©The Archaeological Settlements of Turkey - TAY Project

Beldibi / Kumbucagi

For site maps and drawings please click on the picture...


For photographs please click on the photo...

Beldibi / Kumbuca­ř
Rock Shelter
30 m
Investigation Method:


Location: This rock shelter site is 24 km southwest of the city of Antalya; 3 km north of the village of Beldibi; a village 800 m inland from the shore. It lies at the northern point of a long and thin beach; on a limestone outcrop by the sea; approximately 100 m from the shore. The shelter; which today is hidden by the forest vegetation around it; is difficult to find. The Antalya-Kemer highway is very close to the site. It is possible to reach the rock-shelter by walking down from this highway.
Geography and Environment: Because the limestone outcrop that the rock shelter lies in is locally known as Kumbucagi; this name has been chosen by the site director; Bostanci. However; in several publications the site has also been referred to as Beldibi; after the name of the closest village. There is a small cave; 5 m wide; 4 m deep and 3.5 m high; 12 m above the rock shelter. This cave has a 1.2 m soil deposit within it. It is in a very difficult place to reach. The rock shelter; on the other hand; lies in the terrace beneath; which has a semicircular base and which slopes towards the sea. The shallow rock shelter is 3 m wide. This is where the excavations were conducted. Although Bostanci notes that the depositional fill in the rock shelter is 5 m deep; in his site report he notes that he excavated to 6.20 m. The cave and the rock shelter are both in an area that is suitable for hunting and protection. It has not been reported whether there is a nearby water source. This area was used as a temporary campsite or as a habitation area by Upper Palaeolithic (?); Epipalaeolithic and Neolithic hunters.
Research and Excavation: There are paintings on the back wall of this rock shelter. Excavations here; with an initial 1.5x1.5 m trench by the wall paintings; began in 1959; three years after the site was discovered by Enver Yasar Bostanci [Bostanci 1959a:145]. Because the trench became too narrow for the removal of soil as it got deeper; a new trench had to be opened. This trench; perpendicular to the first is 2x2.5 m in size. In 1959 excavations stopped at 452 cm below the surface. Bostanci reports that the 1960; 1966 and 1967 excavations reached a depth of 6 m. The stratification of the soil in the upper cave was determined by the cross-section of a large hole opened in the cave; probably by treasure-hunting villagers. Most of the spoil the villagers excavated was thrown out of the cave. It takes place in the registered archaeological sites list prepared by Ministry of Culture and Tourism.
Stratigraphy: The excavations uncovered 7 or possibly 8 strata. The first three of these strata were each divided into two layers. Bostanci used letters to label the various strata. The stratigraphy is as follows: Stratum A Layer 1 (A 1): This first level of Stratum A is composed of soil with organic material. Modern pottery is abundant in this top layer. Stratum A Layer 2 (A 2): Classical/ Hellenistic finds were found in this approximately 70 cm thick layer. No change in the soil color or texture is noted. The excavation director dates Stratum A to Modern/Classical times. Stratum B is also composed of two layers. Stratum B Layer 1 (B 1): This is a Neolithic layer which is 53 cm thick. Stratum B Layer 2 (B 2): The pottery from this layer; which is 27 cm thick; also dates to the Neolithic. The tools from this stratum seem to show the continuation of the general Epipalaeolithic trends that can be found in Stratum C. Stratum C Layer 1 (C 1): In this layer the soil has a reddish tint and a mineralized texture because of high iron-oxide contents. The chipped stone industry in this phase shows that Upper Palaeolithic traditions continue along with new microlith technology. Stratum C Layer 2 (C 2): The soil in this layer is also mineralized and red from iron-oxide. The quantity of chipped stone tools and debitage products begins to decrease towards the bottom of the layer. Bostanci believes that Layer C 1 is contemporaneous with Beldibiyen III while C 2 is contemporaneous with the Beldibiyen IV culture. He classifies both layers of Stratum C as Epipalaeolithic/Mesolithic. The 20 cm thick layer of iron-oxide that lies between Layer C and the underlying Stratum D was formed; according to Bostanci; after a series of torrential rains that occurred in the third WŘrm Ice-Age. In this period there is a great decrease in the amount of chipped stone tools found. Stratum D: Even though there are two different layers within Stratum D; Bostanci does not separate them as D 1 and D 2. He reports that while the upper layer of this stratum corresponds to the Epipalaeolithic/ Mesolithic; the lower is Upper Palaeolithic in character. The soil is light in color and sandy. The information for Strata E; F; and G is very fragmentary and we are not supplied with soil descriptions or depths of these strata. Strata D; E; F; and G comprise the Upper Palaeolithic. Bostanci labels these strata after the Kemer municipality this site lies in and calls it the Kemeriyen Culture. Strata D through G are referred to as Kemeriyen I; II; III and IV respectively. While in some publications Bostanci refers to a Middle Palaeolithic stratum (Stratum H); his most recent publications make no reference to such a layer. His claims that there might be a Middle Palaeolithic Layer probably based on the fact that some of the Upper Palaeolithic tools had Middle Palaeolithic qualities. From the publications it is not clear whether virgin soil was reached (For a schematic stratification of the site see table in Bostanci 1967 b:130). Perhaps the most important element of the Kumbucagi rock-shelter is the rock art. While most of the paintings were made with red iron-oxide paint; some are engraved. The engravings are of deer figures outlined with a few simple lines over the natural veins of the rock-face. This antlered animal is depicted in fleeting position with its head turned back [Bostanci 1959:140; fig. III]. Bostanci dates this work of art to the Upper Palaeolithic. Some of the later paintings partially cover the earlier engraving. E. Anati; who also analyzed the wall paintings and carvings; noted that these figures closely resemble wall paintings and depictions on small objects in the Upper Palaeolithic in Western Europe [Anati 1968: 28]. In another publication; Bostanci reports that an ox which is lying down is depicted next to the deer figure [Bostanci 1971:23; end note 62]. He believes that the hunter gatherers who lived in and around the rock-shelter during C 1 and C 2 were the ones who produced the paintings of animal figures. The cave wall facing the sea is adorned with animals painted with red ocher/ iron oxide paints. It is believed that these figures depict hunting and religious scenes. Anati groups the wall paintings in five catagories. Type I refers to four cross shaped figures. Type II refers to anthropomorphic figures composed of a cross with a semi circle on top and forking legs beneath. Type III is also an anthropomorphic figure; very similar to Type II; but here the semi circles; which probably represent the heads; are divided into fours with a white dot in each quarter. Type IV refers to unexplainable triangles while Type V refers to animal figures; especially to a schematic depiction of a mountain goat. Anati believes that this mountain goat is the latest of all the wall paintings. The size of these figures vary between 6 and 23 cm. The remains of ocher and red painted pebbles in layers C 1 and C 2 confirm that this must have been the period when the red wall paintings were made. If the paintings are contemporaneous with C 1 and C 2 then these wall paintings must also date to the Epipalaeolithic/Mesolithic.
Small Finds: Chipped Stone Technology: Stratum C; which is especially rich in microlith tools; is dated to the Epipalaeolithic. This microlith technology includes tranchets; lunates; triangles; backed blades; rough Chateperon micro-points; angle burins and micro-burins [Bostanci 1967b:75-77; table IV]. In addition to the microlith technology; the Upper Palaeolithic includes rounded scrapers; core scrapers; retouched blades; micro stemmed points. Bostanci reports that he found flint fish hook-like tools as well as T and Y shaped flint implements. Thus these tools which include arrowheads and borers have functional importance over quantitative; as has been suggested. The microlith tools found in Upper Palaeolithic Stratum D are surprising; although Bostanci explains that the upper levels of this stratum contains some Epipalaeolithic finds. The technology here is reminiscent of the Nebekian and Kebaran Technologies. The lower layers of this stratum yielded genuine Gravette points while MagdalÚnien burins were also recovered [Bostanci 1967b:83]. Flint core scrapers; angle burins and elongated nosed blades and end scrapers are among the finds. Stratum E dates to the Upper Palaeolithic Period and yielded large scrapers with stepped edge flaking; thumbnail scrapers; rounded scrapers on large and small flakes as well as different types of five-stemmed tools; and Chatelperon micro points [Bostanci 1959a:153-154; fig. X-XI]. Strata F and G have yielded finds similar to Stratum E and the lower parts of Stratum D. These lower strata have slightly different types of triangular points which have been retouched. Chatelperon points on blades and a variety of burins have also been found in Stratum F and G. In one of the lowest strata; an Emireh type point with Levalloiso-MoustÚrien characteristics was found. Bostanci does not describe the chipped stone tools from Stratum G; a stratum that he includes in some of his publications. Other small finds: In Layer C 1; a 6.3 cm long and 4.7 cm wide pebble with a schematic depiction painted in red on one side was found. The painting is of a circle with arm-like extensions which has been elongated on one side and divided into four. Bostanci believes that the painting represents a woman giving birth. A similar small find was recovered from the same stratum; in Layer C 2. This is a flat 15.7x4.8x1.4 cm sandstone pebble with crisscrossing fish scales carved on both sides. Its head and especially its deeply carved eye have been carefully worked. This appears to be the earliest piece of sculpting found in Turkey so far [Bostanci 1976:79; Bostanci 1966:23; fig.1-2]. Small clumps of iron-oxide which are believed to have been used for red paint were found in bunches in Stratum C [Bostanci 1968:54-58; fig.3]. In addition to these finds dating to the Epipalaeolithic/Mesolithic; in the Upper Palaeolithic Stratum D; a pebble painted completely red with iron oxide paint was found. It is possible that this pebble may also come from level C above. A pierced sandstone amulet; several sea shells; deer antlers and teeth; antlers from a kid and fragments of a human skull were found. It is also reported that a pebble found in Upper Palaeolithic Stratum E was shaped like a human face in profile [Bostanci 1966:23;fig.3].
Interpretation and Dating:

To List