In this book, archaeological evidence has been used to re-examine the origins and early history of the Chewa, Malawi's largest ethnic group. Much of the evidence has come from the archaeological site of Mankhamba, located in the southern Lake Malawi area. Until now, scholars have documented the history of the Chewa largely through using oral traditions. Many sources of the traditions agree that Mankhamba was the place where the king of the Chewa, whose title was Kalonga, established his major settlement many centuries ago. The author was fortunate to discover this site, the excavation of which has helped clarify the early history of the Chewa and has shed some light on other aspects of their way of life.

The earliest of the written documents on Malawi go back to the sixteenth century. Some adventurous Portuguese explorers and traders who periodically passed through central and southern Malawi as they sought minerals and other resources in the interior of the region wrote these documents. In their reports home, they sometimes mentioned the places they had passed through and described some of the people they had seen or met. It is from such accounts that historians first read about the ‘Maravi’ people. The term ‘Maravi’ was the Portuguese rendering of the term ‘Malawi’, by which the Chewa referred to themselves. It is also from such records that historians were able to crosscheck the accuracy of oral traditions about Kalonga and other Maravi kings such as Lundu. Some of the Portuguese also described the nature of local villages and towns but very little else.

Because there was not much in these documents to enable historians to reconstruct the history of the Chewa or of any other ethnic group, they resorted to recording oral traditions. Of the more than a dozen ethnic groups in Malawi, historians mostly recorded the oral traditions of the Chewa. This was probably because the Chewa or the ‘Maravi’ were the most mentioned of Malawi's ethnic groups in the Portuguese documents and this gave historians a good starting point for understanding them and their history.

Oral traditions and the current distribution of ethnic groups in central and southern Malawi make it rather obvious that the Mankhamba site was founded by people who spoke Chichewa, a Western Bantu language. Archaeology has provided a time span for the occupation of the settlement but has not given a depiction of what the settlement might have looked like. For this, one may extrapolate from modern village settings in Malawi and from observations made by Portuguese explorers and traders in southern Africa during the seventeenth century. This will help to establish an image of what the Mankhamba settlement might have looked like when Kalonga and his people occupied it. The material remains from the excavations have provided ample evidence regarding the way of life of the Chewa, which will be covered in this chapter. The discussion will concentrate on how they made a living. Long-distance trade will be discussed in detail in the following chapter.

Founding the Mankhamba settlement

Scholars who have analysed oral traditions of the Chewa believe that the pre-Maravi were the first Chewa group to arrive in Malawi and that they were followed by the Maravi. According to Schoffeleers, the pre-Maravi were in full possession of the land when the Maravi arrived. Some of them had settled at Mankhamba, where they established a raincalling shrine. However, oral traditions provided no reliable dates for the arrival of either of the two groups. In fact, for a long time, historians and other scholars have grappled with the chronology of the migration of the Chewa. Their main problem was that the only sources of information they had were documents written by the early Portuguese and the oral traditions themselves.

The earliest of the Portuguese documents were written in the sixteenth century, long after the Chewa had settled at Mankhamba. They contain records of events that took place after, and not before, the fifteenth century. Oral traditions are also problematic in their own way. Unlike archaeology, which involves analysing material remains recovered at archaeological sites and scientifically dating them, oral traditions can be unreliable when they refer to events that occurred many centuries ago because people tend to modify narratives through time. Archaeology has now stepped in to help clarify the chronology and the settling of the Chewa at Mankhamba and other sites.

To appreciate the extent to which the Maravi were involved in long-distance trade, one needs to examine this in relation to what was going on in southern Zambezia, the area south of the Zambezi River controlled by the Mwene Mutapas in Zimbabwe. Long-distance trade has a long history in that area compared to other areas in the interior of southern Africa. Early seventeenth-century written records of the Portuguese show that the Kalongas in Malawi had dealings with the Mwene Mutapas. One of them, Gatsi Rusere, is said to have ‘affirmed that a number of Maravi chiefs, including Kalonga Muzura, were his military allies’. Such an alliance must have had economic benefits too.

Commenting on the nature and intensity of the trade in southern Zambezia, Chirikure pointed out that imported items, such as glass beads, cowrie shells and bronze items began to appear before ad 900. The earliest glass beads were recovered at many archaeological sites, including Chibuene in southern Mozambique, Makuru in south-central Zimbabwe, Schroda in the Limpopo River Valley, and at sites in northeastern Botswana. They belong to a time that archaeologists refer to as the Zhizo period, dated between ad 600 and 900. Bronze was recovered at the site of Berryl Rose Claims in northern Zimbabwe where it was dated to ad 800 and cowrie shells appear elsewhere in the hinterland at about the same time. The distribution of the sites shows that both the elite and ordinary people in southern Zambezia participated in the trade. Some of the commodities used in the exchange of these items were ivory, cattle and perhaps gold. Although the role of gold during this early period is not that clear, the evidence shows that it was being exploited at least by ad 1000. The same might apply to the exploitation of copper. Remains of copper in southern Zambezia were found at several sites, including those that had been occupied by non-elites. This metal was so widely used, it has been argued that Africans considered it more valuable than gold.

The evidence in northern Zambezia, where the site of Mankhamba is located, shows that long-distance trade was not as well established as in southern Zambezia. Although rich copper deposits exist in the north-west of the area, the extent to which copper was exploited during this early period is not clear.

Archaeologists refer to the period of iron-based cultural adaptation and elaboration as the Iron Age. They have divided it into the Early and Late Iron Age periods. The Early Iron Age was the time when iron tools replaced stone tools and ironworking technology became widespread. There is still some debate as to whether ironworking technology in sub-Saharan Africa was indigenously invented or not. In some areas, such as the Great Lakes region, people were producing iron implements by the middle of the first millennium BC. By the beginning of the second millennium ad, more cultural and economic changes began to take shape in much of sub-Saharan Africa, a period archaeologists refer to as the Late Iron Age. These changes included the development of states in areas where such institutions had been unknown, and the intensification of long-distance trade. In many parts of sub-Saharan Africa, state formation and long-distance trade seem to have occurred more or less at the same time, with trade being ‘seen as both a stimulus and a mechanism for state formation’. These activities resulted in a great deal of population displacement and movement, as was undertaken by the Chewa.

The southern Lake Malawi area attracted more attention from archaeologists than all other areas in the country after archaeologists confirmed that the earliest Iron Age communities arrived there early in the third century. During the last five decades of the twentieth century, several archaeologists carried out various research projects in the area (see Map 6.1) where they discovered and described five different Iron Age pottery types. This enabled archaeologists to establish a pottery sequence which C-14 dates have confirmed. The Iron Age archaeology of the southern Lake Malawi area became the best known in the country.

The environment of the southern Lake Malawi area

Lake Malawi and the Shire River, its only outlet, dominate the environment of the southern Lake Malawi area. The lake is one of Malawi's most outstanding topographic features. It lies within the Great Rift Valley of East Africa, which stretches from the Zambezi River in Mozambique to the Red Sea. The Lake Malawi part of the Rift Valley runs almost due north for a distance of about 644 kilometres and it averages 80 kilometres in width. The lake itself measures some 29 604 square kilometres.

This book is a comprehensive account of the origins and early history of the Chewa who began to refer to themselves as ‘A Malawi’ (the Malawi) at some point after they arrived in the southern Lake Malawi area. Seventeenth-century Portuguese explorers and traders recorded the term as ‘Maravi’ whereas modern scholars use ‘Maravi’ interchangeably with the term ‘Chewa’.

Some readers will undoubtedly be surprised that an archaeologist rather than a historian has written this book. They should bear in mind though that archaeology is the only tool available for unearthing (literally) history and culture before any written records. Archaeologists call that time the pre-historic period and that is where the narration of the origins and early history of the Chewa begins. It ends about two decades before the imposition of colonial rule by the British in 1891.

Before delving into the narration, here is a synopsis of the intellectual journey that has brought me to this point. During the first few decades of the colonial period, the British attempted to record the oral traditions of the various indigenous groups in the country, perhaps with a view towards understanding their history. They published nothing from the exercise and any history taught in the country's schools was the history of the British themselves. In primary school, the favourite topics included Dr David Livingstone's exploration of the country and of other parts of southern Africa. Pupils also learned about the efforts of early Scottish missionaries to establish mission stations and schools, and about the arrival in Malawi of early British traders and settlers. The traders established the African Lakes Company, a trading company that locally assumed the name Mandala, meaning ‘spectacles’ (a fascination to the local people), worn by one of the joint managers, John Moir.

In secondary school, students came face to face with the history of the British Empire and the Commonwealth. Teachers drilled them in this history so well that some of those students still vividly recall important events in British colonial history. Perhaps what is unfortunate is that they can intelligently discuss events such as the Boston Tea Party, the Black Hole of Calcutta incident, the Anglo-Boer war and others better than they can the arrival of the Chewa at Mankhamba, or of the Tumbuka, Yao or Ngoni in various parts of the country.

Despite the existence of published oral traditions that clearly indicated where they could find the site of Mankhamba, archaeologists working at Iron Age sites in Malawi made no effort to find it. There seems to be no plausible excuse for this, except perhaps they had more than enough on their plate involving research. They busied themselves with site surveys and the excavation of sites located in Mangochi district, Bwanje Valley, along the Shire River and in the Shire Highlands in southern Malawi. Their focus in central Malawi was the Dedza-Chongoni area, Salima and Nkhotakota districts; in the north, they worked in Karonga, Rumphi and Mzimba districts.

When I completed my doctoral studies, I too gravitated to Mangochi district where I discovered and excavated the Mtemankhokwe 1 site. This turned out, however, to be another burial site with material remains similar to those recovered by Inskeep at Nkhudzi Bay, which forced me to rethink my research strategy and to look for an area that archaeologists had never investigated before. Of the many unexplored areas in Malawi, the area between Bwanje Valley and Salima was the most attractive because, according to oral traditions of the Chewa, the site of Mankhamba was located somewhere there. There was, therefore, a good chance of locating it if a well-planned site survey of the entire area were undertaken.

Current ethnic groups in the area agree that the Chewa were the earliest known ethnic group to settle there, even though they no longer wield the political power they once did. They lost political power to the Yao and the Ngoni who settled there some time during the nineteenth century. Despite that, the Chewa have retained their ritual authority. For instance, they have an active rain-calling shrine, which they located at the Mankhamba site itself, and the Nyau secret society in the area continues to thrive. Retention of these structures, which are important to Chewa culture, means that the Chewa are able to play significant cultural and social roles. When I informed Group Village Headman Kafulama, the senior Chewa headman in the Mtakataka-Mua area, about the project, he was supportive. My interactions with Kafulama and others proved helpful in locating the Mankhamba site.

Finding Mankhamba

Despite being aware of the likely geographical area of the Mankhamba site, finding its precise location was not easy.

According to oral traditions recorded by Ntara, Kalonga began to expand his area of influence before he arrived at Mankhamba. He did this perhaps to prevent other senior members of the ruling family from unilaterally breaking away from him. Two of them, Kaphwiti and Lundu, had done so already.

Kalonga's strategy for expansion was to send out some of his sons and other relatives to various places far away from him to establish new settlements. The oral traditions suggest that he started doing this when he was at Mawere a Nyangu. Among those he sent out were Changamire and Chulu, of the Phiri clan, and possibly Mkanda of the Mbewe clan. Kalonga also let go certain able individuals who requested his permission to venture out on their own, as was the case regarding Mkadzula of the Mwale clan. These leaders became territorial chiefs with the title of Mwini Dziko, (owner of the land) whom Kalonga dispatched with pomp and ceremony. He gave each one emblems of office, which included a flywhisk, a sword or knife, and an iron stool. Once he had established himself firmly at Mankhamba, Kalonga called for a meeting of his elders to allocate land to various individuals. He appointed one of his elders called Chembe to be in charge of land allocation, and Chembe in turn delegated the work to Mgawi, his younger brother. Mgawi instructed those to whom he had given land to divide the land and allocate it to others. This created many lesser chiefs who, together with territorial chiefs, regularly sent ivory and part of the initiation ceremony fees to Kalonga as a tribute.

Kaphwiti and Lundu

Kaphwiti and Lundu were relatives. Kaphwiti was either Lundu's elder brother or uncle. The two broke away from Kalonga and went elsewhere to establish their own settlement. Oral traditions have given various reasons for their departure. Some traditions say that they left because there was either some dispute between them and Kalonga over the practice of evil magic and the administration of mwavi, or there was some dissatisfaction with the choice of leadership after Kalonga's death. Mwavi is a poisonous drink extracted from the bark of the Mwavi tree (Erythrophleum suaveolens), also known as the ordeal tree, which was administered in order to find culprits in disputes involving witchcraft, theft or adultery.

When archaeologists find faunal remains at archaeological sites, they often assume that humans brought them there in their quest for animal protein and other animal products. There is an exception to the rule, though, when it comes to bones found at rock-shelter and cave sites, because humans were not the only users of these sites. Some carnivores and birds of prey such as owls also used them and, like humans, they sometimes brought back body parts or whole carcasses of animals that they had hunted or scavenged. Identifying remains brought in by humans is usually easy because humans butchered the animals and roasted the meat on a fire before eating it. These processes often left cut, burn or scorch marks on the bones. Archaeologists view the marks as evidence that the animal remains went through what some have referred to as a ‘cultural filter’. In other words, the marks on the bones are evidence that humans rather than animals of prey handled these bones.

Meat was not the only useful animal product obtained from animal carcasses. Archaeological evidence shows that humans also broke the limb bones to extract nutritious bone marrow. They used animal skins too for a variety of purposes, such as bedding, carrying children around and for clothing. During the Iron Age, animal skins and some horns were trade items too and people traded these with foreign traders, some of whom had established settlements along the coast of the Indian Ocean starting from the eighth century. Further, people modified bones, ivory and shells by shaping them into useful cultural objects. Among the objects they made from bones were needles, awls, ‘beads and ornaments, tubes for smoking pipes and musical instruments’. From shell and ivory, they made beads, pendants, bangles and other objects.

In general, however, archaeologists tend to find more objects made from bone at Stone Age rock-shelter or cave sites than at Iron Age open sites. The reason for this is that, apart from ivory, most Iron Age people had no need for bone tools and ornaments because they had iron implements for tools, and glass beads, copper and gold jewellery for ornaments.

Success in obtaining information from faunal remains depends on how well preserved the remains are at the time they are recovered by archaeologists.

As noted in Chapter 6, pottery designs and decorations are critically important in archaeology because they often reveal some aspects of the culture of the people who made them. This chapter presents local and imported ceramics as well as stone objects which the Chewa used for tasks that could not be accomplished using iron implements.

Local ceramics

Although the foreign ceramics are important in that they show the involvement of the Chewa in long-distance trade, emphasis will be placed on the local material because this was the handiwork of the Chewa themselves. Pots and bowls, illustrated below, dominated the local ceramics, with smoking pipes and spindle whorls following in dominance.

Pots and bowls

As shown in Table 8.1, one whole vessel and 45 850 potsherds were recovered at Mankhamba. Level 2 produced the largest number of potsherds, with over 40 per cent whereas Level 1 produced the smallest number, with only 9 per cent of the total potsherds. The first step in trying to obtain information from the potsherds was to sort them with an eye to finding fragments that had been parts of the same vessel. The sole surviving whole vessel, recovered in Level 4, was a small shallow bowl that was missing a tiny section of its rim. Recovering a well-preserved whole vessel in an excavation characterised by heavy fragmentation was unexpected. In fact, whole vessels are rarely recovered at settlement sites, but are fairly common at burial sites where they were carefully placed in the graves as containers for burial goods. At the two Late Iron Age burial sites excavated so far in Malawi, Nkhudzi Bay and Mtemankhokwe 1, whole vessels were recovered.

Besides being small — and this was probably why it survived — the Mankhamba whole vessel was not decorated. It had a small hole with a diameter of 7 millimeters intentionally made at the bottom (see Figure 8.1 no i). Drilling holes in pots was a fairly common practice in the southern Lake Malawi area. At Nkopola Swamp, located along the western shores of the eastern arm of Lake Malawi, some potsherds were recovered which had several holes drilled in each one of them.

In Chapters 3 and 4, I have used both published and unpublished sources to summarise the current state of knowledge regarding the early history of the Chewa. Oral traditions were the basis of nearly all the consulted works, with the most prominent being Mbiri ya Achewa. All researchers investigating the history of the Chewa, following Ntara's publishing of his work, have consulted the book, which is considered seminal and the ‘the fullest body of oral tradition’. Ntara's work is of great value to researchers not only because it was the first authoritative study of the Chewa, but also because of the manner in which he presented his work. Ntara did not offer much analytical or interpretive detail, narrating instead the traditions in a summarised format. As a result, it afforded subsequent scholars the opportunity to use his work as if they were using unmodified oral traditions. They commented, analysed and interpreted it to suit the needs of their research. One researcher who made very elaborate comments was Langworthy who, having written an introduction to the translated 1973 edition of the book, proceeded to provide general and interpretive comments to most of its chapters. While this has the disadvantage of making it unavoidable to view Ntara's work from Langworthy's perspective, it rendered the work readily comparable with other subsequent scholarly works on the history and culture of the Chewa. Besides the literature based on oral traditions, there are seventeenth-century documents written by the early Portuguese that refer to the Maravi. The very mention of ‘Maravi’ in the documents is significant in that it confirms that the Maravi, as a distinct group of people, did indeed exist. Some of the documents have helped clarify oral traditions.

Oral traditions

Oral traditions are messages from the past that go beyond the present generation, which are transmitted orally and concern events that occurred beyond the lifetime of the informant. They are expressions of the past or what Vansina referred to as ‘the representation of the past in the present’. Often, oral traditions ‘contain names, titles, or sayings with metaphorical content that also appears on material culture’. Because of this, oral traditions can be useful to archaeologists if properly analysed and interpreted, although it is important not to confuse oral traditions with oral history, which involves messages or statements about the present generation.

Ntara was the first person to admit that it is very hard to identify the specific place of origin of the Chewa and to trace accurately their migration route into Malawi. He said that the old people whom he interviewed gave conflicting answers. They placed their home of origin as far north as Egypt with some saying it was Uganda and others mentioning Mombasa in Kenya. In 1970, Linden interviewed a man whom he said was Ntara's principal source for the first few chapters of his book. The man mentioned Libya as the place of origin of the Chewa. A majority of the oral traditions of the Chewa, however, claim that they came from a place called Uluwa or Luba in the Shaba or Katanga area of the Democratic Republic of Congo (DRC). This claim makes sense when one views the Chewa in the context of the general history of African people called the Bantu. The term Bantu does not refer to race, ethnic group or culture; rather, it is a linguistic term. In many Bantu languages, the word-stem ‘ntu’ means ‘person’. The prefix ‘ba’ indicates plural. Ba-ntu therefore means ‘persons’ or ‘people’. Linguists concluded that Bantu languages are part of one language family called the Niger-Congo language group. The Bantu linguistic homeland is in the general area of eastern Nigeria and western Cameroon which is where the greatest diversity exists among modern Bantu dialects.

Expansion of the Bantu

The expansion of the early Bantu, also referred to as the proto-Bantu, led to the occupation of central, eastern and southern Africa by Bantu-speaking people. Today, there are nearly 900 Bantu language varieties that show a remarkable level of relatedness. They are spoken by more than 200 million people spread out over an area of more than 9 million square kilometres, from southern Somalia in the northeast to southern Africa. Reasons for the Bantu dispersal and the directions or routes they took as they spread out to cover a disproportionately large part of Africa have been the subject of intensive research. The results have generated some disputes among linguists, archaeologists, historians and other scholars. Since the research is ongoing, what is presented below is not the final word. It simply represents the current state of knowledge.

Although many people are familiar with the term ‘archaeology’, very few know what the practice of archaeology entails. It is common for visitors to an archaeological excavation to ask archaeologists ‘how did you know where to dig?’ That question comes after they have seen some of the excavated material. How, they wonder, did archaeologists possibly know that they would find such materials at that precise location? This chapter looks at what archaeology is all about.

What is archaeology?

Archaeology is the study of humans from both the distant and recent past, using material remains — that is, the objects they made, modified, used and then discarded or abandoned. Wherever people have lived even for a short period, they leave behind evidence of their presence in the form of artefacts, ecofacts and features. Artefacts are any portable objects that human beings intentionally made or modified in order to accomplish various tasks and discarded after use. Such objects include tools made from stones, metal, bone or wood; jewellery; and ceramics, including pots, smoking pipes, figurines, beads and other objects. Ecofacts, which are objects found at archaeological sites that humans did not intentionally modify, include bones of both wild and domestic animals with or without butchery marks, firewood on a fireplace, charred or uncharred seeds of both wild and domestic plants, and other objects. Features are any remains of human activity that are not easily movable. Archaeologists observe and study them on site and leave these features there, which include building foundations, floors, irrigation canals, postholes and other features. Qualified archaeologists recover or observe these material remains when they excavate archaeological sites. After recovery, archaeologists take the portable objects to a laboratory where they analyse them in order to answer specific questions regarding the way of life of the people who made them.

Although archaeology is often seen as studying prehistory, that is, the study of cultures that existed before the invention of writing, there is also historical archaeology and the archaeology of modern society. Historical archaeology is the study of complex cultures or civilisations which were literate and have some written records. Considering that biases always creep in whenever people write or produce documents, archaeologists are able to compare if the archaeological record agrees with some of the written relics.

The decline and fragmentation of the Maravi empire began in the eighteenth century, reaching an irreversible point by the turn of the nineteenth century. Some of the reasons for this included the peculiar system of succession of Maravi kings, conflict between them and some of their territorial chiefs, and the rise of the Yao as formidable long-distance traders. The Maravi system of succession created conditions for instability in the absence of strong leadership. As discussed in Chapter 4, the Maravi operated a decentralised political system. Because children could not succeed their father as Kalonga, they came up with an unusual way of choosing successors. They had to be members of the Phiri clan and descendants of Nyangu, Kalonga's sister or mother. This worked well until Undi and Nyangu's unilateral departure from Mankhamba. Since Nyangu was with Undi, it was his responsibility to nominate one of Nyangu's sons to succeed Kalonga. The nominee was, however, subject to the approval of Kalonga’s counsellors at Mankhamba who were of the Banda clan, and he had to marry Mwali, also of the Banda clan. She would be his principal wife and therefore a powerful woman. Simply stated, the succession system gave the Banda clan the upper hand and the Phiri clan little or no say in issues pertaining to Kalonga's successor. With time, the system isolated and weakened the Kalonga, who often had no dependable relations locally.

Sometime during the eighteenth century, perhaps due to the growing influence of the Banda clan, the requirement that Undi nominate Kalonga's successors was no longer enforced which meant that counsellors of the Banda clan at Mankhamba assumed that responsibility. On the surface, it seemed sensible to appoint successors locally. After all, that was how they did it before Undi and Nyangu left. However, times had changed, and this time it did not work smoothly. Oral traditions recorded by Ntara show that the system of succession had become open to all forms of manipulation and abuse by various powerful counsellors. Disturbances began to occur, as people who were not Phiris, such as Mchepera who was of the Mwale clan, began to aspire to become Kalonga.