![]()
The Zambezi river
The Zambezi drains 540,000 square miles of central Africa, hundreds of tributary streams and rivers feeding into the 1,600-mile length of the river. It ranks fourth among Africaās rivers after the Nile, the Niger and the Zaire, and, like them, it changes character as it passes through different geographical zones. The Zambezi flows through floodplains, rocky narrows, vast sandy beds in which the river is almost lost during the dry season and finally a delta with a hundred-mile face on the seaward side. In order to describe the river and its history, it is a convention to consider it in three distinct sections: the Upper Zambezi, which extends from its source as far as the Victoria Falls; the Middle Zambezi, extending from the Victoria Falls to the Cahora Bassa rapids; and the Lower Zambezi from Cahora Bassa to the sea. Throughout its course, the river passes through a number of gorges that have been carved through the mountains that barred its way. The most important of these were the Batoka gorge below the Victoria Falls and the Kafue, Kariba, Cahora Bassa and Lupata gorges. With the exception of the Lupata gorge on the lower river, it was not possible to navigate through these narrows, and in between them there were many smaller rapids and falls, many of them exposed only during the dry season, but all presenting hazards for navigation.
However, the nature of the river is not only determined by geology. More important is the seasonal nature of the rains of central Africa, which swell the Zambezi and transform it from a modestly flowing stream into a huge torrent which overflows its banks and in parts creates extensive floodplains. Rainclouds travel in from the Indian Ocean during the months of October to April, during which time up to 50 inches of rain can fall on the upper reaches of the river and 30ā40 inches on the central African plateau further east. Half an inch of rain can fall in a single thunderstorm and two inches in the course of a day. The floods last from January to mid-May, during which time the Zambezi can rise 20ā30 feet above the dry-season level. The flooding of the Barotse plain is well known, but extensive flooding can also occur higher up the river as well, where the numerous streams entering the Zambezi form a network of waterways.
In December 1885, Frederick Arnot, a Plymouth Brothers missionary who in 1884 had been the first European known to have identified the source of the Zambezi, travelled through country on the right bank of the upper river and described the floods in his diary.
13th Started wading today, and all got into camp very tired I found some huts ā¦ built on dry mounds, the work of former travellers.
14th Water about knee-deep all day; got to another group of half-dry mounds; I pitched my tent with great care, and by setting each foot of my chair on an ant hill, managed to keep above water.
21st Crossed the Lutembwa Valley today. It is flooded, and covered with a forest of immense trees, which have large spreading roots like the mangrove. We waded for some hours through this dense forest. The emerald green mosses, shining through the water, contrasted beautifully with the pale ferns clustering round and among the tree-roots, and with icicle-like mosses, which hung in long delicate festoons from the branches at every opening in the forest.1
The source of the Zambezi is located in the Mwinilunga district in the north-west corner of Zambia, now officially termed the Zambezi Source National Monument, part of the Zambezi Source National Forest, a protected botanical reserve. This point is near the summit of the great watershed which separates the river systems of the Zaire from those of the Zambezi. These two giant waterways are not just rivers but whole drainage systems, fed by countless streams and major rivers that discharge into them from the surrounding plateaux. On its upper reaches, the Zambezi is joined by the Kabompo and the Lungwebungu as well as numerous smaller streams that flow into it every 10 to 15 miles. It is these tributaries that turn what begins as a small stream into a major river which, even a thousand miles from the sea, can be up to a mile wide.
The rains produce a notable rise in the level of the river, but the descent of the water from the upper reaches can take weeks to reach the sea, leading to a number of rises and falls in the level of the main stream in a single year. Moreover, rains are not always regular, so that there may be considerable variations between one year and the next, with prolonged dry periods that may last many years. The rain mostly falls on the high ground of the plateau so that the valley itself can sometimes suffer from drought conditions, and the average annual rainfall at Tete on the Lower Zambezi is only 25 inches. Matthew Hannaford has researched in great detail the cycle of dry and wet periods in south-east Africa over the period 1500 to 1830 and has shown how important they were in their influence on political developments, especially at the end of the sixteenth century and in the early nineteenth centuryāthemes that are discussed in subsequent chapters.2
During the period from April to September little rain falls. Many of the streams on the plateau run dry or are reduced to stagnant pools with sometimes only a trickle of water. At the height of the dry season, the Zambezi itself shrinks in depth and in many places in width. On the lower reaches, it can become a stream less than a hundred yards wide, meandering through a waste of sand, and so shallow that people can cross it on foot.
The vast drainage basin of the Zambezi is a major determinant of the regionās history. The river valleys that feed floodwaters into the Zambezi are sites of settlement, allowing people easy access to water and, although often marshy and liable themselves to flooding, also lines of communication. These valleys are often backed by rocky escarpments or by forested land which can be dense, hostile and sometimes almost impenetrable. The people who live in these river valleys have their lives structured by the seasons, which are not so much the spring, summer, autumn and winter of the northern hemisphere but seasons marked by the dry months and the coming of the rains.
The impassable barriers of the gorges, together with the nature of the delta, which did not provide any reliable navigation channels, and the extreme changes in the flow of the river all served to make navigation difficult and at times impossible, preventing the Zambezi from acting satisfactorily as a major highway from the sea into the interior. As hazardous as navigation of the Zambezi could be, the course of the river did include long navigable stretches. On the lower river, it was possible to sail from the Cahora Bassa gorge downstream to the point where the river splits into the streams of the delta; on the Middle Zambezi between Cahora Bassa and the beginnings of the Batoka gorge, there were long, relatively clear stretches which could be navigated by locally built canoes, while on the upper river the Barotse floodplain provided unbroken navigation from the Ngonye Falls almost as far as the riverās source.
However, nowhere on the Zambezi was navigation easy. During the dry season when the level of the river fell, numerous sandbanks and rocky outcrops diverted the stream into tortuous channels. Then, when the river rose as a result of the rains, the Zambezi would overflow its banks and create extensive shallows and marshes which could disguise the path of the main stream, while the passage over cataracts and through the gorges produced currents and whirlpools that could play havoc with small boats.
When the floods subsided, the peoples who lived on or near the river used canoes, which enabled them to negotiate the shallows and the smaller cataracts, as boats drawing more than a foot could not find enough water in which to float. On most of the river, the dugout canoe was the craft of choice. These varied from small canoes holding two or possibly three people to larger canoes able to carry nine or ten people or the equivalent in tonnage. Canoes operated without difficulty when the river was low, but when the water rose and the flow increased they became less easy to use. On the Middle Zambezi, they were more often employed to cross the river from one side to the other than to travel up- or downstream. On the Lower Zambezi, when the level of the river permitted, traders travelled in boats which writers in the seventeenth century called luzios. These were described in some detail by the Jesuit AntĆ³nio Gomes in the middle of the seventeenth century.
They sail on the rivers in boats made of the trunks of trees, both small and big, the big ones being made of nice planks of which there are plenty. They are trimmed and are not made with nails. They are put together with coir and are so tight and expertly made that they do not seem to have been made by cafres and they can be heavily loaded. They are divided into three sections: one is the shelter, in the middle; the stern section is where the rowing is done by the sailors, who are twenty-four [in number] apart from the helmsman; the prow section carries a big stove where the passengers and the crew cook their food. The middle shelter has a raised floor. Underneath they carry the goods in the bow and in the stern section they load barrels and boxes with the goods that must not get wet. The shelter is divided into two parts: one is for accommodation and can hold two bunks, writing bureaux, baskets and other such things, the other is a verandah with room for four chairs and a table and it is also used to steer the ship. When the sails are in use, the ropes are controlled from here and it is very comfortable for the people travelling in it.3
The luzio seems to have been an adaptation of the seagoing dhow. When Francisco de Lacerda, who undertook the first Portuguese mission to Kazembeās Lunda kingdom in 1798, travelled up the Zambezi from Quelimane he went in a boat he called a ballĆ£o.
BallƵes are also made from a single trunk and differ from the coches [small canoes] in having a deck and in having the cafres row seated in the waist ā¦ but they have short oars made so that the hand in the centre serves as a pivot and the other as the power for this lever. That in which I travelled was 48 palmos long and 8Ā½ palmos in the beam. They are only used for passengers and for such things as must not be damaged ā¦ My ballĆ£o had nine seats and eighteen rowers apart from the pilot (mallemo) and the man in the prow (mucadĆ£o).4
Anything bigger than the luzio or the ballĆ£o experienced huge problems. In 1858, Livingstone brought the first steamer to the Zambezi, and the diaries he and his companions kept are filled with the seemingly endless problems caused by running aground on sandbanks and hidden shallows. In 1898, Major A. St H. Gibbons tried to take a steamer up the middle reaches of the Zambezi as far as the Batoka gorge. He succeeded, just, but the experience underlined the impracticality of this kind of boat. Eventually, special shallow-draft stern-wheel steamers were introduced on the lower river, but their use ran up against another problem. During much of its course between the Cahora Bassa rapids and the sea, the Zambezi has no well-defined or stable banks where jetties, warehouses and mooring facilities could be built to service the steamers. The course of the river changed from year to year, meaning one yearās shoreline might be eaten away by next yearās floods and disappear altogether.
As with transport on land, the internal combustion engine brought about a revolution. In the twentieth century, the outboard motor enabled a large variety of boats of shallow draft to be used, making travel on the river more of a possibility. However, as late as the mid-1930s, travel on the Upper Zambezi was still largely by means of canoes or barges manned by paddlers.
The difficulty of navigation on the Zambezi was a dominant theme in the history of the river. A deep, evenly flowing river would have become a main artery for trade and allowed towns to be built along its course. These could have then become the centres of large state systems, with the river providing vital transport and communication links. It is possible to see how such ports might have developed by looking at the Lower Zambezi, where two towns, Sena and Tete, grew up as river ports and trading hubs from the sixteenth century onwards. These became the urban centres around which the Portuguese creole state was formed, but the lack of secure connections with a sea port, due to the maze of the delta streams, limited the possibilities for commercial development.
Zumbo, at the confluence with the Luangwa, was the only trading town that grew up on the Middle Zambezi. However, this town was short-lived. Flourishing briefly in the eighteenth century, it never formed part of any state system and was abandoned early in the nineteenth century. On the upper river, navigation was relatively easy for a hundred-mile stretch beyond the Ngonye Falls. Here, quite large towns developed near the riverāfor example, Sesheke and Lealuiābut for much of the year the Zambezi overflowed its banks and flooded the land, which made permanent settlements near the river difficult and led to seasonal migrations of the population.
If the course of the river and its difficult navigation made it un-satisfactory as a commercial highway and transport corridor, there were other features which made it an unfavourable environment for large state systems to flourish. Much of the middle and lower reaches of the Zambezi were flanked by rugged escarpments, which insulated it from the central African plateau. The escarpment was made up of barren, mountainous territory which in places was 20 or 30 miles deep. Much of the escarpment presented a very real barrier to easy movement and effectively cut off the people living in the valley from those on the plateau.
Much of the escarpment and the valley below remained without roads and was virtually unexplored through most of the colonial period. Between 1989 and 1991, a decade after the independence of Zimbabwe, Richard Hasler carried out research into resource utilisation in the Zambezi valley near the border between Zimbabwe and Mozambique. He described the difficulties of trying to access the communities in the valley:
Vehicle access ā¦ is via the gravel road down the Zambezi escarpment. The trip is six to seven hours by four-wheel drive vehicle from Harare to Kanyemba, four of which are on the dirt road. When descending from Guruve, the tar ends shortly before making the descent into the valley and the gravel road that traverses the valley floor is very badly corrugated in the dry season and can be impassable in the wet season.5
On the upper river there was no escarpment, but the Zambezi floodplain was bordered by dense forest and in the south by the marshes of the Chobe and the Okavango delta. The Zambezi valley also had an unfavourable disease environment which discouraged settlement. Mosquitoes bred in the marshes and pools left behind by the falling waters after April, and much of the valley was infested with tsetse flies, which made living there impossible for people whose economy was based on cattle rearing.
An overview of Zambezi history
Although the Zambezi is clearly divided into three segments, there is nevertheless a certain unity to its history. Just as the river valleys have attracted settlement, so they have also facilitated migration, and much of the discernible history of the people of the great central African plateaux is one of migration and conquest, as well as permanent settlement. The migrations were often the result of such natural causes as drought, famine and locusts. However, these causes were not easily distinguished from the wars of conquest of warrior elites, who subjected the populations of the river valleys to their overrule and exacted tribute from them. The history of the Zambezi valley reflects this interaction of the relative stability and permanence of the settled valley communities, whose agricultural society is characterised by small, lineage-based villages dependent on extended family networks, and the dominance imposed from above by warlike invaders who established the overrule of a tribute-taking state system.
These invaders, often warrior elites rather than large migrating populations, were able to create state systems governed through the central authority of a king, backed by his office holders and military commanders, and with an elaborate ritual developed to confer mystique and ancestral authority on the ruler. As well as tribute-taking, these were also slave-owning states. The conquerors invariably sought to increase the manpower at their disposal by capturing slaves, attracting clients and dependents and taking wives from the subject populations. However, the degree to which these states depended on slaves differed. The Barotse and Ndebele kingdoms raided extensively for slaves, while slaves were perhaps less important for the Karanga and Marave kingdoms of an earlier period.
The language of the conquerors was also not infrequently imposed as a way of improving communication in a culture that had few means of conveying information apart from word of mouth, and as a way of unifying and consolidating the state. These conquerors always, in the end, reached some accommodation with the settled communities, recognising and incorporating the local spirit cults and rain shrines of the long-settled village communities into their system of control. Ultimately, however, these Zambezian state systems depended on the ability of the elites to mobilise and monopolise the use of force.
The Zambezian states, unlike those that emerged in Europe, never had clearly defined frontiers (although the Zambezi provided a natural barrier to the expansion of raiding and tribute taking, and came to constitute a kind of informal frontier on the lower and middle reaches of the river). The African states that grew up north and south of the river were defined by the allegiance of populations rather than by territorial boundaries. The extent of a kingdom was only as large as the population from which its rulers were able to exact tribute. This inevitably meant that the size of kingdoms could fluctuate considerably over time and that it makes no sense to try to draw their exact boundaries as Europeans tried to do at the time of the partition of Africa. Nowhere were the attempts to define precise boundaries more controversial than in the case of the Lozi kingdom of Barotseland on the Upper Zambezi. When European rulers began to delineate the frontiers of their colonies on the Upper Zambezi, they were faced with all the fluctuating uncertainties that surrounded the claims of the Lozi kings to āruleā vast stretches of the Zambezi valley and its surrounding lands. As Lawrence Flint explained, these claims could be extreme.
At its maximum extent in the late nineteenth century, Lozi rule probably included all the people inhabiting the lands of the Upper Zambezi River basin from approximately 12Ā° S as far downstream as the Victoria Falls. This includes all of present-day Western an...
