Introduction

Figure 1.

Location of survey areas in relation to Wadi es Sir limestone and known karstic caves in Israel.

The Ritual Landscapes Project (RLP) was conceived to test the premise that ossuary burial caves of the Chalcolithic period (4500-3800/3600 BCE) are only found west of the Jordan River. A survey of the karstic regions of Jordan was undertaken, targeting sites resembling the ossuary caves that have been discovered at ever-increasing rates in Israel/Palestine. Similar geological formation, vegetation and Chalcolithic settlement sites are present on both sides of the valley, therefore these caves ought to be present in Jordan. We surmised that the lack of caves in Jordan might be due to slower industrial development there, while in Israel large numbers of new caves are found by bulldozers. There is also a tradition of adventure caving in Israel that is not paralleled in Jordan.

Background

In the southern Levant various Chalcolithic burial types are cited: dolmen fields, cist tombs, grave circles and occasional intramural burials suggesting, perhaps, highly regional expressions of mortuary culture. However, recent discoveries at Nahal Qanah, Peqi'in, Kissufim Road, Givat Ha-Oranim, Horvat Castra, Shoham (Nth) and others have demonstrated clearly that Chalcolithic ossuary cave sites like those published by Perrot and Ladiray (1980) are not confined to the coastal plain and are perhaps the more common form of Chalcolithic burial (van den Brink 1998, 2005) west of the Jordan River. The impressive wealth in the karstic versions of these caves (i.e. caves formed by the dissolution of limestone), e.g. in the form of metal, has made their discovery even more astounding.

Figure 2.

Iona McRae surveys RL5 (al Daher cave) which contained Iron Age material.

Preliminary results

To date the RLP has surveyed two areas of the limestone formation required to form the type of karstic caves, what is termed the Wadi es Sir formation: Survey Area A in the northern highlands and Survey Area B in the Madaba-Mujib area (Figure 1). Note we draw a distinction here between karstic caves and those in other formations that appear to be artificial, i.e. dug out of softer rock, like coastal kurkar (e.g. Azor, Horvat Qarqar). These burial caves are no doubt related to karstic ossuary caves, but are unlikely to be found in Jordan where those geological formations do not appear.

Our results fall into two main categories: RL caves (those with archaeological material) and blank caves (empty caves, rock-shelters or rock-cut caves/tombs). In total we surveyed 302 blank caves and 52 RL caves (Figure 2). Approximately two thirds of the RL caves are located in Survey Area A, and those in Survey Area B are all smaller than 30m deep. This might be due to the fact that the WSL is quite thin in the south, and is exposed by wadi down-cutting, while in Survey Area A it covers a vast distance and is extremely thick (c.200m). In this area there was no shortage of caves, although the largest systems were located in the north east of the survey area - in the higher altitude, higher rainfall zone - and many were partially or completely inaccessible due to collapse (Figure 3).

Figure 3.

Adam Evans surveys RL20, a collapsed cave, now a 'sink hole' near Birgish.

Figure 4.

Jaimie Lovell stabilises the bronze axe heads from RL53, in the Wadi Rayyan.

Figure 5.

Adam Evans and Henry Rockliff work on the cliff face of RL25.

Figure 6.

Selected finds from the RL caves found in 2006 and 2007.

Our survey established the existence of hitherto unknown caves in Jordan, including RL53, which contained two bronze axe heads dating to the Middle Bronze or later (Figure 4; Figure 6.1-2) but Chalcolithic burial caves were more difficult. We did locate stratified Late Neolithic material in a karstic cave in Baun (Figure 6.5) suggesting that karstic cave use was well established by that time.

Tantalising traces of Chalcolithic material are represented in material in scree below caves, e.g. RL25, a group of caves in a single cliff face (Figure 5). Over a dozen caves were loosely associated with Chalcolithic material in this way (RLs14, 25, 58-60, Figures 6.8, 11). One cave, RL58, had a fragment of an undecorated stone ossuary (Figure 6.9) outside it, which could date to a later period. We documented a destroyed cave in the village of Judayta in the Wadi Rayyan described to us by the owners of the house now occupying the site. Chalcolithic pottery was found in the cut made by the house which may have derived from the cave (RL3). Therefore it is possible that Chalcolithic caves have been located and excavated unofficially, thereby explaining Chalcolithic items found in scree.

Conclusions and future work

The survey results do clearly document the existence of karstic cave systems in northern Jordan. However, Chalcolithic caves still appear under-represented. Therefore, there is reason to argue that there were regional expressions of burial in the Chalcolithic in which climate and geology played a major role. Stuart Robinson (University College London) and Stuart Black (University of Reading) have sampled speleothem from RL caves in order to produce the relevant palaeoclimatic data. It is hoped that a comparable sequence to that generated for the Soreq and Peqi'in caves will soon be available which will inform our arguments on cave formation in relation to Chalcolithic settlement phases and agricultural land use.