Mwene Muji: A Medieval Empire in Central Africa?

Abstract

Although the Lower Kasai was identified by Jan Vansina as a likely center for highly complex societies, he failed to recognize that sixteenth-century sources had mentioned the Empire of Mwene Muji as a large polity in that region. Studying the well known and recently discovered literature on West Central Africa, as well as a critical study of oral tradition, shows considerable evidence for the antiquity and existence of Mwene Muji.

When the Italian humanist Filippo Pigafetta published his celebrated description of the Kingdom of Kongo in 1591, he mentioned in passing one of Kongo's eastern neighbors as the “very great kingdom” and even the “empire” of Monemugi, or Moenemugi, as it was labeled on the map that accompanied the book.¹ Pigafetta's account was based on the “writings and reasonings” of Duarte Lopes, a Portuguese New Christian who had resided in Kongo from 1579 until 1583. Lopes came to Rome when Kongo's King Álvaro I commissioned him to be his ambassador, enjoining him in his instructions to “first give a detailed [minutamente] account of what has happened and taken place in these my kingdoms.”²

The name appeared again in published European descriptions, in another equally renowned 1668 compendium of Dutch knowledge, Olfert Dapper's description of Africa.³ As Africanist scholarship developed in the twentieth century, this mysterious empire was largely left out, mostly because of the paucity of information found in these sources, and because Pigafetta's map of Kongo placed it more or less in East Africa. Its location was so uncertain that a summary of European geographical knowledge of Africa published in 1918 firmly placed it there, with the Maravi (in Malawi) and the Nyamwezi (in Tanzania) as its likely inhabitants.⁴

The East African idea stuck. In 1961, anthropologist Erika Sulzmann contacted Jan Vansina to make queries about work he published in 1956 on Kuba history for her own fieldwork on the origins and expansion of the Bolia. She noted that the “Nunu” controlled trade between the lower Kasai and the Maleo Pool, and wondered if they might have played a role in Kuba's foundation. Vansina did not agree, and wrote emphatically “No” in the margin of her letter.⁵ Although she did not mention Monmuge, in his lengthy reply to her, he sent an excerpt from Dapper dealing with this area, and added in a commentary on it “Monmuge was of course located in East Africa.”⁶

Sulzmann replied rapidly to this, and wrote an extensive defense of her identification of the Nunu with Mwene Muji, while effectively demolishing the East African identification.⁷ This was clearly on her mind when Sulzmann did a field expedition the next year (1962–63) to the lower Kasai to follow up on the ideas she presented to Vansina. On 17 May 1963, she noted an interview she had with Mázimawa André, “born in Musye, now a skipper [Bootschauffer] in Inongo,” who told her that that both Mwene Muji and Nimi a Maye were titles of the Banunu lord of Mushie and “so Muene mushye.”⁸ Although her correspondence does not include a subsequent letter to Vansina, he acknowledged her as the source when he published Kingdoms of the Savanna in 1965, now placing Mwene Muji around Mushie.⁹

Vansina accepted the existence and determined the location of this kingdom or empire (which I will write as Mwene Muji), but he apparently did not agree that Mwene Muji should enjoy imperial status.¹⁰ In 1973, while preparing an article on the Jaga invasion of Kongo, Joseph C. Miller asked Vansina about the Jagas’ origin in Mwene Muji, and from Vansina's reply Miller wrote that it was “another half-mythical king whom geographers of the time commonly placed in the unknown interior of the continent.”¹¹ Even Sulzmann did not grant it much status, proposing it was rather a trading hinterland of merchants from Kongo than a proper state, let alone an empire.¹²

Vansina's hesitation to grant more than geographical status to Mwene Muji persisted. In Paths in the Rainforest, Vansina tried to reconstruct the sociopolitical history of much of Central Africa. Using largely linguistic insights bolstered by some archaeological and ethnographic material, Vansina identified a number of centers where linguistic innovations and some archaeological evidence pointed to early and influential development of large and influential kingdoms, one that led to a Teke center and another that would lead to Kongo and Loango.

After examining these centers and their resulting large historically attested kingdoms, Vansina mused on the lower Kasai (which included Mushie) and wrote that the evidence “suggests that there existed on the lower Kasai a center of original development focused on the Boma, Buma, Saa, and Yans groups. Much of the political terminology, especially in the first three groups, is quite original.”¹³ While it should be noted that Vansina wrote before more robust language samples had been studied, and his confidence in the locations of change should probably be diminished, the fact remains that the lower Kasai must be considered as a potential center for these sorts of linguistic innovations.¹⁴

Vansina, however, continued to downplay the significance of the lower Kasai, claiming that it was only later, when the Boma Kingdom developed (in the early seventeenth century), that this region actually achieved the development its terminology seemed to place in an earlier time. Somehow his recognition of what was called an empire in an early source, Mwene Muji, did not qualify as a major polity like Kongo or Teke in spite of the written evidence.

Mwene Muji in Contemporary Documentation

More recent research can elaborate a bit more on the reality, extent, and antiquity of Mwene Muji. The recently discovered Florentine Relation, an account probably written by the Carmelite missionary Diego de la Encarnación in 1586 or 1587, makes it clear that the Carmelites had independent knowledge of an “empire” of Mwene Muji.¹⁵ While doing missionary work in Kongo's eastern province of Mbata, de la Encarnación noted, in a separate report, that he received a visit from a “Lord of many peoples whose country was more than 150 miles away.” In fact, Mushie is about 150 miles from Mbata.¹⁶ He had come that distance even though he was lame in one foot and begged the priest to baptize him.

At first de la Encarnación refused, saying he did not have enough time, but when the Mwene Mbata interceded on the lord's behalf, the priest yielded. “I took a month to catechize and baptize him, and gave him someone to teach them the doctrine there in their country.”¹⁷ The geography and the fact that Mwene Muji was listed as the only eastern country bordering Kongo make this a likely identification and that the contact took place through Mbata adds to that probability.

We learn that a southern portion of Mwene Muji was actually close to Mbata, because Pigafetta noted that the notorious “Jagas” who invaded Kongo around 1567 lived in a province of Mwene Muji before invading Kongo through Mbata.¹⁸ Mbata's eastern border was probably just north of the northern border of today's Angola.

The idea that some sort of evangelization was envisioned also fits with another piece of testimony from Carmelite sources. This was an anonymous account of Kongo completed around 1608 that accompanied a large dossier compiled for Juan Baptista Vives, who had taken on the role of “protector” of Kongo for the Vatican. Diego de la Encarnación had a role in informing, if not writing, the text.¹⁹

The Jaga invasion is mentioned twice in the Vatican manuscript. In the first instance, the wording was clearly drawn from Pigafetta, even if not exactly the same. But the second instance contains new information about Mwene Muji when considering the possibility of finding an overland route from Kongo to Prester John's Ethiopia. Here the author noted that such a trip had not been tried “because it is prevented by the Xacchi who destroyed the Congo in the past years.” But now, the author contended, he had been told by Antonio Manuel, Kongo's ambassador to Rome (who had been communicating with de la Encarnación since 1607) that “already the greatest part of those Xacchi have received the faith, and the others will receive it, if there is anyone who preaches it to the Emperor of Monomengue, of whom they are vassals.”²⁰

This conversion had additional benefits, the Vatican author contended, as now that the Jagas had become subject to Mwene Muji, and following Antonio Manuel's testimony “they are losing their strength with the sweetness [suavita] of Christ's yoke, because, [lacuna].”²¹ It suggests that de la Encarnación's sending someone to the unnamed eastern power had some success, even if it would prove to be ephemeral, although the real effort may well have come from Kongo. Earlier, Diogo I (r. 1545–61) sent his school teachers as lay missionaries to its eastern neighbors, such as Great Makoko, Kundi, Kongo dia Nlaza, and the otherwise unknown “Empalaquate, who is a great lord and lord of great riches, he is a good friend of ours… and very close neighbor to the great queen of Matamba.”²² The Mwene Mbata's entreaties were thus in line with Kongo's interest in the region, and Diogo's successor, Álvaro I might have continued religious diplomacy like de la Encarnación's initiative.

This claim is further strengthened by a report concerning a voyage by the Mexican-born Spanish priest Rafael de Castro in the first decade of the seventeenth century, who traveled from Kongo to some “200 leagues” beyond Okanga in an to attempt to reach Prester John, or well within the domains of Mwene Muji. He reported that the people in this region (which he did not name) were wearing crosses, and “they asked where he came from and where he was going, and he said from Manicongo where the king was Christian and he intended to convert and make Christians, and they said we have been made Christian by his hand.”²³

If the Jagas were under “Christ's yoke” or vassals of Mwene Muji in 1604 (when Antonio Manuel left Kongo), they appear to have become independent some time later. The Yaka Kingdom which probably emerged where the “Jagas” originated in Mwene Muji, was first attested in 1656 when Capuchin missionaries working in Matamba reported its existence. They noted that Kasanje, whose territory was founded around 1635 had been active on both sides of the Kwango River, and the Capuchin Antonio da Serravezza met the son of the ruler of “Quiâca,” who had been captured by Kasanje in a war against his kingdom some years earlier. He was then serving as an ambassador between his father Mbangu a Kitana and Kasanje. His country was north of Matamba and east of the Kwango River as it flows past Mbata's western border.²⁴

One might hope that modern day Yaka traditions could offer some more information, but unfortunately, in the 1750s Yaka was conquered by the expanding Lunda empire. Traditions recorded in the twentieth century related to the Lunda conquerors, and not any prior government, and, much as we will see, the Boma Kingdom's traditions effaced the account of Mwene Muji's tradition outside of its core region of Mushie.²⁵

Establishing the probable southern border of Mwene Muji as in the Yaka Kingdom, then leads us to its northern border. Pigafetta's account is confusing on this, because he claimed that the Jagas, as he understood Lopes, “inhabit around the first lake of the River Nile, in a province of the Empire of Monemugi,” which would probably also be the northern border. Since the evidence on the southern border is more comprehensive, we can put this down to the general uncertainty of Pigafetta's understanding (and not Lopes's original report).

Geographers in the late sixteenth century believed that a lake in today's Ethiopia was the source of the Nile, eventually determined to be Lake Tana.²⁶ However, quite independently of the European visitors to Ethiopia, Portuguese travelers had heard of an interior lake when in Kongo and thought it likely a source of the Nile. Already in 1491, when the first official Portuguese mission arrived in Kongo, they were aware of an inland lake. They believed that it came from the “Mountains of the Moon” often placed close to the lake that was the source of the Nile. When the mission left, a group they left behind were charged with looking beyond Kongo for “various other lands by Prester John and India.”²⁷

Definite evidence of Portuguese knowledge of an interior lake comes soon after. In his instructions to Simão da Silva, designated to lead a Portuguese group to establish a royal factory in Kongo in 1512, King Manuel I requested that they collect information about Kongo's neighbors and trading partners, as well as the lake. He even designated a group of people to devote themselves to this exploration.²⁸ Following royal instructions, in 1526, Baltazar de Castro, then living in Kongo and supported by Afonso I of Kongo, sought to make an exploration of the Congo River to reach the lake.²⁹ If an attempt was made, the limited documentation of the time does not mention it. And in 1536, Afonso ordered Manuel Pacheco to build boats to explore the river with this goal in mind.³⁰

The lake in question was quite likely Lake Mai Ndombe, whose location is appropriate for a border of Mwene Muji, and so it seems reasonable that Pigafetta had correctly heard about the lake and its connections to the empire.³¹ Pigafetta thought that it lay in the general region of (his presumed) upper Nile lake near Ethiopia where there were found “possessions of various lords and some obey Prester John and others the very great King Moenemugi.”³²

Drawing a line from the southern bank of Mai Ndombe to the site of Mbata or the Yaka Kingdom runs about 250 miles (or 400 kilometers), and hence dimensions of a substantial polity, probably large enough to be considered an empire, even if not quite as large as Kongo at the time. However, this estimate would be a minimum assessment, it might have extended even farther south and east, following the course of the Kasai, Kwilu, or Kwango Rivers. It was known in 1580, and could be assumed to have existed in 1567 in order for the Jagas to invade from there (a reported 1561 date on a map is probably spurious).³³ In the midst of these two end points is the center of the region that Vansina had proposed had developed political terminology consistent with a large and complex polity.

However, a second issue arises from this assertion: exactly how ancient can we expect this to be, given that Vansina's data suggested that linguistic innovations associated with such a polity might reach back as far as 1200, his proposed period for the emergence of his highest-order political units. At first glance it might seem that an early date of 1567 is quite late. The Portuguese, as we have already seen, were interested in the interior lake (Mai-Ndombe) from almost the earliest time of their arrival. Why wasn't Mwene Muji mentioned earlier?

In fact, in spite of Kongo's reputation as a very well-documented country, there is very little geographic information available from Portuguese sources for Kongo before the Carmelite mission and Lopes's embassy in the 1580s. Cartographic evidence makes this plain: early sixteenth-century cartography of West Central Africa revealed very little of its interior geography, confining itself to the mouths of rivers. Only the Kingdom of Kongo was labeled, and while the mouth of the Congo River was illustrated, the course of the river was not revealed.³⁴ Instead the central portions of the continent were either blank or filled in with geography based on the assumption that Ethiopia or the Mozambique region occupied the interior. These exaggerated states often intermingled with Classical theories, notably Agysimba (from Ptolemy's second-century geography) and even Biblical fantasies, like the Garden of Eden (or Terrestrial Paradise) or the fabulous Mountains of the Moon, all held to be near the source of the Nile.³⁵

However poorly the Central African interior was represented in maps, there can be little doubt that the Portuguese, at least, knew much more about African geography than was put on maps or published in literature. The Teke Kingdom of Great Makoko, Kongo's serious enemy to the east was mentioned in the earliest sources, but did not appear on a map until Pigafetta's in 1591. The vast destruction of the Portuguese archives in 1755 no doubt cost historians a great deal of sixteenth-century data on Africa. Manuel's instructions to Simao da Silva's 1512 mission to Kongo, for example, instructed him to determine the size of the country, the nature of its governance, and “what kings and lords border on him, their power and mode of living,” as well as their size, composition, and if they made war on Kongo or on each other.³⁶ No doubt such information was collected, probably by multiple people over the earliest years of their contact, but this information is now lost.

After Pigafetta, the Florentine Relation, and the related Vatican account, Mwene Muji is only mentioned one more time in European sources, based on geographical information collected by Dutch investigators during the time they occupied Luanda (1641–48). Their reports of the regional geography resulted in an account, probably originally compiled in the mid-1640s that became the common source for Isaac Vossius's 1666 and Dapper's 1668 descriptions of the region.³⁷ Vossius related that most of his information came from “the slave traders of Congo & Dangola” whose “greed has led them to go deeper and deeper in these regions.” For Dapper, an important source was the otherwise unknown voyage of Jan de Herder to Okanga at the same time. Both writers noted referred to Mwene Muji as “Nimiamye” which Sulzmann, based on Mázimawa André's testimony, identified as an alternate name for the empire — and a not second polity occupying a part of the space, as I have claimed recently.³⁸

Dapper and Vossius both also noted a new title (at least to geographers): “Girubuma” or “Giringboma” (Dapper) or “Giringboma” and “Gingirboma” (Vossius) in their descriptions. At the time of their informants’ visit in the 1640s, this was the powerhouse of the region, controlling some fifteen “kings” and one of the “mightiest kingdoms in Africa.” Giriboma was the title Ngeliboma, used by the ruler of the Boma Kingdom, which lasted until the European conquest in the early 1900s. Its continued presence, and the fact that the other titles, including Mwene Muji and Nimi a Mayi were relatively minor in the early twentieth century, made it appear the truly dynamic and advanced kingdom of the area. It was this kingdom which Vansina claimed was the eventual culmination of the political nomenclature he had identified for the region.

Thanks to the later dominance of the Ngeli of Boma, Vansina did not consider Mwene Muji to be anything like the empire that Pigafetta made it out to be. Perhaps this was because he saw the historical kingdom as the culmination of increasingly large polities following in stages, and he was inclined to see Mwene Muji as earlier and thus less significant than Boma. In Paths in the Rainforest he described “mwéné Mushie” as a “small chiefdom of the Nunu Mushie on the Kwa and makes it clear that centralized government was known in portions of the area by the sixteenth century.” His stepping stone idea was that “both the oral tradition and the later organization of the Boma Kingdom show that there were first small chiefdoms, then principalities, then a final conquest by the ruling dynasty of the Ngeliboma.”³⁹

It would not be until 1882, when Henry Morton Stanley passed through, that a literate witness would comment on this region again. Stanley noted the extensive Boma Kingdom, and the town of Mushie, providing the link to the title Mwene Muji. Shortly afterward, as the Congo Free State developed, Mushie became an important stop, and an assortment of Free State officials, traders, and missionaries began arriving, and accounts of the region become more plentiful, if not always very informative, about life and history of the area.⁴⁰

Oral Tradition and Mwene Muji?

Vansina's mention of the place of Mwene Muji and the Boma Kingdom in oral tradition now leads us to consider how it might fit into our understanding of the empire's role in the oral history of the lower Kasai region. Vansina cited René Tonnoir's extensive study of the oral traditions of the Boma Kingdom, which appears to provide, as Vansina claimed, evidence of a succession from small chiefdoms, principalities, and then final conquest by the Ngeli. But this idea came more from the politics of oral tradition in the colonial period than from an apparently detailed historical tale.

Tonnoir used a great deal of oral tradition collected by earlier colonial officials from the 1920s and added his own field research as a colonial official starting in the 1930s, to publish an extensive account of the Boma Kingdom in 1970, using as its title Giribuma borrowed from Dapper.⁴¹ The resulting work was both lengthy and complex, often quoting extensively from unpublished manuscripts or recordings of oral traditions.⁴²

Tonnoir's account, however, was shaped by the controversies that surrounded the establishment of Belgian rule and the struggle between various local powers to win a place in the colonial government as “chefs médailles.” At the present there are no known accounts of the traditional history of the region before these investigations to delineate local powers began in 1924.⁴³

Prominent in that dispute was the question of the extent of the power of the Ngeliboma, ruler of the Boma Kingdom. The first literate witnesses to visit the area, starting with Stanley in 1882, were under the impression that the ruler of Mushie, who they mistook to be queen Ngankabe (a regent during the minority of Muba, the eventual ruler of Mushie) was the ruler of the “Waboma,” or the whole kingdom as well. Others who followed, officials and missionaries, likewise understood that the ruler of Mushie was the ruler of Boma. Ngankabe indeed explicitly claimed she was “queen of all the Baboumas, my subjects,” in a treaty she made with missionaries on 27 February 1891.⁴⁴ As Boma's capital of Mbali was away from the river and so rarely visited, the claims that the leaders of Mushie ruled the whole area were accepted by the Free State authorities.

Just as the Free State took relatively little interest in the areas away from the river; even when the first colonial reports were written in 1919, the officials of the Belgian Congo did only short visits to these areas and collected what material goods they could, and questions of administration and government waited.⁴⁵ It was only after 1924 that officials took on the job of delimiting traditional authorities who would participate in the rule of the area. Given that the results of these investigations were likely to determine who would benefit from the decisions with regards to income and status, they would naturally lead to lengthy debates between the various traditional authorities. Thus, the first collections of oral tradition now known were gathered in a period in which important decisions concerning the status of many people was hardly propitious for a more impartial history that might have been collected by an early traveler or missionary where historical precedent and fiscal authority were not critically at stake.

The task of determining the political roles and fiscal claims of the areas fell to the territorial administrator Motte, who finished a detailed study of Boma landholding and political organization on 20 January 1926.⁴⁶ Motte's investigations, which began in 1924, relied heavily on oral traditions of authority and organization given by the various rulers, and were mediated by “indigenous tribunals.”⁴⁷ Motte took an extended trip through the country and interviewed high-ranking people who presented themselves as leaders. He collected information about the histories of Boma's constituent parts from Clément Mpanyi (who Motte called “Pagne”) who ruled the territory of Bomenzuri, a relatively low-ranking position, but who was highly regarded as a traditionalist. Motte also interviewed high-ranking officials; Mankutu, who as the head of the Ngeli clan and considered to be the effective leader of Boma, if not holding the title of Ngeliboma, and Bomele, who held the title Nkula and was a senior administrator. In general, one would expect Mankutu to become Ngeliboma and Bomele his companion and advisor. He was unable to interview “seriously” Bompaka, also a leading member of the Ngeli around Nioki, as he was sick.⁴⁸

Mpanyi was an advocate of a decentralized Boma Kingdom, which Motte put down to his position in a tributary district. “All of this must have bothered PAGNE in his self-esteem,” Motte wrote, because Mpanyi claimed that the Baboma maintained that each chief of the land was independent “and absolute master of his land owing tribute to no one.” To which Motte added that Mpanyi thought the Ngeliboma was chief “only of the village of Mbali.” He was such a strong advocate of this position that when errors and inconsistencies in his testimony were pointed out, he confessed that he had “‘forgotten’” [Motte's quotation marks] certain details. Thus, while Motte continued to rely on Mpanyi for this detailed knowledge of the settlement history of Boma, he stood that against the testimony of the higher-ranking Bomele and Mankatu, which he also quoted extensively and tended to prefer on larger issues.⁴⁹

On 25 November 1927, a bit over a year after Motte submitted his report, Mpanyi dictated a lengthy account of Boma tradition to Motte, which was written out in Lingala. It would appear to be a consensus anchored on Mpanyi's acknowledged expertise in the local patterns of settlement but representing a case for a unified Boma “chiefdom.” Administrators tended to handle this report as a sort of official version of the history of the Boma, and it was a critical part of Tonnoir's sometimes flowery account of Boma traditions.⁵⁰

The tradition, which was clearly stated in summary in Motte's 1926 report, had two parts. The first described how the Boma people occupied the area, following a group of leaders who had fled from lands further south along the Kwango River to escape oppressive demands of their elders to work in mines. They had settled in three successive waves, giving way to subdivisions within the group. The second part was then the tale of the conquest of the region by the Ngeli, a senior lineage from whom the younger and thus subordinate lineages had fled in their earlier homeland.⁵¹ Thus, while not claiming the foundation of Boma by the Ngeli, it gave them legitimacy by being elders of the original founders, as well as subsequent conquerors, thus lords by right of genealogy and conquest of Boma.⁵²

While Mpanyi's traditions, as eventually consolidated by 1927, were taken very seriously by Motte, he also had some challenges to them coming from groups that claimed to be outside the Ngeli migration. He noted that Bopili, the ruler of Kampe along the Mfini River, while claiming to stem from the original settlers (and hence to be Baboma) was of a sub-clan whose “real name” is “NTOTE, the same sub-clan as that of the Ntote of MA, the village of the sub-chief invested in Lofunga, who belongs to this sub-clan, just like the chief MUBA of Mushie.” Furthermore, he noted that “All come from the same sub-clan that seems to be a prominent, notable sub-clan that led the first Baboma penetrations into their present Lands.”⁵³ Mpanyi had not said anything about the Ntote in the first part of his tradition, and the discussion in the second part only dealt with its relations with the Ngeli, so Motte's determination must have come from other sources.⁵⁴

Mushie was at the time now ruled by Muba, who had also spoken to Motte, and his testimony swayed Motte into accepting an independent status for lands ruled by the various Ntotes. As Motte noted, the authorities were calling his people the Nunu, and also other designations such as Vanzale, not an ethnicity but an occupational descriptor, meaning “water people.”⁵⁵

Muba was born in the 1860s, and when Stanley visited Mushie, he met Ngankabe, Muba's regent, who impressed him as a “Martha Washington” sort of figure. It was from her claims that the Free State had assumed that Boma was in fact ruled by Ngankape, or rather the Ntote of Mushie.⁵⁶ While modifying Ngankape's claim to rule all the Baboma, Motte believed that the Ntote of Mushie, and other lands on the borders of Boma that were also ruled by Ntotes, so that the “Baboma of Mushie, the Banunu, will have to remain independent I think.”⁵⁷

Following up on Motte's work, Maurice Simon, Motte's superior, wrote a report on 21 June 1928 that summarized the nature of the dispute and, very briefly, the traditional accounts that made up the decisions.⁵⁸ Simon considered the question of the Ntote a puzzle, “Who are the Ntote?” he wrote, noting that there were four groups that were distant from each other, and scarcely knew of each other's existence, and had no recollection of a kinship connection. “One can wonder if these four families are really related, if the word Ntote is not an honorary title rather than the name of a clan.”⁵⁹

While convinced by the claims of Mankutu and Bomele that there was a unified Boma Kingdom, he yielded to Motte's conclusion that the Nunu-Ntote domain be exempted from the unified Boma chieftancy. As Simon saw it, “the Mushie branch, too far from the Ngeli zone of action and also more warlike (fights with the Bateke; the Kimi, the ‘Kundu’, the Europeans) never suffered from foreign influence. It never supplies warriors to the Ngeliboma. It managed to extend its domination beyond the land it originally occupied.”⁶⁰ He decided that the Ntote of Mushie and other lands held by people holding the Ntote title would have local rights but the overall rule would go to the Ngeli. But he also acknowledged that the Ntote territories had not been under Boma's authority and the decision was an administrative one not well supported by historical facts.⁶¹

When Tonnoir became administrator, he did additional research and wrote a report in 1932 which resulted in the colonial government giving greater supremacy to the Ngeliboma and Mushie was declared a sub-chiefdom, while Muba would retain his status as medaillé, “due to his long service,” but it would not pass on to his heirs after him.⁶² In the report he also weighed in on the origin of the Ntote, who “before the infiltration of the NGELI, some Baboma clans, settled along the right bank of the Mimi and the Kwa (between the mouths of the Borua-Mpê and Nzali-Mbali rivers), had been subjected to authority of the NTOTE a Banunu clan, who came from the IBIA region (right bank of the Kasai, upstream from Mushie).”⁶³ This was a sharp break from Motte's conclusion that the Ntote had actually led the first migration, and now put them in as late interlopers, albeit still in position before the Ngeli invasion.

While he did not elaborate on this simple statement in 1932, he would later write an extensive description of how the Ntote came to rule Mushie in his 1970 book, presumably using traditional material that had eluded Motte and Mpanyi. In defense of the idea of a single Boma Kingdom, he claimed that Mushie was “a village founded in the past by the BA-NUNU, but annexed and subsequently assimilated by the people of Kempani [a province of Boma] who, together with the BA-BOMA of the interior, considered the locality as an advanced sentinel watching over the Nzadi [Congo River system].”⁶⁴

The Ntote, Tonnoir claimed, had come in surreptitiously from their original homeland outside Boma's territory in the town of Ibia, some fifteen kilometers upstream from Mushie. There “according to the [unnamed] Baboma chroniclers” the Ntote “secretly decided to conquer them [the Boma] and developed a plan which was to deliver the country to them.”⁶⁵ This was to arrange to have Ngankoti, one of their most beautiful women, seduce Monsienga, the ruler of Kempani, and then use her wiles and influence to get her relatives installed in other points where the Ntote had claims, as along the Mfini River.⁶⁶ Matters finally came to a head when her daughter Mpaleke took over Kempani and the Boma (according to Tonnoir) revolted. In response, a “lightning intervention of a cohort of ‘Ntote’ warriors, coming from Ibia, called by ‘who-knows-who’, quickly put an end to this insurrectional interlude.”⁶⁷ He concluded by claiming that in order to maintain this apparently illegitimate claim, Muba had prevented European explorers from fully understanding Mushie's real role.⁶⁸

Muba, speaking to Tonnoir on other occasions about the tradition of the Ntote, probably restating what he had already said to Motte, contended that his people had settled Mushie first, and claims by the Ngeli elite that the land was empty when they came were false.⁶⁹ He simply said they came down the Kasai and proudly told Tonnoir that he was “the descendant of this distant Mpaleke [who had taken control of Kempani] by her daughter Ngaono, her granddaughter Mabomo-Kelii, then Mosamene, then… then…, etc… (about twenty generations remain).” Tonnoir concluded “such is the origin of the ‘Ntote’ of Mushie.”⁷⁰

Throughout the disputes, none of the disputants paid much attention to the written evidence from the sixteenth and seventeenth century. No one mentioned Mwene Muji, for example, even though it is clear, as we have seen, that it preceded the Boma Kingdom in the 1580s and probably exercised authority over it. Given the uncertainties of Mwene Muji's northern border, it is possible that Boma had not come under its authority, however tradition leans toward Mwene Muji's domination of it. It is equally clear that the Boma Kingdom had emerged as an important kingdom by the 1640s, when the Dutch sources mention it. This allows us to place the traditions in a proper order.

If we accept that Mwene Muji was in fact founded in and around Mushie (based on a convergence of names if nothing else), then it is easy enough to sustain Sulzmann's research that asserted the Ntotes laid claim to both Mwene Muji and Nimi a Mayi as their titles. It is unfortunate that Tonnoir did not report the traditions told to him by Muba more fully, although his interest in the claims of Boma surely explains the omission. Furthermore, the location of the lands controlled by Ntote titles in 1924 clearly border on most of Boma's territory, so much so that Tonnoir noted they “take on the appearance of an enormous pincer tightening around the GIRIBUMA.”⁷¹

While tradition prevented a straightforward interpretation, it seems likely the four Ntote groups not adjacent to Mushie had once been regional commanders who had managed to hold on to some authority and power even as Boma pressed on. Two of them, Mpe and Mushie (and adjacent lands that Boma claimed were theirs) had never fallen under Boma's control, just as the Dutch writers had noted, and the situation described by them in the 1640s had remained more or less fixed until they again appear in the historical record. The Ntote titles that had less memory of a connection to Mushie, it might be added, were not located on larger rivers (like the Kwa or Mfimi) and would be harder for the Mwene Muji to control with naval forces.

Working for a Chronology

In addition to the general facts alleged to be found in tradition, it could also add to the question of chronology. Mpanyi had provided a genealogy of the Ngelis, and this list could serve as a guide to chronology so that by calculating the sequence of mothers within the list of Ngelis there were thirteen generations since Maluma-Biene, the first Ngeli, came to power.⁷² Tonnoir, using twenty years as a guide to the length of a generation, concluded that Maluma-Bieme took power in the mid-sixteenth century.⁷³

Tonnoir's estimate is supported by a slender thread of documents, if we assume that the Mwene Muji of the sixteenth century was partially displaced by the Boma Kingdom by the 1640s. Mwene Muji was in full flower in the mid-sixteenth century (at least it is attested in 1567). Even in 1604, when Kongolese ambassador Antonio Manuel left Kongo and subsequently spoke of Mwene Muji absorbing the Jagas, there was no apparent split.

But this split must have happened soon after, for Portuguese writers living in the colony of Angola mentioned a never before noted kingdom in the east of Kongo that is probably the Ngeli dynasty's first appearance. The cloth trade with the great textile belt of Central Africa was critical to Portuguese success in acquiring slaves by selling luxury cloth from the very region east of Kongo where Mwene Muji lay, and so gave them good knowledge of this region.

According to a fiscal report concerning taxes on goods imported to Luanda in 1612, a number of eastern lands paid relatively few taxes, including the “Kingdom of Ibar.”⁷⁴ Ibar's neighbors included other areas on the eastern side of Kongo, such as the Great Makoko, Nsundi, Mbata, and Okanga. Ibar appears again in 1620 in a memorial written by Garcia Mendes de Castelo Branco, who had lived in Angola from the start of the Portuguese colony. When he spoke of inland trade, he noted that the Portuguese “trade at the kingdom of Macoco, and also the kingdom of Ibare.”⁷⁵ Mateus Cardoso, writing in 1624, also noted with regards to Kongo, “they have great forests which they call infindas, which the ancients left to serve them as fortresses, the principal of them are of Sonho, Bamba, and of Ibar.”⁷⁶

Vansina, who also knew these quotations, read Ibar as a generic term for the Congo River ebale, and placed it near the mouth of the Awumi River.⁷⁷ However, François Bontink had already made a better reading in 1970 of Ibar as “Mbali,” the capital of the Boma Kingdom.⁷⁸ Reading Ibar as Mbali suggests that this represented the emergence of Boma ruled by the Ngeli, for by the 1640s, the Dutch described both Ngeliboma and Mwene Muji coexisting. Given the closeness of fit between Tonnoir's genealogical estimate and these dates, using generations as a measure of time seems justified. Assuming the 13-generation depth of the Ngeli dynasty in oral tradition, and working with 1600 as a starting date, Tonnoir's generation length might be lowered from 20 to 18.5, probably a better length for a generation.

Tonnoir also tried to date the first settlement of Boma's founders, as Mpanyi had reported it in his first document. He developed a method of using village foundations to date a further thirteen generations back, giving a fourteenth century foundation date. But in 1983, Sulzmann wrote a detailed critique of Tonnoir's chronological methods (as well as his devotion to the Boma Kingdom), and challenged this part of his chronology, claiming that it was not correct to assume that the village elders in the early twentieth century had such great genealogical knowledge. She maintained that “the names of the village founders, who were the first owners of the land and handed down to their descendants, who pass on the rights,” only needed to have the founder's name or lineage, and “for this it is not necessary to know the complete chain of ancestors.”⁷⁹

A date for the founding of Mwene Muji might be possible if we use Muba's generation count instead, since his claim rested on an unbroken chain of people bearing Ntote status, and so could be dated directly using Tonnoir's generational method. If we accept a minimum of 24 generations (augmented a bit for the “about” in Tonnoir's report of Muba's testimony), and use 18.5 years per generation as established by the genealogies after 1600, it puts the founding of Mushie at about 1400, or somewhat earlier if more generations were to be added.⁸⁰

But before we become too convinced by generation counts, we should recall the controversy of the 1920s. It is quite possible that Muba, certainly aware that Mpanyi had supplied the authorities with a thirteen-generation Ngeli genealogy in 1926, simply decided to extend his own genealogy when speaking with Tonnoir in 1932. It is troubling in this regard that two earlier but nearly identical reports on the Nunu of Mushie by Hendrick and Hemeni in 1923 counted only nine rulers up to Muba, stretching back no further than the late eighteenth century.⁸¹ However, we can be quite certain that the list of rulers is not the same as that of generations, and Mushie definitely existed long before the earliest of the deeds of these rulers. In his own account of Boma's history Tonnoir admitted that many of the names of Mpanyi's genealogy had no associated information, and “did not leave a transcendent memory, to the point that their very names are sometimes, if not forgotten, at least confused with others.”⁸²

This controversy may well have played out also in Sulzmann's queries of Májimawa André in 1963. She reports with no further information about their discussion that the Ntote of Mushie claimed both the title mwenemuji and nimiamaia.⁸³ Given her earlier correspondence with Vansina, she might have asked Májimawa André directly about the titles recorded in Dapper. And, he may well have been aware of Dapper's text, as indeed some of the earlier claimants like Muba or Mankutu may have also known, or, recognizing the significance of the time depth, he laid the claim to them on Muba's behalf.

Towards a History of Mwene Muji

The Lower Kasai would be a likely center for a major political power. Mushie is on the Kwa River, which is close to where virtually all the branches of the western half of the Congo River Basin converge before entering its main course, making it the end of a major riverine transportation network (see Fig. 1, above). It seems likely that Mwene Muji could have been something of a dendritic polity, whose main authority was exercised by watercraft operating on the extensive river system it probably controlled, occupying positions and dominating polities on the riverbanks while not necessarily intruding into the interior between the rivers.

Figure 1. Map of Mwene Muji and surrounding area

Source: Created by Aharon DeGrassi

This dendritic polity was formed a bit after 1400, if we accept Muba's genealogy, and began expansion along the rivers, using a powerful navy with dominating trade as its principal motive force. That riverine focus in evident in Mwene Muji's alternate title of Nimi a Maye, recorded in Dapper, which was interpreted to Sulzmann to be “the kámba fish, a great river fish which is king… the king of the waters, the king of the fishes” and this mastery probably relied on fleets of large watercraft.⁸⁴

Large watercraft figured in Tonnoir's account of the alleged takeover of Kempani through the seduction of its leader Monsienga by the beautiful Ntote maiden Ngankote; his informants related that Masienga came to bring her to Kempani on an “enormous square-stern canoe, which is called ‘kekupi’” that had no less than thirty paddlers. As they paddled down the river Masienga and his “young wife stretched out under the canopy” in the square stern.⁸⁵ It was in a similar flotilla led by a twenty paddler kekupi that the Ntote Ndjuli imposed an unequal alliance on the lord of Maa sometime later in Tonnier's account.⁸⁶

Tonnoir then followed with a description of the old kekupis, largely put out of business by steamers on the rivers, but which could still be found here and there rotting away in his day.⁸⁷ He actually rode in a twenty-two paddle kekupi doing his duties in 1927–28, which he believed displaced two and a half (metric) tons, and an accompanying supply vessel of the same general dimensions displaced some three tons.⁸⁸

Diego de la Encarnacíon apparently also knew of these vessels, either from the unnamed ruler he catechized, or from seeing such craft on rivers near his post in Mbata around 1585. In the Florentine Relation he wrote there was “a tree with timbers so thick that from a single trunk they make a brigantine of twelve and fifteen oars per side [banda], which the blacks call Almadies; one brigantine made of one log can carry one hundred men and victuals for ten days.”⁸⁹ Mateus Cardoso, a Jesuit working in Kongo who described the country's borders in 1624, mentioned far eastern people who “who come in the lungos, which are great ships” which he believed were Portuguese vessels but may in fact have alluded to Mwene Muji's navy.⁹⁰ Dapper illustrated two boats matching the combined description of square-sterned vessels, even including a canopy, one with eight and the other with four paddlers to a side in his engraving of São Salvador. While it is impossible that such large craft would operate on that minor stream, it may be that the engraver, having a sketch of the boats, decided to place them there to include in the handful of illustrations he made.⁹¹

Only the enormous trees of the rainforest could be made to yield the hulls of these vessels, and so Mushie and lands around it had a clear advantage in constructing and using them over areas with less forest. Those trees stood along with the raffia palms of the Central African textile belt, producing luxury cloth from raffia fibers and exporting them southward, where they were received as far west as Luanda.

Thus, while Dapper noted the importance and power of the Ngeli of Boma, he also observed that the trade to Portuguese bases in Okanga came from Nimi a Maye (the alternate title for Mwene Muji). It was they who sold the cloth in Okanga, and had truces with the Teke Kingdom of Great Makoko to support trade.⁹² It was perhaps this long-distance trade, facilitated by riverine transport, that allowed people from Nimi a Maye visiting the Portuguese positions in Okanga to be able to describe, according to Dapper, the Portuguese possessions in Mozambique.⁹³

While a powerful fleet could dominate trade and the banks of the rivers, they probably did not bother too much with the interfluvial zones which were less vulnerable to naval assault than the riverbank settlements. This would explain the unusual situation of the Jagas before their invasion of Kongo as being in a province of Mwene Muji but not under his authority, though later conquered by him. Ultimately, Mwene Muji lost that tenuous control and the Yaka Kingdom emerged independent sometime before 1656. Similarly, Mwene Muji would be less able to control the area north of the Kwa because the rivers were less amenable to large watercraft and thus left the Ntotes of the north unsupported.

More light may be shed on the extent of Mwene Muji's influence or control from the work of the University of Ghent's BantuFirst archaeological project, which has been doing survey work along the banks of the Kwilu-Kasai watershed since 2018. In 2022, this project restudied the archaeological site of Mashita Mbanza, home of the traditional Pende Kingdom and noted for extensive earthworks that date from the seventeenth to nineteenth centuries.⁹⁴ In the 2023 season, according to BantuFirst member Peter Coutros, the work has “been able to identify the presence of several expansive political/social/economic entities along the Kwilu River and again near Idiofa. One is centered around Bandunduville, and another up the Kwilu near Lusanga – both of which seem to date between the [thirteenth and seventeenth] centuries.”⁹⁵ While still far from identifying these sites as being associated with Mwene Muji, they do suggest that at least some form of political complexity existed in a time frame consistent with Mwene Muji.

A Dutch map printed by Johannes Blaeu in 1662 but based on information collected by de Herder in 1641 in Okango, can help to define the situation further.⁹⁶ It shows a single unnamed river vaguely northeast of the named Wamba River (the border of Okango) which is likely the Kasai, the only southern tributary of the Kwa, the Kwilu having joined it a short distance south of its juncture with the Kwa and the town of Mushie. Informants from Mushie might have told the Dutch of locations up the Kasai from Mushie without elaborating tributaries. On the southern side of this river Blaeu placed a polity he called “Mopenda,” probably the Pende Kingdom associated with Mbanza Mashita.⁹⁷ If we imagine the maximum borders of Mwene Muji were equivalent to its known extent along the Kwango (that is, to the Yaka Kingdom), then extending it the same distance down the Kwilu, it would reach about as far as Mashita Mbanza.

On the other side of the river Blaeu placed a kingdom called “Mosongo.” Vansina proposed that this name could go with the Tsong people of today (on the Kwilu) but the name would also fit the Kuba, whose ethnic name is Bushong, on the Kasai.⁹⁸ The Mosongo might have been on the Kasai side of the Kasai-Kwilu juncture and across from Pende on the Kwilu side to de Herder's informants, thus figuratively on the other branch of the Kasai. The archaeological site of Idiofa is on the Kamtsha (tributary of the Kasai) and thus about as far along the Kasai's tributaries as the Yaka Kingdom is from Mushie. It might then reasonably have extended roughly the same distance up the Lukenye to the east, though there is no evidence from either cartography or archaeology for that region.

Lake Mai Ndombe was probably its northeastern extent if we consider that de la Encarnación believed that Mwene Muji shared domination of the lands around the great lake with another polity that he believed was Prester John's Ethiopia. That polity might have been Bozanga, noted in early seventeenth-century texts as laying in the same area as “Ibar” or the Boma Kingdom.⁹⁹ Mwene Muji's northern border thus probably included the whole of the Boma Kingdom best marked by the distribution of the Ntote titles Mpoka and Ikonia. Finally, on the west it probably shared a border with Great Makoko.

When Stanley arrived in Mushie in 1882, the situation had not changed much from Dapper's time. The rulers of Mushie claimed dominion over the whole region to its north, as Ngankape said, but Boma was definitely a power in the interior. If Mwene Muji had ever included Pende and Kuba in their realm, they had long since broken away, perhaps in the early seventeenth century when Boma and the Yaka broke free. In any case, when the steamers of the Free State began domination of the rivers, Mwene Muji lost control of trade and naval supremacy, and then when the deadly epidemics swept the area in the 1890s, they lost many of their people. Most of the survivors left for elsewhere and by the time the Belgians began collecting tradition, Mushie did indeed seem to be simply a fishing village with a small population and the grand claims of Muba and Ngankape to a once substantial empire were pushed aside.¹⁰⁰