Defenders of Liberty: The Congos and the Question of African Agency in the Haitian Revolution

In April 1802, the leaders of the Haitian Revolution faced a choice that would expose fundamental differences in their vision of liberty in post-emancipation society: join General Leclerc's army or continue the fight against French expeditionary forces. Henri Christophe, Toussaint Louverture, and Jean Jacques Dessalines agreed, whilst Kongo leader Macaya and Sans-Souci refused, the latter describing his forces as “defenders of liberty.”¹ Following his arrest, Louverture famously declared: “In overthrowing me, you have cut down in Saint Domingue only the trunk of the tree of the liberty of the blacks; it will grow back from the roots, because they are deep and numerous.”² The most obvious roots of liberty were Sans-Souci's mainly African troops, known as the “Congos,” who continued the revolution using tactics honed in 1791. But when Christophe, Dessalines, and Alexandre Pétion defected from the French and joined the newly-united “indigenous army,” the conflict over who would lead the “defenders of liberty” persisted, culminating in Henri Christophe's assassination of Sans-Souci, after whom he would famously name his palace.³ Sans-Souci's death foreshadowed a broader discursive victory in post-revolution historiography, which celebrated the role of Creole and mixed-race leaders, and the political ideology of the French Revolution, to the exclusion of African people and ideas.⁴

Following contemporary French sources and nineteenth-century Haitian intellectuals, scholars portrayed the “war within the war” in racial terms, pitting Creole leaders against the mainly African Congo troops. The contempt of the former for the latter survives in contemporary Haiti, where “Congo” means “uncivilized” or “savage,” denoting those who resist the coercive power of the centralized state.⁵ Adherents of this school assume the goal of the Haitian Revolution was to produce a French-style nation state, with free, if coerced, plantation labour. Using this metric, they judge the Haitian Revolution a failure, blaming the refusal of the formerly enslaved Congos to participate in a Creole-led plantation economy.⁶ This Eurocentric narrative not only silences, as Michel-Rolph Trouillot famously argued, the history of Haiti's Kongo founders, but it fails to explain the longue durée of Haitian history.⁷

By contrast, scholars especially from Haiti, portray divisions in the “war within the war” and the Haitian Revolution more broadly in economic terms: pitting members of the plantocracy, many of whom were not only free but owned enslaved workers, and the formerly enslaved, whose definition of “liberty” required escaping the horrors of plantation labor.⁸ These contrasting visions proved irreconcilable. As a result, post-independence Haiti was characterized not by plantation monoculture but by the lakou system of communal land ownership and local markets, what Jean Casimir terms a “counter-plantation system,” an “institutionalization of an alternative to large-scale agriculture.”⁹ From the perspective of the formerly enslaved, the Haitian Revolution was not just a war of independence but a more important “victory over slavery and the entire corresponding economic system.”¹⁰ At the heart of these contrasting visions of the Haitian Revolution are two radically different visions of liberty in post-emancipation society, raising the question of how African ideas and institutions influenced the political imaginaries of Haitian Revolutionaries. Scholars have argued that recovering the history of African people and ideas requires going beyond the silences and biases of colonial sources more broadly.¹¹

The latest front in this historiographical “war within the war” is David Geggus's “Kongomania,” which uses revised estimates of the number and origin of West Central Africans in Saint Domingue to present a forceful jeremiad against the role of the Congos in the Haitian Revolution. However, there are serious limitations to his argument, which uses the exact number or proportion of West Central Africans in Saint Domingue to measure their impact on the Haitian Revolution, an exercise that risks reducing the value of the enslaved once again to numbers, and assumes ethnicity as a static category somehow determinative of human history. Geggus's exclusive and uncritical reliance on European sources leads him to not only make incorrect conclusions about the origin of captives, but repeat negative tropes about Africans on both sides of the Atlantic. By privileging European documents and categories, the narrative not only fails to recover African ideas, but reproduces a racial interpretation of the Haitian Revolution. Geggus's assertion that though Sans-Souci's “revolt began what became the war of independence, most continued fighting not only the French but also the Creole generals… so great was their hatred of the Creole elite” (245) is the most recent example of a racial argument as old the Haitian Revolution itself.

The crux of the debate over African agency has coalesced around West Central Africans, the largest group of Africans in the colony. Geggus disputes estimates derived from the Trans-Atlantic Slave Trade Database (TSTD), Gabriel Debien, and Jean Fouchard, and argues the overall proportion of West Central Africans was smaller than previously imagined.¹² However, his estimates correlate with conservative estimates from the TSTD, a more representative and comprehensive measure of the composition of the enslaved population. Geggus appears to misread the summary of overall slave trade statistics in my dissertation as erroneously referring only to West Central Africans. From the period beginning at the end of the Seven Years War in 1763 until emancipation in 1793, French slave traders disembarked 474,895 captives—239,361 recorded from West Central Africa, or 50 percent of the total.¹³ When applied to Geggus's lower estimate overall proportion of the enslaved population born in Africa, the estimates produced mirror Geggus's findings—one half of one half.

These numbers could well have been higher. The TSTD covers an estimated 90 percent of French voyages; the place of embarkation is known for nearly 80 percent.¹⁴ This discrepancy may explain why, for some years, French records cite thousands more captives than TSTD.¹⁵ The database cannot account for French slave traders who under reported or failed to report voyages to evade customs duties, which records show increased after the 1784 repeal of the lucrative but corrupt acquits de Guinée and the 1789 prohibition of travelling directly from India to West Central Africa without stopping in France.¹⁶ Illegal or unreported voyages explain why, despite the fact the French had nearly exclusive control of trade on the Loango Coast after 1784, the number of captives purchased declines in TSTD. A more accurate statement given the limitations of the database is that between 1763–1793, at least 50 percent of captives disembarked originated in West Central Africa. Ignoring voyages for which no port of embarkation was listed, West Central Africans represented 57 percent of disembarked Africans. If the tendency of one-half holds for the missing ports, the figure would rise to 60 percent.

Geggus's estimates derive from a database compiled from plantation inventories, which must be used with careful source criticism, quantitatively and qualitatively. While the database he has created will no doubt advance scholars’ knowledge of the demographic contours of plantation societies, it is not possible to extrapolate reliable averages from the current sample. In order for an average to provide meaningful information, the dataset must be representative of the overall population and have a symmetrical distribution. Neither rule holds for the database Geggus has compiled. But the broader question remains of whether or in what ways these numbers and categories are useful. Do West Central Africans need to be the majority of the African population for their lives to be of interest to historians? There is a danger of judging the worth of the lives of the enslaved based on the metrics used by slave traders and owners, reducing human beings, once again, to numbers.

An equally significant problem is the categories used to compile the database, which derive from unreliable ethnonyms used by slave traders and owners. One example will suffice to illustrate the problem: the Mondongue. In the eighteenth century, the French believed the Mondongue came from the interior of central Africa, where they practiced infanticide, cannibalism, and human sacrifice. Europeans recognized them by their scarification and filed teeth. Moreau argued the enslaved “butchers of human flesh kept their odious inclination” for cannibalism in the colony.¹⁷ As evidence, he presented the 1786 case of an enslaved midwife accused of infanticide and cannibalism despite the fact there was no physical evidence of foul play, and moreover, colonial physicians had already attributed the symptoms of the deceased newborns to mal de mâchoir, or neonatal tetany.¹⁸ In fact, Moreau's description of the Mondongue quoted a popular but unreliable travel narrative of central Africa, which itself repeated tropes in earlier narratives, which were themselves rooted in ancient myths about Africa as inhabited by related groups of nomadic, cannibalistic barbarians.¹⁹

In fact, the Mondongue were neither a “nation” nor an “ethnicity,” nor is there any evidence for Africans eating human flesh on either side of the Atlantic.²⁰ In my research, I have shown the French ethnonym “Mondongue” derived from one of two West Kongo terms for captives in the eighteenth century: mvika and ndongo. ²¹ Mvika referred to captives who were abandoned for sale due to debt, poverty, or ambition. The word Ndongo signified captives sold by caravanners turned slave raiders whose commerce West Kikongo-speakers associated with warfare, chaos, and witchcraft. A Mu-Sombe referred simply to a bought person: somba means “bought.”

As for their geographical origin, linguistic evidence from eighteenth-century Saint Domingue suggests captives purchased by the French on the Loango Coast spoke West Kikongo languages, meaning they originated not from the southern state of Kongo or far interior but from the Loango coast polities and Mayombe rainforest hinterland—still a large area that could easily qualify as the “interior” especially to Europeans, who did not venture into the Mayombe rainforest until the nineteenth century.²² French slave trader Louis de Grandpré concluded that captives came from far away, yet he insisted they all spoke the same language.²³ This research is in line with recent scholarship on the impact of slaving on coastal African polities.²⁴

However, scholars of the Americas, unfamiliar with African history and the European genealogy of knowledge about Africans in the early modern period, can easily repeat the European myth of the Mondongue. The Mondongue even appear on eighteenth-century French maps, located in the interior of central Africa alongside other alleged tribes of anthropophagus: the Jagas, Zimbas, and Mumbos.²⁵ In the absence of eyewitness accounts, Enlightened cartographer Jacques-Nicolas Bellin explained, mapmakers relied on dubious travel narratives and ancient texts that posited the interior of Africa was inhabited by related groups of barbaric, nomadic man-eating tribes.²⁶ Following these sources, Geggus erroneously claims the “Mondongos” were a non-Bantu speaking ethnic group who “lived northeast of the Bakongo among the Tio” (225–226). The Mondongue are an example of why scholars must approach European categories about Africans with extreme caution. At best, the “nations” used by slave traders and subsequently slave owners vaguely indicated a broad linguistic region or port of embarkation. At worst, the categories relied on unsubstantiated myths found in dubious early modern travel narratives, the product of a genealogy of knowledge about the interior of Africa rooted in ancient Greek barbarian myths.²⁷

There are also problems with categories closer to home—that of the “Creole.” Geggus's argument for the numerical superiority of Creoles rests on the inclusion of children, who represented “close to half of enslaved Creoles,” meaning the number of Creole and Kongo adults “must have been quite similar” (22). Furthermore, “As the Congo's sex ratio was much higher that [sic] the Creoles’, it seems likely that the armies of the Haitian Revolution contained somewhat more Kongolese than Creoles” (243). There are several problems with Geggus's reasoning. Enslaved children more than likely had at least one African parent. As Geggus points out, Toussaint Louverture reportedly learned his Aja-Fon parents’ language. The survival of African vocabulary in Vodou chants, where they are known as langaj, provides evidence African parents transmitted linguistic and cultural knowledge to their children, raising the question of how or in what way Creole is a useful marker of difference.

No doubt, children born in the colony had a greater chance of learning Kreyòl and perhaps even to speak or, less likely, write French. The advantage is reflected in the fact that the leadership of the Haitian Revolution, who survive in French archives, were, largely, those who could correspond in writing with colonial authorities. However, as Jean Casimir, Gérard Barthélemy, and others have shown, the Kreyòl language served, after independence, not as the language of the Francophile elite but as “the primary instrument for the construction and preproduction of this sovereignty” of the counter-plantation system.²⁸

There is a danger in repeating the widespread bias of French slave owners and intellectuals who categorized enslaved children born in the colony as racially distinct from and superior to their African parents. Following the European belief that climate as well as “race” determined physical, moral, and intellectual characteristics, Moreau asserted “The Creole nègres are born with physical and moral qualities, which gives them a real right to superiority over those who have been transported from Africa.”²⁹ The idea of Creole Degeneration (or generation, as Moreau applied it to the enslaved) has happily been relegated, alongside the idea of biological race, to the dustbin of historic scientific racism.

There is no evidence of cultural (as opposed to economic) distinctiveness cohesive or widespread enough to warrant the thorny term “ethnicity” in reference to enslaved children born in Saint Domingue. On the contrary, as one Vodou chant asserts, “I am a Kongo Creole.”³⁰ More broadly, the Petwo rite of Haitian Vodou, long considered Creole, is now recognized to have Kongo origin.³¹ These examples serve as reminders that colonial identities were fluid and not mutually exclusive, suggesting scholars need to move beyond ethnicity, rooted in static notions of African culture, as a meaningful category of difference.

The goal of recovering the African history of the Haitian Revolutionaries is not to revive the French thesis of race war, nor to view African origin as deterministic of future choices or actions, but instead to recover alternative political imaginaries for the Haitian Revolution. However, the silences and biases of the colonial archive represent a fundamental challenge for scholars of Africans in the Atlantic world. Haiti's Kongo founders rarely appear in the written archive on either side of the Atlantic. There are no written sources documenting the interior slave trade to the Loango Coast. After enslavement, Africans survive in unreliable accounts of slave traders and owners. As a result, when Congos do appear in the archive, it is often as slaves, Bossales, savages, or cannibals. Once in Saint Domingue, and especially during the Haitian Revolution, many Kongo and other formerly enslaved people created communities away from the plantation in maroon settlements and later on the lakou. How do we recover the history of people who spent much of their lives outside of or maligned by the gaze of Europeans? How do we use archives responsibly, without reproducing the problematic power dynamics, or reducing African people to European stereotypes?

In my work, I approach this problem with a multidisciplinary methodology to identify, with greater specificity than possible with written documents alone, the geographic origin of captives sold on the Loango Coast. I employ a historical linguistic methodology known as a borrowing study, or “words and things,” tracing the parent language of the words the Kongo and their descendants used in Saint Domingue and independent Haiti. Though French slave traders insisted the captives they purchased originated from cannibalistic “nations” in the interior of Africa, linguistic research reveals that captives spoke the West Kikongo languages of the Loango Coast and Mayombe forest. This does not preclude captives originating from further inland, but, if they did, there is scant evidence outside the derogatory ethnonyms of Europeans.

Understanding how the West Kongo responded to enslavement on both sides of the Atlantic requires recovering the African history from the local point of view. The West Kongo were not “Atlantic Creoles” nor did they originate from European-style hierarchical kingdoms. The West Kongo understood polities as heterarchical composites of visible and invisible agents woven together in a complex social fabric.³² They ensured the prosperity—defined as health and wealth—of the group through a “communal ethos” that emphasized reciprocity, autonomy, and redistribution, ethical values central to the creation of the lakou in Haiti.³³ In Saint Domingue, enslaved Kongo men and women drew on specific West Kongo ideas and practices to address the material, spiritual, and biological affliction of enslavement. During and after the revolution, the Kongo created new communities using West Kongo public healing practices, the building blocks of a new social order: the “counter-plantation” system of communal land tenure regulated, as on the Loango Coast, by a public healing institution known in Haiti as Vodou.³⁴ From this point of view, the Kongo, as a people and a set of ideas, were the “Defenders of Liberty” of the Haitian Revolution.

However, the creation of the Haitian Revolution's most radical and long-lasting transformations was not discretely Kongo or African, but was “Kongo Creole.” The history of the Kongo is a reminder that to reconstruct the African history of the Haitian Revolution, scholars must move beyond static notions of identity and culture rooted in early modern notions of race and ethnicity. If the West Kongo represented a linguistic group and cohort of enslaved people, in Saint Domingue, Kongo became an identity, a lexicon of power, and a set of social and political ideas that transcended geographical origin.

The debate over the role of the Congos in the Haitian Revolution acts as a cipher for debates over the role of non-European ways of knowing and being, including the role of marroons, Haitian Vodou, and Africans.³⁵ As we have seen, if the debate is framed as quantitative, it is fundamentally qualitative, reflecting colonial assumptions about what individuals and groups are capable of producing social action and political philosophy.³⁶ Geggus's characterization of “Africans and children of Africans” as having “in their native languages … no word for ‘liberty’” is an example of European prejudice against African political ideology, the basis of a racial interpretation as old as the Haitian Revolution itself.³⁷ If Trouillot's work has drawn attention to the silencing of the Congos in the archives, “Kongomania” is a timely warning that what is contained in the archives can be just as problematic as what is absent. Only by moving beyond the European archive can scholars hope to avoid repeating the discursive violence of the past and recover the history of the most radical ideas and proponents of “liberty” in the Haitian Revolution.