Social implications of water distribution in the ancient oasis of Ma’rib

1It is unnecessary to repeat here the obvious about the massive infrastructure of the so-called great dam of Ma’rib and its sluice gates, the obvious having been stated so many times before. Instead, I would like to draw attention to some implications that these physical constructions had for the farming communities. The dam was, of course, not a dam in the strict sense, designed to store water in an artificial lake. Rather, the dam’s purpose was to block the flood and direct flood waters immediately into the two main branch canals taking off on either side of it. Functionally speaking, the term “diversion barrage” is the more suitable designation for the Ma’rib dam.

2Looking at the technique of the dam’s construction, it can be suggested that the barrage was in fact designed to break. From the sophistication of the stone work contained in the north and south sluices I argue that the original builders could have built a more substantial dam, if they had so wanted. But the problem with a substantial dam is that if there is an exceptionally large flood, the rampaging water can cause havoc to the entire system.

3This problem was exacerbated by the accumulation of many metres of sediments as the sediment-rich flood waters dropped their loads onto the downstream fields during many centuries of irrigation. As a result, the diversion barrage had to be raised again and again, its main purpose now being to retain the water at a height that would still allow for it to be diverted onto the elevated fields under gravitational flow. But there is a danger inherent to retaining a large volume of water at an unnatural height, because the water will seek a way by cutting a path to a lower level.

4It is not surprising, therefore, that erosion gullies cutting through the deep sediment accumulations of the ancient field systems were a familiar phenomenon at least in later Sabaean times. Unable to cope with extremely large volumes of water, such as would be delivered by a “century flood,” the Sabaean irrigation engineers made extremely sophisticated provisions for the overflow of excessive water through spillways. In case of a severe flood, it would be preferable that the centre part of the main barrage gives way, leaving all the delicate distribution network intact. A breach may have been desirable from time to time for a second reason, namely to allow the sediment accumulated behind the barrage to wash away. Re-building the breach would be critical, but it would be a fairly simple earth-moving operation, not one of complicated engineering. The epigraphic evidence underlines the fact that, at least during the later Sabaean period, the dam had to be re-built regularly.

5From the beginning of the 4^th century CE until the beginning of the 7^th century CE, when occurred the final breach of the dam described in the Qur’an, inscriptions specifically mention five incidents of the dam breaking: at the beginning of the 4^th century CE, at the beginning of the second half of the 4^th century CE, in the year 449 CE, again in 450 CE, and around 544 CE. From differences, or discordances, in the cross-bedding of the sediments behind the remnants of the dam, Brunner concluded that a sixth breach occurred towards the end of the 6^th century CE.

6I would like to draw attention to the inscription CIH 541, which reports events in the year 544 CE. In actual fact, although the inscription mentions the dam breaking, the first third of the inscription deals with general social and political issues. The text refers to disturbances amongst tribes in the kingdom that forced the king to send a military contingent to settle them. Yet, as soon as the news spread that the dam in Ma’rib was broken, the inscription simply states that the king sent—literally—not an order, but a simple request, to the tribes to participate in the repair work. The numerous activities involved and the enormous supplies that were expended are then listed in detail. But no mention is made of the army going out to conscript the people in a forceful way, and it is pretty obvious that the tribesmen must have realized that their own livelihoods depended entirely on the repair works: repairing the breach would be critical to the next growing season. This was, no doubt, a major reason for their willing participation and shared responsibility.

7Communal cooperation is even more obvious when we look at the operation of the water distribution network. Distribution of the water to the individual fields was accomplished by way of branch canals and a network of smaller feeder channels. Because of the extremely limited duration of any given spate, the trapping, diversion, and distribution of the flood water had to be completed as quickly and as efficiently as possible so that the largest possible extent of the cultivatable area could be irrigated. Sub-areas of the oasis, consisting of a number of individual fields, were thus irrigated as one unit by operating a distribution device along the canal branches. This system presupposes both a consensus on water distribution and the compliance of individual farmers with the agreement.

8It may seem trivial to ask: who opened what sluice gate and when, and why did nobody cheat and take more than that to which they were entitled? All too easily an individual could have come in the middle of the night and manipulated the field walls surrounding his property in order to receive an additional quantity of water. But individual disregard of the irrigation schedule would have had drastic consequences for the entire community: the disruption of reduced efficiency, overall loss of irrigated farmland, and quarrelling would result in a breakdown of the entire system.

9Farmers must have understood the danger of this principle, that in the end the benefit of any short-sighted individual advantage would be outweighed by the disadvantage caused to the community. Communal responsibility and cooperation in fact form the fundamental secret to the long-term effectiveness of the system. Without group responsibility and cooperation, there would simply be a remarkably impressive engineering scheme as useless as some of the grandiose development schemes that the modern world has observed. Ma’rib would not have worked the way it did for so many centuries without the cooperation of its users. And the system was far too complex to work on the basis of force superimposed by the state.

10When I published my dissertation dealing with the practical aspects of irrigation farming in the ancient oasis of Ma’rib, as volume V of the German Archaeological Institute reports, I did not hold this position. I relied more on a model which envisaged strong state control in the ordering, maintenance, and operation of the system. I have modified my position since then and placed greater emphasis on the need for a community’s willing participation because of my subsequent involvement in the Canadian Archaeological Mission’s Zabīd programme, which takes us into the Islamic period.

Water allocation rights in the mediaeval Wādī Zabīd

11In the case of Zabīd, the existence of numerous texts and the greater breadth of Islamic historiography made it much easier to extract nuances about social interaction from the written sources and to match them with the reality of the physical systems. In comparison, for the Sabaean period only few epigraphers take the information contained in the inscriptions one step further, into detailed social history.

12For the Wādī Zabīd one can find plenty of citations that refer directly to or imply the role of negotiation in matters of water distribution. I wish to present therefore an overview of the critical points for what we know about the possible beginnings of this practice.

13The town of Zabīd was founded in 820 CE. The earliest physical evidence for engineered irrigation systems in the Wādī Zabīd stems from the 11^th to 12^th centuries. One may, however, assume that in order for the city to be fed, the Wādī Zabīd was already being controlled for irrigated agriculture to a certain degree at least from the 9^th century onwards.

14Until recently, the earliest reference to the allocation rights of the water was connected to the name of Shaykh Isma^cil al-Jabarti to whom oral tradition gives the credit for codifying the law before 1473, the year of his death. Serjeant already raised the possibility that al-Jabarti was only arbitrating, using rather than formulating the law.

15In a recent MA thesis by N. H. Salameh, a new manuscript on the subject has been brought to light which gives us the priceless information that the unique Wādī Zabīd water law was codified in response to violent disputes among the farmers of the Wādī Zabīd over water towards the end of the 14^th century CE. These disputes happened during Qadī Muwaffaq al-Dīn ^cAli ibn Abī Bakr al-Nashirī’s term in office as the qadi of Zabīd (1391 to at least 1400) in the reign of the seventh Rasulid sultan, al-Malik al-Ashraf Isma^cil bin al-^cAbbās (1377-1400). The new water law replaced the allocation of spate water according to the Shari^ca principle of upstream priority. Instead the new law followed the principles of customary law (^curf) and allocated the water by prescribed calendar dates.

16The 17 canals irrigating the fields in the Wādī Zabīd were classified into three groups. Group I, consisting of six individual canals that irrigated the upper field systems, had the right to use the water between 19 October and 2 August. The seven canals of Group II, feeding the middle field systems, received the water in the system from 3 August to 13 September, while the four canals of Group III had the water allocation for the low-lying field systems between 14 September and 18 October. Apparently it was felt that in this way, with the relative unpredictability of the floods, the risks as well as the bounty were spread equitably across the entire system.

17This was, however, not the end of the disputes among the farmers of the Wādī Zabīd about flood water allocation. The premise that the system functioned smoothly from then on is false. While one might conclude that the documented legal disputes and the constant need for arbitration are evidence for an absence of community consensus, I take it that the very fact that these cases were debated reflects in essence what we would call a democratic way of balancing individual and communal rights. It reflects established legal and socially accepted practice. The face of anarchy would look totally different. In addition, this allocation right has been effective more or less unchanged until the present. Like the ancient oasis in Ma’rib, the Wādī Zabīd is a man-made landscape which bears the marked imprint of many centuries of successful irrigation farming.

18These thoughts have been reinforced by my involvement in studies conducted for international development agencies in Yemen.

Present developments

19Yemen’s cultivatable lands and rural population have been dependent for generations on scarce and unreliable rainfall. As a result, Yemeni farmers have developed great skills in water harvesting, water storage, and spate irrigation that I described before. Hand in hand with these skills went an excellent understanding of the ecological imperative of soil and water management for sustainable basic food production.

20Rural communities are spread throughout the thousands of minor catchments linked to the major wādī systems throughout the country. Thus tens of thousands of local communities, in qurā (sing: qarya: villages), mamāsī (sing. mamsā or mamsiyah, groups of villages under the non-formal authority of a ^cāqal), and ^cuzal (sing. ^cuzla: clusters of mamāsī linked within tribal networks, usually represented by a shaykh), have an immediate impact upon the greater wādī systems that tend to be the focus of studies on Yemen’s irrigation systems.

21Therefore, throughout both the major arteries and the branch areas there existed networks of group management and shared task allocations between men and women farmers, their village communities, and the tribal networks, all welded together in a shared ‘association of interests’ known as ta^cawwun—the local necessity for “a spirit of cooperation.” It was this non-formal, shared responsibility which formed the basis for the success of sustainable resource management in Yemen.

22Many of the problems visible today throughout the country are related to a major ecological collapse. Beginning in the early 1970's, extensive deforestation, abandonment and subsequent collapse of terracing systems, and overgrazing have increased surface run-off and triggered massive soil erosion. A general neglect of flood diversion structures in the wādīs and the increasing over-extraction of ground water resources added to the problem. Sustainable natural resource management, traditionally driven by practical community-based initiatives, had been abandoned for various reasons.

23In the former north Yemen, a major reason was the exodus of large segments of the male population, especially to Saudi Arabia and the Gulf, which left whole villages without a work force. In many villages the remaining community consisted of women, children and the elderly. Natural resource management is men’s work; their absence meant that the traditional rural networks responsible for the practical initiatives collapsed, and that significant maintenance operations did not get done. In the past the village women contributed to this work through their own activities, such as the important task of culling fire wood in a way that regular re-growth of the trees is guaranteed. Today, two factors play against this. The first is that due to increasing population growth more and more fire wood needs to be collected, forcing the women to go farther and farther afield into the marginal areas of their neighborhoods, and being more ruthless in their practices. The second factor is increasing road building activity and access for four-wheel drive vehicles which allow greater volumes of wood to be cut for commercial use. In both cases basic economic pressure forced communities to ignore traditional safeguards for the protection of their sustainable resources, and to adopt more selfish measures instead.

24In addition, the remittances from the men working abroad meant that their home communities were no longer completely dependent upon their own productivity for survival, a situation that in turn reinforced the tendency to make maintenance of the existing land-use systems less essential. For many years, returnees had invested capital in cash generating schemes, such as installing pumps or generators. This trend was exacerbated in the 1990 Gulf War crisis when the need to reap a quick return on investment was perceived. Short-sighted individual activities such as well-drilling were multiplied, thereby contributing to the dramatic over-exploitation of the water resources.

25In order to address and tackle the pressures of these inter-related ecological and agricultural problems of the present, a way must be found to revive the spirit of communal responsibility and cooperation. This was the starting point of a consultancy field mission in which I participated for the German Agency for Technical Cooperation (GTZ) in 1996 in the Ibb and Ta‘izz Governorates. The amazing result was that there actually is a new movement of cooperation visible in Yemen’s agricultural communities, which may be partly due to the farmers realizing that the individualistic approach has proven unsuccessful for the majority of farmers. These new associations of interest need encouragement, training in some practical and administrative skills, and attending professional advice where needed.

26I have therefore great confidence that if a productive relationship with funding donors can be established, then the crisis of ecological degradation can be reversed, and we would have an opportunity to modernize rural society around natural resource renewal.