Published as Chapter 22 of The Social and Environmental Effects of Large Dams: Volume 1. Overview. Wadebridge Ecological Centre, Worthyvale Manor Camelford, Cornwall PL32 9TT, UK, 1984. By Edward Goldsmith and Nicholas Hildyard.
The Sonjo, a North Tanzanian tribe with a population of some 4,500, have practised irrigation agriculture from time immemorial. Indeed, their whole way of life – from their world-view to the social and cultural organisation – reflects their preoccupation with irrigation. Thus, Robert Gray who made a detailed study of Kheri, a Sonjo village, in his book, The Sonjo of Tanganyika writes that “Irrigation is a central theme in Sonjo religion as well as in agriculture”. 
Rainfall, for instance, is thought of as “the overflow from a celestial irrigation system” and the tribe’s formal prayer for rain is “stated as a hydraulic metaphor – ‘Open the sluices of Belwa’”; Belwa being the mythical name of the Sonjo.  So, too, the Sonjo see natural springs as sacred objects.
“The springs are treated as sanctuaries and may not be approached by ordinary people who are involved in sexual activities and the passions of everyday life . . . Thus, they are effectively insulated against contact with the profane world, which might contaminate and weaken their sanctity and introduce disastrous change.” 
For the Sonjo, proof of the sacredness of the springs lies in the very predictability of their flow: they have never been known to fail and are thus “conceived of in a different category from unpredictable blessings such as rain or health”. Should the springs fail, however, the results “would be more catastrophic than in the case of drought”.
Quite sensibly, therefore, “the basic policy with regard to springs is directed at maintaining the unchanging state of affairs that has prevailed in the past”. In fact, one could argue that it is only because the springs are sanctified that they remain productive. In a more materialistic world – Texas for example – they would undoubtedly have been run dry.
With regard to rain, the Sonjo believe that so long as the appropriate rituals are performed – and the traditional moral code is upheld – then their God, Khambageu, will protect them. If rain does not fall it is assumed that same basic rule of conduct has been violated – and that Khambageu has thus been angered.
The Sonjo cultivate three types of land which they refer to as ‘hura’, ‘magare’ and ‘isirene’ land:
The hura land, which is the most important, is situated in the flat alluvial river valley. Every year, during the rainy season, it is flooded by the river, which deposits on it a thick layer of silt. During the dry season, however, it must be cultivated entirely by irrigation, the water being derived from springs and from the river itself. There is clearly no fertility problem since the land is fertilised annually by the silt deposited on it by the river, nor is there likely to be a salinity problem since the salts are washed out by the floodwaters.
The magare lands are located on the sloping lands above the water, which makes them very vulnerable to erosion. Moreover, since no silt is deposited on them by the river, their continued fertility is not assured by natural means. In addition, they are more prone to salinity since the salt in the soil is not washed out by the annual floodwaters.
The Sonjo have adapted to these ecological constraints by dividing their magare land into two different categories which are cultivated on alternate years – just as did the inhabitants of the Euphrates River. The exact amount of magare land actually under irrigation at any one time depends on the amount of water available for irrigating it. Tenure of the magare fields gives certain rights to irrigation water, since without the water the plots would be worthless. In a very bad year – when there is very little rain, or when the rain falls at the wrong time – the magare land may produce no crops at all, but the hura fields will always produce enough food to avert famine.
The isirene land is situated on several natural terraces down-stream from the hura valley. It consists, in all, of thirty acres. A very elaborate irrigation system is required to cultivate the isirene land because the different elevations of the terraces make it necessary, in several places, to build aqueducts – hollowed-out logs which bridge the irrigation furrows – so as to bring water to inconveniently situated plots.
When there is a shortage of water, the hura plots are given priority and not occasionally there is not enough left to irrigate the isirene land. Not surprisingly, individual Sonjo people do not depend entirely for their livelihood on isirene plots; they also own hura and magare plots. In order to adapt to water shortages later in the year, the isirene plots are planted very early, in the hope that crops will mature before irrigation water becomes scarce. Nowadays, cassava – a recently introduced crop which requires little irrigation – tends to be grown on the isirene plots. 
The agricultural system itself is very simple. The water is controlled by means of sluices and is channelled through furrows to where it is required. The system itself is best seen as “a fine network of small channels reinforced by a super-imposed coarser network of larger channels”.  That network is tilted towards the lower end of the valley – effectively ensuring that when there is too much moisture, the surplus drains away to rejoin the mainstream as it leaves the valley.
At the same time, water is not allowed to enter the irrigated area unless it is strictly required. Nor is it allowed to pass through the hura plots during the rainy season; the furrow which takes water to the magare plots only being opened after the annual rains and once the ground starts to dry out. That furrow is then only kept open until the grain is ripe after which it is closed off to remain idle for the rest of the year. There is thus no over-irrigation – an achievement of considerable importance since salinity is reduced and, at the same time, the spread of water-borne diseases is avoided.
According to Sonjo mythology, their irrigation system dates from when the tribe’s six villages were first founded by semi-mythical heroes. It is thus considered the sacred duty of each generation of Sonjo to ensure that the system is kept in good repair – the Council of Elders, or Wenamiji, being responsible for organising the necessary maintenance work.
After the flood waters have subsided, for instance, it is the Elders who organise the repair of any damage done by the rains. All able-bodied men in the village (with the exception of the blacksmiths who are banned for ritualistic reasons) are expected to take part – the principal tasks being to de-silt channels and to rebuild dykes which may have been washed away. If, during the course of the year, further damage is done by unexpected rain, the village men are called out again to repair the damage.
In general, however, the annual maintenance work takes about three days – although, before the introduction of iron hoes, it took much longer. Thus, in the past, the work was carried out using a digging stick – a tool considered by many to be ‘primitive’ but which was, in fact, quite adequate for the task at hand. 
The co-operation shown by the Sonjo in keeping their irrigation works in good repair is by no means exceptional. Indeed, such co-operation is a feature of all the traditional systems we have studied – and with good reason. The fact that maintenance work is done voluntarily, and with the help of the minimum technology, is of critical importance, since it means that it is not hindered by financial constraints, nor by a lack of technicians or a shortage of spare parts. It also means that the farmers do not have to be charged for water, which could cause serious problems among the poorer ones.
The irrigation cycle lasts about 14 days, during which time all the plots are supposed to be soaked once – the length of the cycle being determined by the necessity to water the hura crops at least once a fortnight. Again, the system is under the control of the Council of Elders, whose members have first call on irrigation water – the rest of the people being organised into several categories, each having different water rights.
Each 24-hour day is divided into four periods of about six hours each. Primary rights over water are allocated to each farmer for a 6-hour period. The first 6-hour periods are allocated to the seventeen Wenamiji who, between them, thus use up 4 days of the cycle. The privileges of the Wenamiji are accepted by all, not least because they are seen to be derived directly from Khambageu, the Sonjo cult hero. Those privileges enhance the status of the Wenamiji – and, hence, serve to maintain the tribal social structure.
After the Wenamiji, the next group to have access to the irrigation water are the eighteen ‘minor’ Wenamiji or Wenamiji barirage. They, in turn, are followed by the wakiama – a group of between 20 and 25 elders (known individually as mokiama) who have no hereditary rights to the water. They must therefore pay ‘tribute’ (usually in the form of goats or produce) to the Wenamiji. 
The number of watering periods assigned to the wakiama varies – not least because of the need to keep the length of the irrigation cycle down to 14 days. The water ‘bought’ by the wakiama – who are all members of established families and can thus usually afford to pay their ‘tribute’ – is not negotiable with an individual Wenamiji.
The price to be paid is set by the Council of Elders itself. Significantly, where the tribute is paid in the form of a goat, it is usually used in ritual sacrifices: indeed, such sacrifices are considered essential for the smooth running of the irrigation system. The tribute is thus seen as benefiting the whole community and not just the individual Wenamiji to whom it is paid.
Because of the limited number of irrigation periods available, the size of the wakiama is restricted, even so – when there is a real water shortage – sane mokiama may not be granted any rights of water in order to ensure that at least some crops are properly irrigated. 
Fewer than half of the farmers in the village which Gray studied, fall into the three above categories. The rest are ‘clients’: that is, they have to obtain their water from ‘patrons’ amongst either the Wenamiji, the minor Wenamiji or the wakiama. In most instances, that proves relatively easy as the patrons themselves can normally soak their plots in two hours rather than the allotted six and are thus likely to have a considerable surplus of water.
Moreover, a client is likely to have a close relative – a father or brother for instance – within one of the patron groups, and it is to him that the client will address himself. If he does not have any such relative, then he has to depend on his resources of honey, grain or money for obtaining water.
But even if he cannot obtain any water by these means, he can still obtain some illegally by making a break in a channel near his plots and flooding them at night. This is not considered a serious crime and the penalty imposed by the Wenamiji is one goat, which is considered relatively little. There appears to be little moral stigma attached to the offence. In effect, even the least privileged always have access to water.
The Wenamiji are also the village political leaders. Because, as Gray points out, they have “the power of depriving individuals of the water which is an absolute necessity for raising crops”, they are in a position to apply “an effective sanction for enforcing obedience to tribal laws and their own administrative orders”. 
Indeed, the whole elaborate system seems to work in such a way to maintain the authority of the Wenamiji. This could be a disadvantage if the Wenamiji were politicians or bureaucrats whose interests did not coincide with those of the tribe. But this is not the case. On the contrary, the role of the Wenamiji, like that of all tribal elders, is to act as custodians of the tribal wisdom and ensure its application – thereby assuring the society’s cohesion and the maintenance of its traditional way of life.
This being so, a system which maintains the authority of the Wenamiji is also that which ensures the cohesion of the social group and the stability of its relationship with its natural environment.
|1.||Robert F. Gray, The Sonjo of Tanganyika: An Anthropological Study on an Irrigation-based Society. Oxford University Press, Oxford, 1963; p.52.|