 2018)

Fig 18: Unit 06, looking south, after cleaning work (December 2018)

24The apse-like protrusion has one loophole window and three holes, possibly to hold beams. It is topped with a very shallow arch and a flat roof. Its floor is higher than that of U06 and has similar mortar. South of the apse-like protrusion, on the top level of the west wall is a loophole window.

25The wall that used to separate U06 from U07 is now preserved at the foundation level and during the cleaning work it became clear that it was rendered with a plaster with brick powder about 5-6 cm in thickness on its both faces (Fig. 17). Cleaning work at unit U07 brought to light no flooring but three layers of plaster on the walls.

2.2. Decoration

26Mosaic Flooring in U01: The mosaic flooring at U01 was previously covered with geotextile and sand by the Bodrum Museum of Underwater Archaeology; therefore, the author has not been able to see it. However, the following can be reported from previous publications and old photographs (Fig. 8):

27The polychrome mosaic pavement fills the entire U01 measuring about 478 x 300 cm and its borders follow the walls of the unit, thus clearly indicating that it was made for this unit; that is, it does not belong to any earlier or other structure. The outermost wide border features a scroll of large ivy leaves. Five ivy leaves on the long sides and four on the short sides are placed in alternating direction. Then comes a set of two narrow bands flanking a wider band of guilloche. The central panel is contoured with a thin border and features a figural composition. In the centre of the panel is a large crater upon which two speckled leopards jump. The tails of the leopards curve up filling the southeast and southwest corners of the panel.

28In their study on this building, Collins and Zäh (2006: 296-298) state that the white-grey crater with black contours on a white background is filled with a red liquid (wine or blood); the leopards are yellow-brown in colour; the guilloche is in white-red/orange shades; and the outermost border of vegetal scroll has black ivy leaves on white background. It is also stated that the central and the right side of the panel are partially well-preserved. The leopard figures belong to the repertory of Roman and Byzantine mosaic art (hunting and circus scenes).

29Comparable examples with regards to the theme can be cited as the mosaics of the Great Palace in Constantinople (Istanbul) (sixth c. CE); Collins and Zäh (2006: 299) also cite the mosaics from a villa at Miletus, now housed at the Pergamon Museum in Berlin (about 200 CE); the mosaics of the Church of St. Christoph at Qabr Hiram in Lebanon, now at the Louvre (about 575 CE); and the Dar Buc Amra mosaic of Zliten in Tripolis, Libya (first c. CE). Floor mosaics of similar quality are also known from the Early Byzantine monastery complex at Torba in the close proximity, which are currently also concealed under a protective layer of geotextile and sand (Özet 2009: 71-82).

30Fresco Remains in U05: The window in the south wall of U05 is arranged into a semi-circular niche in the wall; its intrados was originally painted with a figural fresco, which is in a very poor condition today (Fig. 19). However, Collins and Zäh (2006: 295) state that a soldier’s head with a helmet of Late Antiquity is depicted there and give for comparison a follis coin depicting Justinian I (r. 527-565 CE) wearing such a helmet.

Fig 19: Fresco traces in the intrados of window of U05 (2018)

Fig 19: Fresco traces in the intrados of window of U05 (2018)

2.3. Structural Notes

31Holes for tie beams and holes for wall beams: At the springing level of the vaults in all three wings is a series of square holes for placing tie beams, which originally improved the strength of the vaults (Figs. 9a, 10, 18, 11).

32In the walls, there are square holes cutting across the thickness of the walls and their positions from the ground level vary in height. These holes do not exhibit a pattern of placement except the fact that they are mostly either in the middle or lower parts of the walls. It is plausible that wooden beams were placed in them, possibly for improving the strength of walls (Figs. 10, 11, 17, 18).

33The only relieving beam attested is along the entire south façade at the level of the window sills, extending further about 1 m on the west façade. The place of the beam was later filled with a lime mortar containing brick fragments and powder (Fig. 4).

34Partition walls rising only 2 m: The walls separating U04 from U05 and U06 from U07 seem to have been only to a height of about 2 m as inferred from the traces on the body walls and previous photographs. At first glance, they seem to be built adjoining the body walls at a later date but a closer examination shows that they actually belong to the original construction phase because they do join into the body walls without any plaster layer in between. It is clear that these partition walls were not meant to carry any load and as far as it could be observed during the cleaning work, they do rise on foundations. The joining method of the partition walls is also attested for the joining of the protrusion niches between U03 and U04 (Fig. 13).

35Plaster: Walls were plastered over both on the inside and the outside as inferred from traces. On the exterior, the plastering did not cover the entire surface of especially larger stones. On the interior, particularly in the U01, U02 and U07, there is a very thin layer of plaster, which is not concealing the texture of the masonry. At the southwest and northwest corners of U07 three different layers of plaster were attested. The bottom one is a thin layer; the middle one is about 3-4 m in thickness and the top layer with high brick powder content is 6 cm thick, similar to that seen in U06 (Fig. 20a). The Early Byzantine monastery complex at Torba has a similar plaster with high brick powder content (Figs. 20b-c) used as bedding for marble facing. However, under present circumstances there is no evidence attested for similar marble facing at Gerenkuyu.

Fig 20: After cleaning work (December 2018)

Fig 20: After cleaning work (December 2018)

a: Northwest corner of U07, layers of plaster; b: Thick plasters with high content of brick powder from the west wing; c used as bedding for marble facing at the Early Byzantine monastery complex in Torba (2018)

36 In U04 and U05, the protrusion niches and partitioning walls bear traces of lime plaster with less brick powder content. The vaults still bear concrete plaster applied most likely using planks.

37 Supporting arches for the vaults: Only the north wing’s vault had two supporting arches but the south and west wings do not have. This indicates interventions to the superstructure in time.

3. Construction phases

38Based on the current situation it can be proposed that the structure has had the same overall dimensions throughout its history. There are no signs visible for any annex built or demolished. On the exterior, the masonry over the level of the windows is slightly different from the lower parts, on all façades. This different masonry is attested on a larger area in the eastern part of the southern façade and part of the east façade. This hypothesis is also supported by the anomaly visible in the upper part of the north wall of U03 (Fig. 21). This anomaly consists of two square depressions, recalling shallow but larger holes for ‘tie beams’, and a little over them the wall seems to be cut and continued somewhat narrower. All these give the impression that the upper parts of the structure were rebuilt. Based on the examinations it is possible to propose two scenarios for the original building: a– It was built as a single-story building; b– It was a two-story building; when the sixth-century earthquake hit, the upper floor fell and never been rebuilt. In this case U02A could be the stairwell in the original construction.

Fig 21: North wall U03, showing the anomaly and holes for tie beams (2018)

Fig 21: North wall U03, showing the anomaly and holes for tie beams (2018)

39First Phase: The structure was built together with all its partitions and units. Although partitioning walls and protrusion niches seem to have been added later, as mentioned above, they belong to the original construction phase. Nothing is known about the original superstructure but most likely, it consisted of barrel vaults as it is now. The mosaic pavement in U01 belongs to this phase. The fresco in U05 should also have belonged to this phase. Unit 03 may have had a window on its south wall which was converted into the current loophole window but there is no clear-cut evidence for it. Even a window on the east wall of U03 may be proposed for the place of the loophole window (Plan 2).

40Second Phase: The first structure lost its superstructure –and its upper floor if there had been one– and part of its south wall in the east, and possibly part of its eastern wall, as inferred from the masonry on the exterior, most probably in an earthquake. These were rebuilt creating the anomaly in the upper part of the north wall of U03. However, the partitioning walls were not raised up to the vault. The new masonry in the upper parts does not contain any bricks but rather, small creek stones inserted in between the bigger ones. Probably in this phase, the wall between U02 and U02A as well as the one between U01 and U02 were not rebuilt. The doorway between U02A and U03 was blocked with similar masonry. In this phase, it is also possible that a window in the south wall of U03 was converted to the loophole window we see today. The apse-like protrusion was topped with a flat masonry roof. All the loophole windows were built in this phase.

41Third Phase: Period of decline; the doorway between U05 and U06 was blocked with masonry employing mud mortar (most probably by shepherds or others sheltering here). The main doorway in the north and the supporting arches of the north wing were built, possibly after some earthquake damage. Destruction was accelerated by illicit digs and earthquakes, and culminated in the final collapse of the northern half in the earthquake of July 2017.

4. Dating

42The evidence available for the initial construction of the structure consists of the mosaic pavement in U01 and the fresco remain in U05. The figural composition in the mosaic panel may be attributed to the sixth, at the latest seventh century CE based on comparative materials cited above. The fresco remain in U05 was attributed to the reign of Justinian I (527-565 CE) by Collins and Zäh (2006: 295). In this case, it should be plausible to ascribe the initial construction to the first half of the sixth century CE at the latest. The second phase should not be distant in time from the first as the masonries are quite similar and the second phase is not even attested easily. Guidoboni (1994: 338-339) cites a severe earthquake between 554 and 558 CE, centred at Kos razing it to the ground, and causing a seismic sea wave (tsunami). It is highly likely that this earthquake was the reason for the end of the first phase.

43After the arrival of Turks in the region in the early thirteenth century, the building was not used extensively by the Muslim population in the later centuries as inferred from partial survival of the figural mosaics and frescoes. However, Ruggieri attributes the structure to the post-Byzantine period and the apparent reason for this should be the construction style of the main doorway –now in ruins. Anyone strolling downtown Bodrum will certainly see many examples of gates built in the same fashion; Bodrum’s traditional Chian-type houses with courtyards –attributed to the nineteenth and twentieth centuries– have gates with similar construction, opening into the courtyard from the street (Başak and Bektaş 1983: 86-92; Yücel Besim 2007: 40-41). This suggests that –at least– the main doorway of the structure was rebuilt at some point in the last two centuries. Therefore, it is necessary to check the earthquake data available. Ambraseys and Finkel (1995) list earthquakes attested for 1500-1800 from sources for Turkey and adjacent areas. #32, 35, 62, 68, 92, 109, 126, 211 and 331 cite earthquakes centred at Rhodes, Kos and Milas. Of these, #62 in 1616 at Rhodes, #68 in 1631 at Milas, #109 in March 1673 at Kos, #126 in 1685-86 at Rhodes, #211 on 31 Jan. 1741 at Rhodes may seem to have had their tolls on this structure. Furthermore, Altunel et al. (2003) cite various earthquakes with a magnitude ≥6 in the vicinity for the twentieth century. Altunel et al. (2003, Fig. 2) marks a fault line extending along the north shore of Gökova Gulf and earthquakes in 1926 on the east coast of Rhodes, in 1933 centred at the Gulf, in 1957 and 1961 offshore Marmaris and Fethiye, in 1959 at the eastern part of the abovementioned fault line, and in 1956 off the southwest coast of Kos. All these earthquakes may have had their tolls on the structure and this may explain the late fashion of the main doorway. And this also indicates that the structure remained in use by people, who were not hostile to images. Furthermore, the barrel vaulting –or, at least the “concrete plastering” on them-, the “concrete” layer on top of the roof and the supporting arches in the north wing may also be attributed to this phase.

5. Function of the Structure

44Church: In the original registration inventory, the structure is noted as a church (Fig. 1), most likely based on the apse-like protrusion. For a church, the apse is expected to be on the side facing east (or Jerusalem) but the direction of the apse-like protrusion at Gerenkuyu is on the west side. Furthermore, houses of prayer always aim to gather as many people as possible based on the size of the community. The houses of prayer in the major monotheistic religions of Judaism (synagogue / kahal), Christianity (church / ekklesia) and Islam (mosque / djam‘i) all are attributed a term simply meaning “community / congregation” and “gathering place”; thus, it is inferred that the aim is to gather as many people as possible. However, the structure at Gerenkuyu is divided into smaller chambers not suitable for holding a “congregation” –perhaps suitable for a Jewish minyan only– yet, the direction of the presumptive ark (=the protruding apse-like niche) is again wrong. No chamber in the structure seems to be meant for Christian liturgy –perhaps, U06 only as an oratory within the residence.

45Bathhouse: The cultural inventory (Diler 2013: 1360) records the building as a bathhouse based on the article by Collins and Zäh, who proposed the bathhouse identification based on the apsidal layout and comparison with similar structures (2006: 301-305). The Early Byzantine monastery complex at Torba has a bathhouse, which features a series of small chambers with hypocaust and furnace (Ruggieri 2005: 122-135, esp. 128-132). A similar layout is also attested at the bathhouse with hypocaust uncovered to the east of the Istanbul Archaeological Museums within the premises of the Topkapı Palace (Müller-Wiener 2001: 50, Fig. 27; Yegül 2006: 300, Fig. 324). The bathhouse at Şeytan Bükü has multiple small rooms covered with barrel vaults (Ruggieri 2003: 245-250). At Alakışla, to the west of Keramos (Ören) and east of Gerenkuyu, there is a group of bathhouses within the settlement and among these T3 with its apsidal rooms of small dimensions is given as a comparative structure by Collins and Zäh (2006: Fig. 14.5, cited as ‘Ala Kilise Therme II’; Ruggieri 2003: 188-190 Fig. P28). Late Roman baths at Syedra in Cilicia is also given for comparison (Collins and Zäh 2006: Fig. 14.1). The renowned baths within the Qusayr ‘Amra palace complex of the Umayyads in today’s Jordan has multiple apsidal rooms for comparison (Collins and Zäh 2006: Fig. 14.2). All these structures cited for comparison by Collins and Zäh, feature apsidal rooms and barrel vaults. The bathhouse complex uncovered at Amorium has a rectangular main building with a spectacular polygonal structure attached to it; it features the typical hypocaust system of antiquity (Lightfoot et al. 2005: 233-241). The sixth century East Baths in Andriake also feature the typical halls and hypocaust system (Çevik and Bulut 2014; Niewöhner 2012: 224-228). The same is also valid for the bathhouse, located by the odeion in Kibyra (Özüdoğru and Dökü 2012). In addition, the restored bathhouse at Aspat is covered with barrel vaults; has a hypocaust; is dated to the Middle Byzantine period (Diler 2013 III.1: 52-53), and its current look with barrel vaults recalls that of Qusayr ‘Amra. As for Gerenkuyu, barrel vaults – yes; but the apsidal room is a question mark because the apse-like protrusion is more like a big niche, which is slightly bigger than a semicircle, whereas in the examples given by Collins and Zäh for comparison and others at Istanbul, Amorium, Andriake, and Kibyra cited above, the niches are almost as wide as the walls of the chambers. In the Gerenkuyu structure, the chambers, or units, are not physically isolated with walls from each other as in the other examples cited above.

46However, for a structure of the Early Byzantine period to be identified as a bathhouse it needs to have a hypocaust system for heating, baked clay pipes in the walls, water supply and drainage systems, stokehole, furnace, waterproof plastering, etc. (ODB “Baths”; Berger 1982: esp. 102-108; Berger 2012; RE “Bäder”; RE “Hypocaustum”). None of these pre-requisitions –except the plaster– have been attested in this structure. Local workers of the Bodrum Municipality stated that previously they had filled in the pits in U02 and U04 dug by illicit diggers and they noted no cavities that may point to a hypocaust system, or any flooring at all. In February 2019, the structure was targeted by treasure hunters once again; this time, a “well” about 2 m in depth and less than 1 m in diameter was dug in the floor of U05, in front of the window with a niche, cutting through the blockage of rubble and mortar (Fig. 16); again, no hypocaust was attested. Thus, the only evidence to support the bathhouse function is the traces of thick lime plaster with brick powder content. Furthermore, one would expect to find a settlement around or nearby a bathhouse but the heaps of stones in the fields around do not look satisfactory enough to envisage a settlement. To the north of the building are two water wells only, built with some ancient spolia, and still in use, for irrigating the surrounding land.

47Our proposal: The author is of the opinion that this building was most likely a house from the Early Byzantine period. The interior of the building does not show any signs for a heating or water supply system to support a bathhouse function proposal. However, the floor mosaic in U01, the protrusion-niches with revetment, and the traces of a fresco all suggest that the original owner of this structure was not a commoner. On the other hand, the absence of a bathing area and toilets within the building proper indicates that the toilets were located outside the house, in the yard –a tradition attested in Anatolia until very recently. About 30-40 m north of the structure are two wells, which could have supplied the needs of this estate. The level difference of about 1 m by its south wall and the heaps of stones may suggest that there were auxiliary structures around and possibly a terracing.

48The block form of the structure and the way it is standing alone suggest that it was the residence of a squire who owned the surrounding land. Niewöhner’s study (2015: 36-40) on the Late Antique origins of the block form palaces of the Late Byzantine period, e.g. the Tekfur Palace in Istanbul and the Laskarid Palace at Nymphaion (Kemalpaşa), cites as examples some rural houses from the Early Byzantine period: at Sinekkale in Cilicia (Eichner 2008), at Kirse Yanı in Caria (Giese and Niewöhner 2016), and at Andriake in Lycia (Niewöhner 2012: 228-231). However, each of these three houses is two-storied and the closest one at Kirse Yanı has an annexed bathhouse. The Gerenkuyu example distinguishes itself with its single-story block form construction; its windows facing the sea direction and the absence of a bathhouse. However, this may remind us of the other scenario: in the first phase, the structure might have been two-storied; then the upper story fell in the earthquake mentioned above, and had not been rebuilt during the reconstruction. Furthermore, the thick layer of plaster with high content of brick powder attested in U06 may be the bedding for marble facing, as suggested from the monastery complex at Torba (Figs. 20b-c). Then, this would allow us to imagine a somewhat luxuriously decorated residence of a squire from the Early Byzantine period.

Haut de page

Bibliographie

Altunel, E., Stewart, I.S., Barka, A. and Piccardi, L., 2003: “Earthquake Faulting in Ancient Cnidus, SW Turkey”, Turkish Journal of Earth Sciences 12: 137-151.

Ambraseys, N. N. and Finkel, C. F., 1995: The Seismicity of Turkey and Adjacent Areas. A Historical Review, 1500-1800, Istanbul, Eren Publishers.

Başak, S. and Bektaş, C., 1983: Bodrum: Halk Yapı Sanatından Bir Örnek, Istanbul, Apa Ofset.

Berger, A., 1982: Das Bad in der byzantinischen Zeit, Miscellanea Byzantina Monacensia Heft 27, Munich, Institut fur Byzantinistik und Neugriechische Philologie der Universitat Munchen.

– 2012: “Bizans Çağında Hamamlar”, In: Ergin, N. (ed.), translated by Özbay Erozan, A., Anadolu Medeniyetlerinde Hamam Kültürü: Mimari, Tarih ve İmgelem, Istanbul, Koç University Publications: 67-83. Original publication: Ergin, N. (ed.), Bathing Culture of Anatolian Civilizations: Architecture, History and Imagination, Peeters, 2011.

Collins, B. and Zäh, A., 2006: “Terme bizantine in Caria: Una struttura termale protobizantine a Gerekuyu Dere inferiore presso Bodrum / Byzantinische Thermen in Karien: Eine früh byzantinische Thermenanlage im unteren Gerekuyu Dere bei Bodrum”, Quaderni Friulani di Archeologia XVI: 291-307.

Çevik, N. and Bulut, S., 2014: “Andriake Doğu Hamamı: Bölgenin Hamam Mimarlığına Işık Tutan Yeni Bir Örnek [The East Baths at Andriake: A New Example Casting Light on the Bath Architecture of the Region]”, Adalya 17: 221-262.

Diler, A. (project director), 2007: Muğla Valiliği Muğla Kültür Envanteri – I: Bodrum Kentsel Sit Halikarnassos, Muğla, Muğla Sıtkı Koçman Üniversitesi Karya Araştırma ve Uygulama Merkezi Yayınları.

Diler, A., Gürbıyık, C. and Gümüş, Ş., 2007: “Bodrum’un Türk Dönemi Tarihi ve Eski Belgelerde Bodrum”, In: Diler, A. (ed.), Muğla Valiliği Muğla Kültür Envanteri – I, Muğla, Muğla Sıtkı Koçman Üniversitesi Karya Araştırma ve Uygulama Merkezi Yayınları: 32-37.

Diler, A. (project director), 2013: Muğla Valiliği Muğla Kültür Envanteri III.1 & III.2: Bodrum Yarımadası Arkeoloji ve Sanat Tarihi Kalıntıları, Muğla, Muğla Sıtkı Koçman Üniversitesi Karya Araştırma ve Uygulama Merkezi Yayınları.

Diler, A. (project director), 2013: “Değerlendirme”, In: Diler, A. (ed.), Muğla Valiliği Muğla Kültür Envanteri – III.1, Muğla, Muğla Sıtkı Koçman Üniversitesi Karya Araştırma ve Uygulama Merkezi Yayınları: 31-97, 52-54, 90-96.

Eichner, I., 2008: “Sinekkale – Herberge, Kloster oder Gutshof?”, Olba 16: 337–360.

Giese, S. and Niewöhner, Ph., 2016: “Das frühbyzantinische Landhaus von Kirse Yanı in Karien”, with a contribution by R. Descat, Istanbuler Mitteilungen 66: 293-352.

Guidoboni, E., 1994: Catalogue of Ancient Earthquakes in the Mediterranean Area up to the 10th Century, Rome, Istituto nazionale di geofisica.

Lightfoot, Ch., Arbel, Y., Ivison, E.A., Roberts, J.A. and Ioannidou, E., 2005: “The Amorium Project: Excavation and Research in 2002”, Dumbarton Oaks Papers 59: 231-265.

Müller-Wiener, W., 2001: İstanbul’un Tarihsel Topografyası, transl. by Ü. Sayın, Istanbul, Yapı Kredi Yayınları. Original publication: Bildlexikon zur Topographie Istanbuls: Byzantion, Konstantinupolis, Istanbul, bis zum Beginn des 17. Jahrhunderts. E. Wasmuth, Tübingen, 1977.

Niewöhner, Ph., 2012: “Andriake in byzantinischer Zeit” In: Seyer, M. (ed.) 40 Jahre Grabung in Limyra, Vienna, ÖAI: 223-240.

Niewöhner, Ph., 2015: “The late Late Antique origins of Byzantine palace architecture”, In: Featherstone, M., Spieser, J.-M., Tanman, G., Wulf-Rheidt, U. (eds.) The Emperor’s House. Palaces from Augustus to the Age of Absolutism. Berlin – Boston, De Gruyter: 31-52.

ODB, 1990: The Oxford Dictionary of Byzantium, 3 vols. A. Kazhdan (ed.-in-chief), Oxford.

Özet, A., 2009: “Bodrum-Torba Monastery Mosaics”, Journal of Mosaic Research 3: 71-82.

Özüdoğru, Ş. and Dökü, E., 2012: “Kibyra 2011”, Anmed 10: 46-52.

RE “Bäder”: “Paulys Realencyclopädie der classischen Altertumswissenschaft”, Vierter Halbband, cols. 2743-2758.

RE “Hypocaustum”: Paulys Realencyclopädie der classischen Altertumswissenschaft, Siebzehnter Halbband, cols. 333-336.

Ruggieri, V., 2003: Il Golfo di Keramos: dal tardo-antico al medioevo bizantino, Rubbettino, Rubbettino Editore.

Ruggieri, V., 2005: La Caria Bizantina: topografia, archeologia ed arte (Mylasa, Stratonikeia, Bargylia, Myndus, Halicarnassus), Rubbettino Editore, Rubbettino.

Yegül, F., 2006: Antik Çağ’da Hamamlar ve Yıkanma, transl. by Erten, E., Istanbul, Homer Kitabevi. Original publication: Baths and Bathing in Classical Antiquity, MIT, 1992, 1995.

Yücel Besim, D., 2007: Özgün Kentsel Mekanların Okunması ve Belirlenmesi Üzerine Analitik Bir Çalışma: Bodrum Türkkuyusu Örneği [An Analytical Study on the Reading and Identification of Authentic Urban Spaces: Case of Bodrum Türkkuyusu], unpublished PhD dissertation, Ankara University.

Haut de page

Table des illustrations

Titre Fig 1: Original registration inventory form dated 10 November 1999
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-1.jpg
Fichier image/jpeg, 660k
Titre Fig 2: General view from the north, before cleaning work (November 2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-2.png
Fichier image/png, 745k
Titre Fig 3: General view of the east and north façades (2010)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-3.jpg
Fichier image/jpeg, 172k
Titre Fig 4: General view of the east and south façades (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-4.png
Fichier image/png, 1,0M
Titre Fig 5: West façade (2010)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-5.jpg
Fichier image/jpeg, 1,6M
Titre Fig 6: Detail of the west vault and the mud-stone filling and ‘concrete’-like flat roofing (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-6.jpg
Fichier image/jpeg, 2,6M
Titre Plan 1
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-7.jpg
Fichier image/jpeg, 132k
Titre Plan 2
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-8.jpg
Fichier image/jpeg, 80k
Titre Fig 7: General view of the north wing, looking west (2010)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-9.jpg
Fichier image/jpeg, 168k
Titre Fig 8: Mosaic pavement in U01 (2012)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-10.png
Fichier image/png, 917k
Titre Fig 9: Units 02 and 02A a: in 2010; b: after cleaning work in December 2018
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-11.png
Fichier image/png, 2,0M
Titre Fig 10: Doorway leading into U04, showing springing of the supporting arch and holes for tie beams (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-12.png
Fichier image/png, 1,8M
Titre Fig 11: Unit 03, looking east (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-13.jpg
Fichier image/jpeg, 2,1M
Titre Fig 12: After cleaning work (December 2018)
Légende a: Southern protrusion niche; b: Northern protrusion niche; c: Detail of revetment in the northern protrusion niche
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-14.png
Fichier image/png, 2,6M
Titre Fig 13: Joining of the partitioning walls (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-15.png
Fichier image/png, 1,7M
Titre Fig 14: General view of the south wing looking west: protrusion niches in the front, U04, partitioning walls, U05, doorway opening into U06; after cleaning work (December 2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-16.png
Fichier image/png, 1,8M
Titre Fig 15: Niche-window arrangement in the south wall of U05 (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-17.jpg
Fichier image/jpeg, 1,8M
Titre Fig 16: The “well” dug in U05 by illicit diggers in February 2019
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-18.jpg
Fichier image/jpeg, 192k
Titre Fig 17: View of the apse-like protrusion and U07, after cleaning work (December 2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-19.png
Fichier image/png, 1006k
Titre Fig 18: Unit 06, looking south, after cleaning work (December 2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-20.png
Fichier image/png, 1,8M
Titre Fig 19: Fresco traces in the intrados of window of U05 (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-21.jpg
Fichier image/jpeg, 3,5M
Titre Fig 20: After cleaning work (December 2018)
Légende a: Northwest corner of U07, layers of plaster; b: Thick plasters with high content of brick powder from the west wing; c used as bedding for marble facing at the Early Byzantine monastery complex in Torba (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-22.png
Fichier image/png, 3,4M
Titre Fig 21: North wall U03, showing the anomaly and holes for tie beams (2018)
URL http://journals.openedition.org/anatoliaantiqua/docannexe/image/825/img-23.jpg
Fichier image/jpeg, 2,9M
Haut de page

Pour citer cet article

Référence papier

İnci Türkoğlu, « Early Byzantine Structure at Gerenkuyu Mevkii of Yalı–Bodrum »Anatolia Antiqua, XXVII | 2019, 109-125.

Référence électronique

İnci Türkoğlu, « Early Byzantine Structure at Gerenkuyu Mevkii of Yalı–Bodrum »Anatolia Antiqua [En ligne], XXVII | 2019, mis en ligne le 31 janvier 2022, consulté le 20 mai 2024. URL : http://journals.openedition.org/anatoliaantiqua/825 ; DOI : https://doi.org/10.4000/anatoliaantiqua.825

Haut de page

Auteur

İnci Türkoğlu

Pamukkale University, Faculty of Letters and Sciences, Department of Conservation and Restoration of Cultural Heritage, Denizli
inciturkoglu@hotmail.com

Haut de page

Droits d’auteur

CC-BY-SA-4.0

Le texte seul est utilisable sous licence CC BY-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search