ORE Open Research Exeter

TITLE The Wadi Faynan Project, Southern : a Preliminary Report on Geomorphology and Landscape Archaeology

AUTHORS Barker, G.W; Creighton, O.H; Gilbertson, D.D; et al.

JOURNAL Levant

DEPOSITED IN ORE 28 May 2008

This version available at

http://hdl.handle.net/10036/28552

COPYRIGHT AND REUSE

Open Research Exeter makes this work available in accordance with publisher policies.

A NOTE ON VERSIONS

The version presented here may differ from the published version. If citing, you are advised to consult the published version for pagination, volume/issue and date of publication -fhe G.S7. BARKER ET AL. Wadi Favnan Proiect.Southcrn Tordan 33 best understood at present is the impact of past met- study area to yield an equally complex history of alliferous mining, where both primary waste materi- settlement and land use. als and slagscan be seento have covered large areas In the Late Pleistocenea period of erosion in the of the landscape and to have entered into the river mountains and deposition at lower elevationsled to systems,where they now produce distinctive copper- the accumulation of alluvial fans and fluvial deposits rich terrace gravels. The primary constraints on nearly forty metres thick (which we have termed progressare the absenceof reliable knowledge of the the Ghuwayr Beds) and at a later date by further antiquity of the deposits present and the absenceof fluvial deposits (the Shayqar Beds). The evidence similar studies from adiacent areas with which com- for human occupation at this time recovered by the parisons can be drawn. Future fieldwork must widen team consists of sporadic chipped stone artefacts, the focus of study to establish whether or not the mostly of middle palaeolithic type, found within and sequence established this year holds true further on top of thesetwo setsof deposits.However, recon- afield, and seekto establishan independent absolute naissancesurvey in April 1996 by Bill Finlayson and chronology for the regional picture that emerges. Steve Mithen immediately prior to our own cam- Our preliminary study of the nature and distribu- paign discovered an epipalaeolithic (final palae- tion of the Quaternary deposits present in the \fadi olithic) site (WF16) consisting of flints, quernstones Faynan and its tributaries has revealed a complex and other groundstone artefacts on the interfluve history of events from c.150,000 years ago to the between the Wadis Ghuwayr and Shayqar, possibly present, embracing environmental changes of great associatedwith circular and semi-circular platforms, magnitude, and our study of the archaeology of the and other smaller collections halfway up the $7adi field system, along with the investigations of other Dana and on the plateau edge (Finlayson and archaeologicalteams, indicates the potential of the Mithen 1996). The Epipalaeolithic in the Near East

Figure 12. Looking northeasttowards the herring-bonefields in WF4.3, trapping oaerland water and gully flow and leading this water into thefields below. (Photograph: G. Barker). 34 LEVANTxxrx leeT is of course of critical interest as a key transitional stage to the emergence of agricultural communities, being characterized by social elaboration, the spe- cialized exploitation of plant foods and herd ani- mals, and a degree of sedentism especiallyat lower elevations, with upland sites usually intrepreted as task-specific sites related to residential sites in the lowlands (Henry 1989; Kaufman 1992;Valla 1995). \(rhilst detailed information on the narure of epi- palaeolithic settlement in the study area will only come from the detailed investigation of these sites that is planned, it is clear that a fully-fledged agri- *^ cultural community was established here early in the Holocene, probably by the eighth millennium e.c. ": 'tVadi -a Ghuwayr 1', the PPN site being excavatedby - \o ! o\ the Department of Antiquities, is of a date o\ Jordanian T range similar to the well known village of Beidha a d q) *s few hours walk away on the plateau near Petra (nE (Kirkbride 1966), and has a similar range of struc- ,3 q tures and artefacts. The settlement / R was located at ^s the point where the $(zadi Ghuwayr egressesfrom a narrow defile to form a significant floodplain a kilo- metre up-wadi from the confluence J< (Fig. 4), immediately by the spring that now irrigates S the market gardens of the Faynan. S7hilst the botan- I N ical and faunal data from the site are currently q under l study, the spring-side location N is typical of many a I R early farming settlements i in the Near East, suggest- -, l S M ing a reliance on naturally-irrigated land for crop 6- 6 farming and animal grazing (Bar-Yosef 1995). s Through the Holocene, the environment of the study area has been dominated by a seriesof mainly \)6 fluvial events: rivers have U4 incised their floors, erod- !4 ing the Ghuwayr and Shayqar Beds, as well as orher \-r pre-existing surficial deposits and bedrocks. Traces R of these rivers exist up-wadi, in the form of terrace N deposits of sediments and erosion surfaces, some- a times displaying significant cementation and cal- a.i' cretization, whilst in the confluence area, distinctive \ layers of fluvially-deposited boulders gravels R and cl) E were laid down. The discovery within the Upper q) Faynan Beds of the presence of a more-or-less (|- o' perennial stream adjacent to Tell Wadi Faynan, and associatedwith the site's occupation in the sixth and TI a fifth millennia 8.c., presents q' a strikingly different 'Fa picture from today, of a relatively rich and diverse q) aquatic landscape. Interestingly, the pottery and z.--I- mortar from the site contained a mixture of reeds q) and grassas well as straw (Najjar et al. l990r 4l) R

The excavations at Tell \7adi Faynan found traces FS of simple rectangular dry-stone buildings associated l-r with storage pits, sickle blades, carbonized cereals (r1 and a faunal sample dominated c) by the bones of L goat. sheep or We found traces of what may be a b0 boulder wall in the river section fifty metres or so t --

G.\7. BARKER ET AL. The Wadi Faynan Project,Southern Jordan 35

east of the settlement, within the same sedimentary 1972). The social and economic conrext of these horizon as the Tell Wadi Faynan settlement, so an transformations, for example the nature of the con- important question to try to resolve in future years trol of the ores and of the trade in their products, will be whether this community was growing its and of the impact of the expansion of copper mining crops on moist backswamp soils, as Sherratt (1980) on local economic and social organization, are much for example has argued for neolithic settlements in debated (Finkelstein 1995; Joffe Lgg3), and form basin-edge locations such as Qatal Htiyrik in Turkey the principal research focus for Dr $Tright's investi- and Nea Nikomedeia in Greece, or whether people gation of site \UfF100. in the \7adi Faynan were now beginning to experi- Whilst her initial fieldwork at the site is being ment with systems of floodwater farming using reported separately, the principal result was an primitive diversion walls. appreciation of the remarkable size of the settle- By the later fifth millennium B.c. the Tell \7adi ment: the boundary walls of \7F4.13 measure Faynan community was extracting malachite copper approximately 400 by 200 metres) and the area ore on a small scale from the local sandstones,and within is carpeted with chalcolithic and early bronze smelting it on site, creating quantities of iron-silicate age pottery, together with other diagnostic artefacts 'Canaanite' slag (Najjar et al. l99O). This development accords of the periods such as blades and frag- with the evidence from the region as a whole, that ments of basalt bowls. Some of the enclosure walls the systematic exploitation of copper ores began in are certainly later, presumably of Nabataean/ the Chalcolithic and then gready expanded in the Roman/Byzantine date, but others, including sec- third millennium B.c. during the Early Bronze Ag., tions with turret- or bastion-like structures attached accompanied by technological improvements allow- to them somewhat akin to the towers at contempo- ing the large-scale use of low grade ores (Adams rary settlements such as Arad (Amiran et al. 1978), 1991; Adams and Genz L9951'Hauptmann 1989 may also be associatedwith the early bronze age and 1992; Hauptmann et al. 1992; Rothenberg settlement. As noted earlier, habitation traces of

* F[1:

Figure 14. The water catchment st/ucture WF24, on the basisof surface material likely to be of chalcolithic and/or early bronze age date. (Photograph: G. Barker) 36 LEVANT xxrx 1997

these critical periods were also noted elsewhere N within the field system. ' b. 5r- i $(rhilst our understanding of the land use systems c.i that supported the chalcolithic and bronze age \o settlement transformations in the study area will o\ o\ certainly require excavationof sites such as WFl00, it is noteworthy that our survey identified a series of *-/ --^-xYQAR +f ":a) circular or oval silt-filled structures on the low-angle ,.o,6#: fans along the southern edge of the floodplain that q) R may have been used for water catchment and/or U. storage at this time. WF24, for example, located within $fF4.11, consisted of a circular depression some fifteen metres in diameter bordered by a rough S,) dry-stone wall (Fig. l4). Immediately to its south, q)a q) upslope, was a single boulder wall (5059) con- N structed along the slope to lead any surface runoff a into a small (three by two metre) rectangular srruc- a. ture (5059) and thence into the main catchment. a, The pottery around these structures was entirely 'i hand-made, the shapes and decoration broadly indicative of a chalcolithic/early bronze age date, and >' was associated with other typical finds N' such as the & fragment of a pierced stone weight, a tabular flint \ R scraper, a sickle blade and borers. The silts were cored, and are currently \) being analyzed. Other sim- "s ilar catchments were noted to the south of \7F4.8 \ (lil7F39) and of $fF4.18 (\ilfF4O). Although direct

q) information regarding chalcolithic and early bronze v, age cultivation systemsin the region is still extremely \-r .i poor, there are some archaeological indications from q) the Negev that the settlement transformations at this .s time were associatedwith the development of simple floodwater farming systems (Cohen and Dever ,.(_\ q) 1980, s-a ^\ 42; Levy 1995, 230), and the thesis is sup- ported now by phyolith evidence in cereal crops (- o\ q-., (Rosen 1992). \\ ho \B\\'.-, Our initial study of the field system has demon- ...---- l\ strated the likely longevity and complexity of its his- tory, with indications of piecemeal development of .\ the main suite of walls. This hypothesis is \l supported -- --'.11 v% ' 2 qn q) . z by the occurrence of a number of Nabataean- ( (- t'; Roman farmsteads on the slopes immediately south ? SF i.x of the field system (such as til7F35 and WF36), and *bn 6 vA' i'Tt, I O.S burial cairns of similar ages around these farms and \l \,\ ., * sd within \V \B the enclosed areas, and also by the variability ),\ sPts in the distributions of Nabataean, Roman and { i, r\ Slr Byzantine pottery indicated by our initial studies of \.-| .>-., \ s$ the abundant surface material. The fact that water ( SS catchment was only one element in the system also ,'l) 3: c-r I emphasizes the fact that we are not dealing with a \n cn u*L unitary system laid out to some grand design, as 'R (r ,, t) does the relationship of the field system ro rhe Dana doS Beds, which in pan seem to be contemporary with IJlP the construction and use of fields (as with the con- G.W. BARKER ET AL. The Wadi Faynan Project, SouthernJordan 37 struction of the aqueduct) but elsewhere clearly taken from the sedimentary sequencesexamined by post-date them. Evidence for re-plastering the con- the project's geomorphologistswill provide a further duits to the Khirbat Faynan water mill, with the indicator of the changing scalesof land use and min- upper level consisting of an opus signinum and tJne ing activity through these periods. Certainly the use lower level of a coarse lime plaster, and of re-plas- of the field system seems to have been associated tering the inlet conduit to the reservoir, is another with the widespread deposition of loamy depositsby indicator of the likelihood of a development wind and water, whilst its eventual abandonment sequenceof water management associatedwith the may in part be related to the lowering of the water field system (Appendix). table as the river incised itself into the wadi floor In this respect, it is noteworthy that the scale of after the deposition of the Dana Beds. mining activity in the Wadi Faynan through the Beyond the field system, it is clear from the little Nabataean, Roman and Byzantine periods also reconnaissancesurvey that we undertook, for exam- seems to have been highly variable through time: ple on the upper slopes on either side of the Wadi charcoal analysis from the major smelting site of Shayqar, that there is a rich suite of archaeological Khirbat an-Nahas indicated that in the earlier cen- remains including lithic scatters, terrace walls, turies of the first millennium B.c. the activity was domestic structures and graves,as well as numerous relatively small scale,with little impact on the envi- examples of boulders decorated with pictographs ronment - people were able to gather sufficient fuel with human, animal and abstract motifs. A number wood for the smelting within walking distance of the of structures containing a room or rooms and a pen site without significant damage to the local ecology or enclosure were located, some with Roman and (Engel 1993). The Roman period, however, was Byzantine pottery, others clearly very recent in date, characterizedby a massiveexpansion in exploitation and we also found a number of rent footings of (Hauptmann 1992). It is hoped that the samples (recently?) abandoned Bedouin camps. Classifuing

'l- jd.:. F-*. . i;$ t# F{-'-r "*. [-:; *;d, $i

Figure 16. A modent irrigation resen:oir in the Wadi Faynan. The water is piped doutnfrom the Wadi Ghuwayr spring (on the right, at the foot of the hilk in the distance) into such reserz;oirs,and then pumped onto the surrounding fields. (Photograph: G. Barker). 38 LEVANT xxrx leeT and dating this archaeology is likely to be extremely over A2, a flat coarse white lime plaster with gravel and difficult, but the potential clearly exists in the traces of grass or straw and chalk or lime lumps, 22 mll- archaeology record outside the field system, as limetres thick. The flat interface suggests that the pink within it, to document changing patterns of arable plaster is a secondary phase. The frred clay is probably and pastoral activity, and the extent to which they crushed tile, brick, or pottery. B. Coarse white lime mortar with gravel were integrated or separated in terms of social and charcoal, about 28 millimetres thick. organization, from the time of the first agricultural C. Roughly faced white lime mortar or plaster, 22 milli- settlement by the $(radi Ghuwayr spring to the pre- metres thick, with gravel, chalk, charcoal, and traces of an sent day. over-layer of crushed tile (?) plaster. The recent introduction of irrigated agriculture Chemical separation with dilute hydrochloric acid left for water-dependent crops such as tomatoes, water residues which were identified and graded. They were melons, courgettes and cucumbers poses a very real composed of basalt-type rocks with quartzite, qlJartz, threat to the remarkable archaeological landscapes sandstone and some flint or chert. Sample A1 had an of the \fadi Faynan. At the present momenr there upper layer composed almost entirely of crushed ceramic are thirteen water reservoirs in use or near comple- material with some quartz sand which may also have come from the ceramic. The charcoal probably tion (almost all built in the past rwo years), and the was a lime- burning residue. The table below summarizes the particle cultivated area covers about a third of the field size distribution analysis. system (Figs. 15 and 16). The area taken back into cultivation is thus growing rapidly, sometimes pre- Table 2. Wadi Faynan mortar/plaster samples:particle size ceded by the destruction of the extant ancient walls distribution analy sis. 'graael' 'send' 'sih' and other structures, sometimes by bulldozing sim- Sample soluble %ocomments ply the surface material. Even ploughing of the fields >2 mm. <0.15 2 mm. <0.15mm. has an impact on the archaeology, however, remix- A13633469 upper) ing and reburying the surface artefacts. It is clear bricVtile A2 83 l2 5 46 lower that within a few more years there will have been a 867 25 8 43 significant loss of information on the ancient settle- c55 30 l5 48 ment and farming systems of the \7adi Faynan. For this reason) the intensive field-walking and struc- The soluble value approximates to the lime content, but tural survey of the field system area have to be the here it may be slightly high due to the presence of chalk or main priority for future work, with the study of the limestone, although this may in fact be unburnt lime. The results show that the coarse abundant archaeological remains outside the sys- mortar layers are fairly similar in composition, although the parricle size grades tem - themselves also under threat especially from are not. This may be partly due to the very small sample tomb robbing - the second urgent priority. size and numbers. The very high lime contenr for the problems S(rhilst the of modern destruction are ceramic-based material is of interest, though. It does profound, the scale of the archaeological record appear to be very similar to classic Roman opus signinum immense, and chronological resolution likely to be brick- or tile-based mortar used for water-resisting struc- especially daunting, there is every reason to expect tures. This also had a higher lime content than ordinary that the continuation of a carefully integrated pro- mortars and plasters. gramme of geo-archaeological fieldwork in the $7adi Faynan will provide a complex and detailed land- Acknowledgernents scape history that will add considerably to our understanding of the human history ofJordan and of Thanks are especially due to: the institutions that funded the region as a whole. the fieldwork - the Humanities Research Board of the British Academy, the University of Leicester (Research Appendix: mortar/plaster samples frorn the Wadi Fund) and the British Institute at Amman for Archaeology Faynan (GCM) and History; Alison McQuitty (Director), Pam \il7atson (Assistant Director) and Caroline Middleton (Secretary) Samplesof mortar/plasterwere collectedin 1995 from of BIAAH for invaluable advice, encouragement, and three locationsin the Wadi Faynan:the inlet channelinto practical support; Dr Karen Vright (and her team: Aiex the watermill (sampleAl, upper;sample A2, lower);the Wasse, Ariane Marcar, Gail MacKinnon) for much- outlet channelfrom the reservoir(B); and the inlet chan- appreciated practical help and wisdom; Dr Ghazi Bisheh, nel to the reservoir(C). The sampleswere analyzedmicro- Director-General of Antiquities, for his help and support; scopicallyand chemicallyto determinetheir structureand Dr Mohammad Najjar, excavator of the Tell ril7adi composition. They were all lime-based, with varying Faynan, for his generosity in encouraging the geomorpho- aggregates,and canbe describedas follows: logical and palaeoecological sampling at the site; our rep- Al. Pink fired clay and lime plasrer,7-10 millimetres, resentative from the Department of Antiquities Mr. Emad --

G.W. BARKER ET AL. The Wadi Faynan Project, SouthernJordan 39 ad-Drooz; the many local people who made us so wel- Gilbertson, D. D. (1996) Explanations - environment as come, and who provided much invaluable information, agency. Pp. 291-317 in G. Barker, D.D. Gilbertson, including Abu Fawaz, Abu Khalil, Abu Mustafa, G.D.B. Jones, and D.J. Mattingly Farming the Desert:the Mohammed Rashayda and Sayl Rashayda; and last but in UNESCO Libyan Valleys Archaeological Suruey: Volume no sense least, Aladdin Madi, a wonderful cook respons- One - Synthesis. LTNESCO: Paris. Department of ible in no small part for the efficiency and commitment of Antiquities: Tripoli, and Sociery for Libyan Studies: the field team. 'We London. would like to thank Debbie Miles for preparing the Gilbertson, D. D. and Hunt, C.O. (1996) Romano- figures. Libyan agriculture: walls and floodwater farming. Pp.l9L-225 in G. Barker, D.D. Gilbertson, G.D.B. Jones, and D.J. Mattingly Farming the Desert: the Bibliography UNE'SCO Libyan Valleys Archaeological Suruey: Volume One - Synthesis. I-INESCO: Paris. Department of Antiquities: Tripoli, and Society for Libyan Studies: Adams, R. (1991) The Wadi Fidan project, Jordan, 1989. Leztant33, 181-183. London. Glueck, (1935) Adams, R. and Genz, H. (1995) Excavations ar Wadi N. Explorations in Eastern Palestine II. American Fidan 4: a chalcolithic village complex in the copper ore Schools of Oriental Research: New Haven. CT. district of Feinan, sourhern Jordan. PEQ,8-20. Amiran, R.C., Paron, U., Shiloh, Y., Brown, R., Tsafrir, Hauptmann, A. (1989) The earliest periods of copper Y. and Ben-Tor, A. (1978) Early Arad: The Chalcolithic metallurgy in Feinan, Jordan. Pp. 119-136 in A. Settlement and Early Bronze City I: First-Fifth Seasonsof Hauptmann, E. Pernicka and G.A. \Tagner (eds) Old Excavations, 1962-1966. Israel Exploration Sociery: World Archaeometallurgt. Bochum, Deutsches Bergbau- Museum, . Der Anschnitt, Beiheft 7. - (1992) Feinan/S7adi Barker, G., Gilbertson, D.D., Jones, G.D.B. and Feinan. AJA 96, 510-512. Mattingly, D.J. (1996) Farming the Desert:the UNESCO Hauptmann, A., Begemann, F., Heitkemper, E., Libyan Valleys Archaeological Surey: Volume One - Pernicka, E. and Schmitt-Strecker, S. (1992) Early Synthesis; Volume Two - the Site Gazetteer and Pottery. copper produced at Feinan, Wadi Araba, Jordan: the IINESCO: Paris. Department of Antiquities: Tripoli, composition of ores and copper.Archeomaterials 6,l-33. and Society for Libyan Studies: London. Henry, D. (1989) From Foraging to Agricubure: the Leztant Barnes, FI., Flender, M., Ruben, I., Shafiq, R. and at the End of the Ice Age. University of Pennsylvania McQuitty, A. (1995) The Wadi Feynan Project,Phase 3. Press:Philadelphia. Amman, British Institute at Amman for Archaeology Joffe, A.H. (1993) Settlement and Society in the Early and History, unpublished reporr. Bronze Age I and II, Southern Leztant. Monographs in Bar-Yosef, O. (1995) Earliest food producers - pre- Mediterranean Archaeology 4. Sheffield Academic pottery neolithic (8000-5500). Pp. 190-204 in T. Levy Press:Sheffield. (ed.) The Archaeologjt of Society in the Holy Land. Kauf'man, D. (1992) Hunter-gatherers of the Levantine Leicester Universiry Press:London. Epipalaeolithic: the sociological origins of sedentism. Beaumont, P. (1989) Drylands: Enztironmental Journal of Meditenanean Archaeologt 5, 165-201. Management and Deuelopment.Routledge: London. Ktg, G.D. (1989) Feinan. Pp. 225-227 in D. Homes- Cohen, R. and Dever, \Ufl.G.(1980) Preliminary report of Fredericq and J.B. Hennessy (eds) Archaeologt of Jordan 'Central the second season of the Negev Highlands il.l. Field Reports. Surueys and Sites A-K. Akkadica Project'. BASOR 236, 4l-49. Supplementum 7. Leuven. Engel, T. (1993) Charcoal remains from an iron age Kirkbride, D. (1966) Five seasons at the pre-potrery copper smelting slag heap at Feinan, Wadi neolithic village of Beidha in Jordan. PEQ 98, 5-61. flordan). Vegetation History and Archaeobotany 2, Lagrange, M.-J. (1898) Phounon. RB 7, lL2-115. 205-2tt. Marco, S., Agnon, A., Stein, M. and Ron, H. (1994) A Finkelstein, I. (1995) Lizting on the Finge: The Archaeologt 50,000 year continuous record of earthquakes and sur- and History of the Negex, Sinai and Neighbouring Regions face ruptures in the Lisan Formation, the in the Bronze and lron Ages. Monographs in Graben. U.S. Geological Survey Open File Report Mediterranean Archaeology 6. Sheffield Academic 94-568, tt2-t14. Press: Sheffreld. Millington, A.C. and Pye, K. (1994) (ed.) Enaironmental Finlayson, B. and Baird, D. (1995) The Dana Change in Drylands: Biogeographical and Geomorphologi- Archaeological Surztey, Phase 1 Report, Jordan. cal Perspectiztes.\Y lley: Chichester. Unpublished report, Centre for Field Archaeology, Musil, A. (1907) Arabia PetraeaII. . Alfred Holder: Edinburgh. Vienna. Finlayson, B. and Mithen, S. (1996) The Dana-Faynan Najjar, M., Abu Dayyeh, A. es-S., Suleiman, E., (south Jordan) Epipalaeolithic Project: report on reconnais- Weisgerber, G. and Hauptmanfl, A. (1990) Tell \7adi sance suruey, 14-20 April 1996. Unpublished report, Feinan: a new Pottery Neolithic tell in sourhern Jordan. British Institute at Amman for Archaeology and ADAJ 34,27-56. Historv. Amman. Quennell, A.M. (1984) The western African Rift System. .-

40 LEVANT xxlx 1997

Pp.775-88 in J.E. Dixon and A.H.F. Robertson (eds) Ethnographiques.CNRS, Monographie du CRA No.6. The Geological Evolution of the Eastern Mediterranean. Paris. Geological Society of London, Special Publication 17. Rothenberg,B. (1972) Timna: Valley of the Biblical Copper London. Mines. Thames and Hudson: London. Rabba'a, I. (L994) Geologt of the Al Qurayqira (Jabal Sherratt, A.G. (1980) Water, soil and seasonalityin early Hamra Faddan), Map Sheer No. 3051 il. Amman, cereal cultivation. World Archaeologt lI, 313-330. Hashemite Kingdom of Jordan, Ministry of Energy and Thomas, S.G. and Middleton, N.J. 0994) Desertification: Natural Resources, Natural Resources Authority, Exploding the Myth. John Wiley: Chichester. Geology Directorate, Geological Mapping Division, Valla, F. (1995) The first settled societies-Natufian Bulletin 28. (12,500-10,200 BP). Pp.169-187 in T. Levy (ed.) The Rosen, S.A. (1992) Phytoliths as indicators of ancient irri- Archaeologt of Society in the Holy Land. Leicester gation farming. Pp. 28L-287 in Prdhistoire de University Press: London. I'Agiculture: Nouaelles Approaches Experimentales et