Roots of Agriculture in South Arabia

University of Minnesota Archaeological Expedition in Yemen, February-March 1998

by This e-mail address is being protected from spambots. You need JavaScript enabled to view it

Yemen Update 41:1999


Figure 1. Sampling a 12m sediment profile in the Wadi Sana, Hadramawt.

With the support of AIYS through a NMERTA Project Grant, Dr. Joy McCorriston of University of Minnesota, led an interdisciplinary collaboration of scholars and students (Dr. Abdalaziz bin Aqil, Dr. Ingrid Hehmeyer, Zakariah Johnson, Russanne Low, Dr. Louise Martin, Abdal Basit Nama'an, Dr. Eric Oches, Dr. Pieter Vlag) to Hadramawt Province of southern Yemen. There the team conducted an archaeological survey and test excavations to explore early settlement and early agricultural strategies that took place during a dynamic period of global climatic change 9000-5000 years ago The team included geologists, archaeologists, a Yemeni ethnographer, and General Organization of Antiquities, Museums and Manuscripts representatives and spent several months in extremely remote regions of southern Yemen's arid plateau, today only occupied by Bedouin herders and their kin in isolated tiny farming oases. Without the generous logistical support of CANADIAN OCCIDENTALPETROLEUM LTD. (CANOXY-YEMEN) with its outstanding commitment to preserving natural and cultural resources in its concession area, it would have been nearly impossible for the RASA team to work in so remote a region. But the effort and persistence in exploring an archaeologically new area led to exciting new discoveries.


Figure 2. Extensive burnt surface exposed by recent erosion in the Wadi Sana. Such surfaces have been clearly burned in situ (as determined by magnetic profiles, reddened substrate, and grain size) and indicate much greater vegetation cover probably linked to more moist climate in the past.

The RASA team hoped to find evidence of the first peoples to reoccupy southern Arabia after the hyper arid phase of the last Ice Age. We think people moved or expanded into Arabia,which offered a new frontier when stronger seasonal monsoon winds blew greater rainfall over the subcontinent in the first half of the Holocene Little is known of these people--they carried with them stone tool technologies reminiscent of the eastern Mediterranean, but they occupied a land at the crux of the Indian Ocean. What else did they obtain from eastern Mediterranean lands, and what links did they have with peoples there? Did they also have links with East Africa or with South Asia? In all three regions, independent peoples domesticated different packages of crops and animals using different technologies Since domesticates from all these regions appear in later Arabia, the earliest occupants settled a frontier infarming&emdash;potentially they could mix and match from a wide array of domesticates. The particular mixes and timing of agricultural adoptions and adaptations in Arabia can tell us about the fundamental reasons people chose agriculture, whether because climates and environments are changing so that food procurement must also change,or because people are what they do, choosing to farm as their kin or trading partners do to express a sense of identity with them

With these interests, we headed for the southern Jol, the mountain plateau between southern Yemen's coast and the well-known Wadi Hadramawt of the interior. Our reasoning was thus: early Holocene prehistoric people have left almost no trace near the sites where Iron Age Hadramawt kingdoms--Raybun and Shabwa--sprang up. Yet such kingdoms could not arise without a deep cultural history of settlement, agriculture, and exchange. We sought the evidence for these societies in the uplands where water and landscape would once have been attractive for settlement but now are too arid to sustain more than a few scattered communities and mobile herders With few modern people around to disturb them,archaeological remains should be well-preserved.


Figure 3. Burnt surface in section. One of three stratified burnt surfaces in this profile, the visible surface as a dark horizontal line on the right has been radiocarbon dated earlier than 5880 BP (4775 BC).

In the Wadi Sanaa drainage system where the German Archaeological Institute and other travelers had already documented stone tools and monuments dating from 5000 to 2000 years ago, the RASA team set out to investigate a gravel stream bar densely strewn with 5000 year old stone tools and heavy blocks of stone. Our initial hopes that this might mark the remains of a settlement seem unfounded after several weeks of careful excavation. We recovered hundreds of tools and blanks of several raw materials from the surface but not a chip of worked chert below ground. The blocks of stone, some as tall as an adult and nearly a meter wide, had been dislodged through strong stream currents. But because lighter, fine, uneroded silts underlay the blocks, it seems more likely that humans,rather than water, carried blocks to the site. If ever a settlement had been, it was swept out and jumbled by wadi action before late Neolithic (so-called Habarut) people fashioned their stone tools on the site.

At this particular site, in Wadi Shumylya,the team was still compelled to explain the jumbled alignment and concentration of large stone blocks. As a working hypothesis, The RASA team believes we discovered what may be of the oldest evidence of agricultural technology in Arabia--a now nearly destroyed agricultural check dam. A layer of stone tools dating to about 5000years ago overlies the stream gravel bar that washed out and collected in a strong flood around the stone blocks we think we reused to build the check dam. Because the stone tools appear to be in the place they were dropped, we know that our putative dam (if we are correct in our interpretation of the stone blocks) and its destruction must be older than 5000 years. No similar agricultural technology is known from this date, but Tony Wilkinson and MacGuire Gibson's work suggests terraces in northern Yemen's highlands may also date back about 5000 years. Field silts from the desert margins also began to accrue around this period, suggesting that as people began to practice agriculture, they used different technologies indifferent environments.

The project objectives include investigation for evidence of early landscape management and environmental impact by humans. Agriculture--practiced by erecting shallow check dams to slow water flow and encourage silt deposit in gullies and stream beds--would play an important part in landscape management. Project geologists Drs. Rick Oches and Pieter Vlag in the field and Dr. Subir Banerjee of the Institute for Rock Magnetism devised other approaches that also promise insights into former environments and changes through time. The RASA team cleared, described and sampled 7 wadi sections in the mouth of Wadi Shumylya. By matching field and laboratory descriptions of grain size and sediments with profiles of various magnetic parameters (e.g., magnetic susceptibility) project geologists expect to develop sophisticated understanding of sediment transport and formation, perhaps even correlating sections widely spaced in the wadi system. Such studies offer an important tool for understanding the changeable wadi environment in which human occupation must have occurred. Only if we understand the environmental context can we make sense of our current puzzle--Where were human settlements?

The RASA team conducted an archaeological survey in the lower Wadi Shumylya, but we found only evidence of monuments and agricultural works. We would have to conclude from this season's results that people practiced agriculture without settlement, a proposition that runs counter to basic tenets of economic lifestyle elsewhere in prehistory. So to probe this enigma,we expanded our area of coverage to include the upstream (southern)tributaries of Wadi Idem, where we hoped to find settlements that we could compare with Wadi Shumylya. After several weeks, the team did finally find evidence for long-term (sedentary or semi-sedentary)human occupation.

The RASA team discovered an early settlement of more than 80 structures alongside deposits made by a now defunct set of springs. We excavated hearths and midden dumped into the construction of a terrace, and we hope to use the layout of houses,the debris within them, and especially wood and bone from hearths to understand how people organized themselves and their settlements,what they ate and how they procured it. We suspect this settlement at Shi'b Munayder is the earliest reported in Hadramawt (perhaps around7000 years old), but we await the results of C14 dating of charcoal from a hearth. Local Hadrami people and Yemeni archaeologists and are excited by the discovery of a settlement they recognize as

'complete in enough of the different kinds of structures and in such a state of preservation that...it will open wide horizons to the project in the future study of this kind of settlement. A detailed study from sufficient evidence of the domestic and economic life, the political organization, the architecture, the society, the people and the culture, [might someday offer the potential to] distinguish the kinship relations among [inhabitants] and...[to clarify the cultural sequence of the region].' (Dr. Abdalaziz Ja'afur bin Aqil, General Director of the General Organization of Antiquities and Museums in Hadramawt Province and member of the RASA team, speaking to the daily AL-AYAM in Mukalla, Yemen, April 1998)


Figure 4. Partially excavated Neolithic Pit House constructed and occupied between 5806-5616 BP (4689 BC-4456 BC).

Somewhat to our surprise, we turned up no indications in the course of fieldwork that the occupants of this settlement practiced agriculture. There were no clearly associated agricultural works nearby, no tools convincingly for reaping or hoeing, and no seeds. Thus we seem to have a paradox--Settlement without agriculture, Agriculture without settlement.

There is, of course, much work to be done on the samples brought back from this season and in planning how best to address the questions we raised. We borrowed for study many stone tools and brought sediment and charred plant fragments for analysis at the University of Minnesota. Anthropology students and graduate supervisors are working on the analysis of plant remains and stone tools in McCorriston's Archaebotany Laboratory, while in the Institute for Rock Magnetism Dr. Subir Banerjee and postdoctoral fellow Dr. Pieter Vlag are supervising magnetism measurements. Dates from radioactive carbon isotopes, amino acid racemization, and optical hermoluminesce will be processed elsewhere but will help enormously in interpreting our unique discoveries.


Figure 5. Mapping an ancient shruj of undetermined date in the mid-Wadi Sana.

RASA benefited from the support and generosity of people too numerous to mention here, but the project would like to thank them and the General Organization of Antiquities,Museums and Manuscripts, AIYS, CANOXY-YEMEN, The University of Minnesota, the National Science Foundation, and many individual staff of these organizations.


Figure 6. Disused check dam of undetermined date in a tributary to the mid-Wadi Sana.

Search Site

Search Library Collection