About Bouar Megaliths of Central African Republic

Across the central plateau of the Central African Republic, between the Lobaye and Mambéré drainages and approximately 350 km north-northwest of Bangui, an archaeological complex of roughly seventy clustered groups of standing stones extends across an area sometimes given as 7,500 km² in the UNESCO tentative-list dossier (file 4003, inscribed 11 April 2006). The complex is centered on the town of Bouar in Nana-Mambéré prefecture, with outliers extending westward into the highlands of north-western Cameroon. Local Gbaya-speakers call the monuments tazunu (sometimes transliterated tajunu, glossed "standing stones"). The earliest sustained archaeological documentation appeared in Pierre Vidal's La civilisation mégalithique de Bouar: prospections et fouilles 1962–1966 (Études oubanguiennes 1, Firmin-Didot, Paris, 1969). Later work by Nicholas David and by Etienne Zangato of the Université de Paris X-Nanterre revised both the chronology and the interpretation, producing the picture used by mainstream Central-African archaeology since the 1990s.

The Bouar Plateau

The Bouar plateau is a granitic peneplain on the southern flank of the Adamawa highlands, sitting between approximately 950 and 1,100 m elevation. The bedrock outcrop is the source of the laterite-capped granite blocks used in the monuments. UNESCO's 2006 dossier describes a zone roughly 130 km long by 30 km wide, with the seventy or so distinct megalithic groups distributed across watersheds that drain south into the Sangha and Ubangi systems. Each cluster is situated, according to the same dossier, at the head of a watercourse and oriented either toward the east or in the direction of the running water — a recurrent placement that has been read as either hydrographic, ritual, or both, and is the closest thing to a documented spatial regularity in the corpus.

Vegetation at Bouar today is a savanna mosaic with riparian forest galleries, sitting on the northern margin of the Congolian rainforest block. Palynological reconstructions of the central-African Holocene — the synthesis assembled by Bostoen, Clist, Doumenge, Grollemund, Hombert, Muluwa and Maley in Current Anthropology 56(3), 354–384 (2015) — show a forest-perturbation phase beginning around 4,000 BP at the rainforest periphery and reaching its core around 2,500 BP, with mosaic landscapes of open forest and grassland savanna opening corridors for both farming populations and large-scale earthwork projects. Bouar sits squarely inside that perturbation belt, both geographically and in time.

The Gbaya, who today are the dominant population around Bouar, are speakers of an Adamawa-Ubangian (Niger-Congo, non-Bantu) language and arrived in the region only in the 16th century CE, several millennia after the monuments were built. They preserve no oral tradition claiming descent from the builders. The same is broadly true of neighbouring Mboum, Sara and Banda communities. The cultural identity of the builders is therefore not directly recoverable through ethnographic continuity, and every interpretive claim about ritual, calendrical, or astronomical purpose passes through that gap.

Sixty Centuries of Use, or Fifteen?

The chronological literature is unsettled and presents two roughly competing frames. Vidal's 1969 monograph, working from typological and stratigraphic comparison without a developed radiocarbon series, proposed a late-Neolithic horizon broadly compatible with the c. 3500–2700 BCE range cited by UNESCO's tentative-list entry. Nicholas David's 1982 Azania 17(1): 43–77 article Tazunu: Megalithic Monuments of Central Africa excavated two tumuli incorporating megalithic features and reassessed Vidal's stratigraphy in light of new radiocarbon determinations, concluding that the principal monumental phase falls in the first millennium BCE rather than the late Neolithic.

Etienne Zangato's 1991 doctoral thesis at Nanterre (Étude du mégalithisme dans le nord-ouest de la République Centrafricaine) and the published synthesis Sociétés préhistoriques et mégalithes dans le Nord-Ouest de la République Centrafricaine (Cambridge Monographs in African Archaeology / BAR International Series, Archaeopress, 1999) refined this picture into three cultural periods, with calibrated ranges given in the monograph running from approximately 2135–1612 cal BCE to 1263–932 cal BCE for the earliest documented occupation, and a megalithic florescence between roughly 950 cal BCE and the second millennium CE. Iron metallurgy appears in the sequence from the 9th century BCE onward, a finding consistent with David's first-millennium framing for the densest building activity.

The popular framing that the Bouar megaliths predate Stonehenge depends on which dating one accepts. Under Vidal's late-Neolithic proposal (or UNESCO's c. 3500–2700 BCE summary), the earliest stones at Bouar would be roughly coeval with or older than the earliest sarsen settings at Stonehenge (c. 2500 BCE). Under the David and Zangato chronology, which is the one accepted by current Central-African archaeology, the bulk of the monuments are 1,500 to 2,500 years younger than that and post-date Stonehenge's main phases. The conservative scholarly statement is that the Bouar complex is one of the oldest and largest concentrations of standing stones in sub-Saharan Africa, with first-millennium-BCE peak construction and possible earlier antecedents that are not yet well dated.

The Quarry, the Block, the Mound

The construction sequence at Bouar is reconstructible in broad outline from the excavated sites, though the details differ from cluster to cluster. The granite of the plateau outcrops at the surface in sheets and tors, and the builders worked the natural fracture planes rather than detaching blocks from a face. Zangato's 1999 monograph identifies several probable quarry zones within two to four kilometres of the major monument groups, where the density of split blocks, wedge marks, and rejected outliers indicates extraction activity rather than naturally weathered scatter. The largest single uprights at the principal clusters are in the 2.5 to 3.5 metre range, with weights estimable at one to four tonnes — within the range of village-scale transport using rollers, sledges, or the levered-pivot method documented for comparable African monument traditions.

Once at the construction site, the laterite rubble fill was raised first as a low mound, and the granite uprights were set into prepared sockets either at the perimeter or at the apex. The stratigraphy David documented in 1982 at his two excavated tumuli suggests a single-episode construction rather than a long accretion: the laterite layer is compact and homogeneous, with no buried palaeosols or stratigraphic breaks that would indicate phased building over multiple seasons. This is the structural basis for David's inference that a village labour force could have raised a single tumulus in a single dry season, with the dry season at this latitude running roughly November through March.

The architectural vocabulary is restricted. There are no megalithic chambers in the Western European or Saharan sense — no dolmens, no passage graves, no covered cists. The Bouar tumulus is fundamentally a mound with peripheral or capping uprights, not a constructed interior space. Some clusters show concentric rings of secondary stones around a central tumulus, and a few have aligned uprights extending outward from the main mound, but the variations are local elaborations on a single underlying form rather than distinct architectural traditions. Zangato's 1999 typology recognizes three main variants on this theme, distinguished by the relative prominence of perimeter ring, capping uprights, and outlying alignments.

What the Excavations Found

The monuments themselves take a recurrent form. A typical tazunu is a tumulus of laterite rubble, roughly 10 to 30 m in diameter, contained or crowned by upright granite blocks ranging from less than a metre to over three metres tall. Vidal's 1962–1966 prospections documented approximately seventy distinct groups across the prefecture, with concentric arrangements visible at the better-preserved sites and traces of perimeter alignment at others. The granite was quarried locally, with several outcrop quarries identified by Zangato within a few kilometres of the largest groups, indicating an economy of effort that constrained how far material was moved.

Excavation has consistently failed to recover articulated human burials inside the tumuli. David's 1982 Azania report, which excavated two tumuli to bedrock, recovered no skeletal material; Zangato's later sondages and full excavations across multiple sites repeated that result. Small finds within the monuments are limited to lithics consistent with a Late Stone Age and early-Iron-Age technological level, occasional pottery sherds, and ferruginous concretions. The absence of human remains has driven the cautious phrasing in Zangato's 1999 monograph, where the function is described as memorial or commemorative rather than directly funerary — the monuments behave like cenotaphs more than like graves.

The stratigraphy at several sites is shallow, with the principal building event apparently compressed into a single construction episode. David inferred that an individual tumulus could have been raised by a village-scale labour force in a single dry season, which constrains the social-organization model: the builders need not have constituted a large or hierarchically complex polity, only a coordinated village-cluster economy able to mobilize seasonal labour. The total monument count across the seventy clusters is, on Zangato's reckoning, in the low thousands of individual standing stones.

The Eastward Orientation Problem

The closest thing to a documented spatial regularity at Bouar is the observation, repeated in Vidal's 1969 prospection report and codified in UNESCO's 2006 tentative-list description, that the megalithic groups are oriented toward the east, or alternatively toward the running water of the source they sit above. This is a generalization across seventy clusters, not a per-site survey of azimuths, and it has not been published in the peer-reviewed archaeoastronomical literature with measured declinations, sightline analysis, or precision estimates.

An east-facing orientation is consistent with sunrise as a directional referent, but "east" in this loose sense is also consistent with any number of non-astronomical referents — prevailing wind, slope, water flow, the direction of arrival of the builders' ancestors in regional cosmologies that often code an eastward homeland. The Gbaya themselves, who name the stones, have no traceable astronomical doctrine tied to the monuments; their ethnographically attested ritual calendar centres on agricultural cycles tied to the bimodal rainy season rather than to solstitial or equinoctial sunrise.

Until a systematic azimuth survey with calculated declinations is published — the kind of work Clive Ruggles and Bernardo Mendonça Lima have done for the Carnac alignments in Brittany or for the Brazilian megalithic site of Calcoene — the eastward-orientation claim should be treated as a typological observation about the corpus rather than as evidence of astronomical intent. Wikipedia's List of archaeoastronomical sites by country and the broader review literature catalogue Bouar as a site meriting investigation, not as a site with demonstrated alignment.

Astronomical Hypotheses Tested and Untested

Several speculative hypotheses circulate about the astronomical content of the Bouar complex. None has been published in a peer-reviewed archaeoastronomy journal with measured azimuths, atmospheric-refraction corrections, and statistical tests against the null of random orientation. The principal hypotheses, in roughly decreasing order of plausibility given the available evidence:

The first is solstitial sunrise. At Bouar's latitude (approximately 5°57' N), the summer-solstice sunrise rises at an azimuth of roughly 66° from north and the winter-solstice sunrise at roughly 113°; the equinoctial sunrise sits between, at 90°. A loosely "east-facing" complex is compatible with any of these, but discriminating among them requires per-site azimuths and horizon profiles — work that has not been published. The hypothesis is testable; the data are not in the literature.

The second is heliacal-rising orientation, in which an alignment is built to a specific star's first annual appearance before sunrise. This is the hypothesis that has been most rigorously confirmed for other African sites (Nabta Playa in Egypt, dating to before 5000 BCE, is the canonical case, with peer-reviewed orientation work tied to circumpolar and equatorial stars). For Bouar, no such analysis has been published. The orbital precession of the equinoxes (P 25,772 years) means that any star alignment proposed for c. 1000 BCE must be back-calculated to its position then, not to its current rising azimuth, which is a frequent error in popular treatments.

The third is lunar standstill alignment, which has been documented at sites such as Callanish in Scotland and explored at Stonehenge. The major and minor standstill cycles of 18.6 years produce extreme rising and setting azimuths that, in northern temperate latitudes, are striking. At Bouar's near-equatorial latitude, the lunar standstill range is compressed and harder to read in the landscape, which makes lunar-standstill hypotheses inherently less compelling here than at higher-latitude sites.

The Limits of Inference Without Ethnographic Continuity

What separates Bouar from Mesoamerican, Egyptian, or even ancient British archaeoastronomy is the absence of textual or oral-traditional continuity between the builders and any living community. At Chichén Itzá, the Yucatán Maya kept ritual practice and codical knowledge into the contact period; the Dresden Codex preserves the Venus tables that allow modern researchers to verify alignment hypotheses against a documented astronomical doctrine. At Stonehenge the situation is worse, but the prehistoric British landscape is densely surveyed and the corpus of comparable circles is large enough to support statistical reasoning about orientations. At Bouar, neither resource exists.

The Gbaya term tazunu is a name applied by a population that arrived three millennia after the monuments were built. The original builders' language is unrecoverable; the dominant pre-Bantu language families in central Africa at the relevant period were Adamawa-Ubangian and earlier substrates whose phonology and lexicon are reconstructed only fragmentarily. There is no surviving body of myth or ritual practice that can be assumed to descend from the builders — every interpretation of the monuments' meaning is therefore an inference from the physical evidence alone.

This is not a counsel of despair. Comparable rigour has been brought to the Senegambian stone circles documented at Wassu and Sine Ngayene, to Göbekli Tepe in Anatolia, and to the Calcoene alignment in Brazil, all of which lack ethnographic continuity. The procedure is the same: measure orientations across the whole corpus with calibrated instruments, calculate the expected distribution under a null hypothesis of randomness, and test whether the observed orientations are consistent with any astronomical target visible from the site horizon at the relevant epoch. That work is what is currently missing for Bouar.

Where the literature is honest, it says so. Zangato's 1999 monograph does not claim astronomical alignment; UNESCO's tentative-list dossier records the eastward orientation as a defining typological trait but stops short of an astronomical reading; David's Azania paper restricts itself to chronology and burial-function inference. The gap between this scholarly reticence and the popular framing of Bouar as "Africa's oldest astronomical observatory" is wide, and the gap is doing most of the rhetorical work in the popular sources.

The Northern-Cameroon Extension

The Bouar complex does not terminate at the modern Cameroon–Central-African-Republic border. Scott MacEachern's 1994 Antiquity 68 paper Interpreting Standing Stones in Africa: A Case Study in North-West Cameroon documented a comparable set of standing-stone sites in the Diamaré and Mandara highlands immediately to the north-west, and argued for a single broader monumental tradition extending across what is now the international boundary. These sites share the laterite-and-granite construction, the absence of buried bodies in the excavated examples, the village-cluster placement at watershed sources, and the broad first-millennium-BCE chronology.

MacEachern's contribution is methodological as much as descriptive. He pointed out that the interpretive frame Western archaeology brought to African standing stones — funerary monuments as expressions of hierarchical social organization, on a pattern abstracted from European megaliths — does not necessarily fit the African evidence, where labour-mobilization capacity does not require state-level hierarchy and where the absence of bodies in many monuments resists straightforward burial reading. The Bouar–north-Cameroon corpus is, in MacEachern's framing, a case study in how a sub-state agricultural society organized memorial labour without producing the archaeological signature of stratified rank.

This trans-frontier scale is consequential for the question of builder identity. A monumental tradition extending across what is today three or four hundred kilometres of upland implies a population network larger than any single village cluster, and the linguistic and demographic candidates that fit such a network in the first millennium BCE include the eastern Adamawa-Ubangian-speaking groups and the pre-Bantu agricultural substrate of the central-African belt. None of these populations is identifiable from the archaeological evidence with the kind of precision that historical-linguistic reconstruction or ancient-DNA work could supply, but the corpus scale rules out the most localized interpretations.

Where Bouar Sits in the Comparative Record

Among African megalithic complexes, Bouar is distinctive for three reasons. First, it is large — the seventy clusters and several thousand individual stones rank it alongside the Senegambian circles at Wassu and Sine Ngayene as one of the densest concentrations on the continent. Second, it is old — even on the conservative David and Zangato chronology, the principal building horizon at c. 1000–500 BCE makes it older than the Senegambian complexes (which Sutton and others have dated to c. 300 BCE–CE 1500). Third, it sits in the cultural void between the Saharan megalithic tradition to the north (Nabta Playa, Adrar Bous) and the southern Bantu-speaking world, which has no comparable monument tradition.

The Senegambian circles (Wassu, Sine Ngayene, Kerbatch, Wanar) share with Bouar the absence of human remains in many monuments and the use of locally quarried stone, but their laterite-pillar form and the larger number of monuments per site distinguish them. The Saharan tumuli of the early Holocene are funerary in clear cases. The northern-Cameroonian standing-stone sites surveyed by Scott MacEachern (1994 Antiquity) are immediate neighbours of the Bouar complex and the most relevant comparison — same broad period, same general form, same Adamawa highland setting, possibly the same builder culture extending across what is now the Cameroon–Central-African-Republic border.

The absence of comparable megaliths elsewhere in the Central African Republic and in the Congo basin proper sharpens the question of why this corner of the Adamawa plateau produced a monumental tradition. Two factors converge: the granitic geology supplies workable stone in a region where most landscapes do not, and the perturbation of the rainforest documented by Bostoen and colleagues opened savanna corridors at exactly the right time for village-cluster economies to take hold. The monuments may be the visible record of communities crystallizing around those new ecological openings, with the eastward orientation reading as either solar, hydrographic, or both — a question that future surveys will have to settle.

Sixty Years of Field Seasons

The history of fieldwork at Bouar is short by comparison with the European or Egyptian megalithic literatures, and the brevity is itself part of the explanation for the interpretive gaps that remain. Before Vidal arrived in 1962, the standing stones had been noted by colonial-era explorers and administrators — passing references occur in French ethnographic reports from the 1930s and 1940s — but no systematic recording had been undertaken. Vidal's four field seasons between 1962 and 1966, published in 1969 as La civilisation mégalithique de Bouar, produced the first cluster map and the first typological inventory, and remain the baseline for any subsequent work.

The next sustained programme was Nicholas David's work in the late 1970s and early 1980s, culminating in the 1982 Azania 17(1) paper that excavated two tumuli to bedrock and produced the radiocarbon series that revised the chronology. David, then at the University of Calgary, framed the Bouar work within his broader interest in central-African Late-Stone-Age and early-Iron-Age archaeology, and the Azania paper sits in dialogue with his contemporaneous publications on northern Cameroon and the Mandara highlands.

Etienne Zangato's research programme, conducted from the Université de Paris X-Nanterre and unfolding from the late 1980s through the 1990s, was the most thorough single body of work on the complex. His 1991 doctoral thesis was followed by the 1999 BAR monograph and by a series of journal articles documenting specific sites and the calibrated radiocarbon sequence. Zangato's work is what current Central-African archaeology means when it cites "the Bouar chronology," and his three-period framework (a pre-megalithic phase, a megalithic florescence, and a post-megalithic continuation overlapping the Iron Age) is the operational reference.

Field seasons since 2000 have been irregular and constrained by political conditions. The Central African Republic's political instability since the mid-2000s, including the 2003 coup, the 2012–2014 conflict, and the prolonged insurgency in the western prefectures, has made sustained foreign archaeological work difficult. Sporadic surveys and documentation projects have continued — the UNESCO tentative-list dossier was prepared in 2006, and several Central African Republic and French researchers have published shorter pieces — but the kind of multi-season excavation programme that would advance the chronology or test the orientation question has not been mounted.

Conservation, Access, and the Slow Recovery

The site is not formally inscribed on the World Heritage List; it has been on the tentative list since April 2006, and the dossier remains a tentative-list submission rather than a formal nomination. Political instability in the Central African Republic since the mid-2000s, including the 2012–2014 civil conflict and intermittent insecurity in the western prefectures, has impeded both the formal nomination process and field archaeology more broadly. Recent fieldwork has been limited; the most systematic body of evidence remains Zangato's 1991–1999 work.

The monuments themselves are physically robust — large stones laid into laterite rubble are difficult to damage casually — but they are vulnerable to opportunistic stone-robbing for construction material, to bush-fire damage of the surrounding vegetation that obscures their setting, and to the erosion of context as roads and settlement patterns shift. The clusters near Bouar town are the most studied and the most exposed; outlying groups are less documented and less disturbed. A systematic archaeoastronomical survey of the kind described above would be the natural complement to a renewed formal nomination, and is the work most likely to produce a sharper picture of the builders' intent if it is ever undertaken. Until then, the corpus stands as the most consequential under-studied megalithic tradition in sub-Saharan Africa — a baseline established by Vidal, refined by David and Zangato, and waiting for the next generation of fieldwork to push the chronology, the orientation question, and the builder-identity problem past the cautious framing the current literature can support.

Purpose

memorial / commemorative (Zangato 1999); no human remains recovered

Precision

no per-site azimuth survey published

Modern Verification

David 1982 Azania 17(1):43–77; Zangato 1991 (Nanterre thesis), 1999 (BAR); palaeoclimate context Bostoen et al. 2015 Current Anthropology 56(3):354–384

Significance

The Bouar megaliths are the largest and oldest documented megalithic complex in central Africa, with approximately seventy distinct clusters and several thousand individual standing stones distributed across roughly 7,500 km² of the Adamawa plateau on the Cameroon–Central-African-Republic border. UNESCO placed the site on the World Heritage tentative list on 11 April 2006 (file 4003), recording a date range of c. 3500–2700 BCE drawn from Pierre Vidal's 1969 prospection report. The principal building horizon, on the revised chronology of Nicholas David's 1982 Azania reassessment and Etienne Zangato's 1991–1999 work at Nanterre, falls in the first millennium BCE, with possible earlier antecedents in the third and second millennia BCE.

The site is consequential for three converging reasons. First, it documents organized monumental construction by a non-Bantu, pre-iron-working population in a region of central Africa where almost no other monument tradition exists. The builders predate the Bantu expansion's main rainforest-route pulse (Phylogenetic work by Grollemund et al., PNAS 2015) and may belong to an Adamawa-Ubangian or earlier substrate population whose linguistic and ethnic identity is irrecoverable through ethnography. Second, the corpus exhibits a consistent eastward (or hydrographic) orientation across seventy clusters, the closest thing in the central-African record to a candidate archaeoastronomical regularity — though the systematic azimuth survey that would test the astronomical hypothesis has not been published. Third, the dating brackets the rainforest-perturbation phase documented by Bostoen et al. (Current Anthropology 56(3), 354–384, 2015), placing the monumental tradition exactly in the ecological window when savanna mosaics opened across the central-African belt and village-cluster economies could mobilize seasonal labour for large constructions.

The site is also a case study in the limits of inference. With no human remains recovered from the tumuli, no documented oral tradition tying living populations to the builders, and no published azimuth survey, every interpretive claim — funerary, calendrical, astronomical, ritual — is at the level of hypothesis. The honest scholarly framing of Bouar is therefore as a corpus whose physical regularities suggest deliberate placement and orientation, whose dating places it among the oldest such corpora in sub-Saharan Africa, and whose meaning remains open to the kind of systematic comparative work that has begun to be done at the Senegambian circles and at Göbekli Tepe. The popular framing of Bouar as "older than Stonehenge" is defensible only under the loosest version of Vidal's dating and is rejected on the David and Zangato chronology accepted by current Central-African archaeology.

Connections

The Bouar complex sits in conversation with several other archaeoastronomical sites surveyed elsewhere on this site. The closest formal comparison is the Wassu and Senegambian stone circles in the Gambia and Senegal, another large sub-Saharan corpus built by a pre-state agricultural society, surveyed for orientation regularities and treated as funerary-commemorative.

For the absence of ethnographic continuity problem, the Anatolian site Göbekli Tepe presents the same interpretive challenge in a far older form (c. 9500 BCE), and the methodological response — systematic corpus-wide orientation survey — is the same approach Bouar awaits. The Egyptian site Nabta Playa is the canonical case of African archaeoastronomy with confirmed alignments, dating to before 5000 BCE, and provides the methodological benchmark against which Bouar is conspicuously under-studied.

For the eastward-orientation problem and the question of solstitial versus equinoctial alignment, Stonehenge is the most heavily surveyed comparable case in temperate latitudes; the latitude difference is consequential, and the contrast is instructive. For star-rising alignments specifically, the broader treatment at heliacal rising sets out the procedure that would be applied to Bouar if a systematic survey were undertaken. The general framework for distinguishing genuine astronomical alignment from coincidental cardinality is covered at precession of the equinoxes, which is the correction any back-calculation to c. 1000 BCE would have to apply.

Further Reading

  • Vidal, Pierre. La civilisation mégalithique de Bouar: prospections et fouilles 1962–1966. Études oubanguiennes 1. Laboratoire d'Ethnologie et de Sociologie comparative, Université de Paris X-Nanterre. Paris: Firmin-Didot, 1969. The first sustained archaeological documentation of the complex, based on four seasons of fieldwork. The chronology proposed here (late Neolithic) was revised downward by later work but the typological inventory and the seventy-cluster map remain the baseline reference.

  • David, Nicholas. "Tazunu: Megalithic Monuments of Central Africa." Azania: Archaeological Research in Africa 17(1), 1982, pp. 43–77. The radiocarbon reassessment that moved the principal building horizon into the first millennium BCE. Excavates two tumuli to bedrock and establishes the no-human-remains pattern that has held through later work.

  • Zangato, Etienne. Étude du mégalithisme dans le nord-ouest de la République Centrafricaine. Doctoral thesis, Université de Paris X-Nanterre, 1991. The most thorough single archaeological treatment of the complex, with calibrated radiocarbon series and the three-period chronological framework subsequently published in monograph form.

  • Zangato, Etienne. Sociétés préhistoriques et mégalithes dans le Nord-Ouest de la République Centrafricaine. Cambridge Monographs in African Archaeology / BAR International Series. Oxford: Archaeopress, 1999. The published synthesis of the 1991 thesis. Documents iron metallurgy from the 9th century BCE and a megalithic florescence in the first millennium BCE. Standard reference for current chronology.

  • Bostoen, Koen, Bernard Clist, Charles Doumenge, Rebecca Grollemund, Jean-Marie Hombert, Joseph Koni Muluwa, and Jean Maley. "Middle to Late Holocene Paleoclimatic Change and the Early Bantu Expansion in the Rain Forests of Western Central Africa." Current Anthropology 56(3), 2015, pp. 354–384. The palaeoclimatic context for the rainforest perturbation phase that opened savanna mosaics across the Bouar region in the third millennium BP, providing the ecological framing for the late-monumental horizon.

  • UNESCO World Heritage Centre. Les mégalithes de Bouar. Tentative List file 4003, submitted 11 April 2006 by the Republic of Central Africa. The official site description, including the eastward / hydrographic orientation observation and the c. 3500–2700 BCE dating drawn from Vidal's 1969 framework.

Frequently Asked Questions

Are the Bouar megaliths older than Stonehenge?

The answer depends on which dating is accepted. Pierre Vidal's 1969 monograph La civilisation mégalithique de Bouar proposed a late-Neolithic horizon of roughly the fourth and third millennia BCE on typological and stratigraphic grounds, and UNESCO's 2006 tentative-list dossier (file 4003) cites c. 3500–2700 BCE drawn from this older framework. On that reading, the earliest stones at Bouar would predate or be roughly coeval with the sarsen settings at Stonehenge (c. 2500 BCE).The reassessment by Nicholas David in Azania 17(1) (1982), based on radiocarbon determinations from excavations of two tumuli, moved the principal building horizon into the first millennium BCE. Etienne Zangato's 1991 Nanterre thesis and 1999 BAR monograph refined this into three phases, with the megalithic florescence between roughly 950 cal BCE and the second millennium CE. Iron metallurgy in the same sequence appears from the 9th century BCE. On the David and Zangato chronology, the bulk of Bouar's monuments are 1,500 to 2,500 years younger than Stonehenge's main phases.Current Central-African archaeology accepts the David and Zangato framework. The conservative scholarly statement is that the Bouar complex is one of the oldest and largest concentrations of standing stones in sub-Saharan Africa, with first-millennium-BCE peak construction and possible earlier antecedents that are not yet well dated. The flat assertion that Bouar is older than Stonehenge depends on the older Vidal dating and is not the current consensus.

What does the word tazunu mean and who uses it?

The term tazunu (sometimes tajunu) is a Gbaya word, glossed in the Gbaya–French lexicographic tradition as "standing stones" or "stones-that-stand." The Gbaya are speakers of an Adamawa-Ubangian language (Niger-Congo, non-Bantu) and are the dominant population of the Bouar region today. They arrived in the area in the 16th century CE, several millennia after the monuments were built. The name is therefore a Gbaya descriptive term applied to pre-existing features in the landscape, not a name inherited from the builders themselves.This linguistic gap is consequential for interpretation. No oral tradition recoverable from the Gbaya, neighbouring Mboum, Sara, or Banda peoples claims descent from the builders or preserves a doctrine of the monuments' purpose. Etienne Zangato's 1999 BAR monograph treats tazunu as a convenient archaeological label rather than as a window onto builder intent. Pierre Vidal in 1969 used the same approach. The honest scholarly use of the term is descriptive: it labels the corpus, it does not interpret it.The transliteration varies. UNESCO's 2006 dossier uses tazunu; David's 1982 Azania paper uses tazunu with the diacritic conventions of the time; popular sources sometimes write tajunu or tanzunu. All refer to the same Gbaya morpheme. The variation is a transcription artifact of mapping a tonal Adamawa-Ubangian phoneme onto French and English orthographic conventions, not a reflection of distinct underlying terms. Within the scholarly literature on the complex, tazunu is the standard form, and that is the term Zangato uses throughout the 1999 BAR monograph that is now the principal reference work.

Has any astronomical alignment at Bouar been verified in peer review?

No. UNESCO's 2006 tentative-list dossier (file 4003) records that the megalithic groups are oriented eastward or in the direction of running water at their source, and Vidal's 1969 prospection report had observed the same regularity. Beyond this typological generalization across the seventy clusters, no per-site azimuth survey with calculated declinations, horizon-profile corrections, and statistical testing against a null hypothesis of random orientation has been published in Archaeoastronomy, Journal for the History of Astronomy, or any other peer-reviewed archaeoastronomical venue.The corpus-wide reviews of African archaeoastronomy — including the relevant entries in Clive Ruggles's Handbook of Archaeoastronomy and Ethnoastronomy (Springer, 2015) and the Wikipedia List of archaeoastronomical sites by country — catalogue Bouar as a site meriting investigation rather than as a site with demonstrated alignment. Nicholas David's 1982 Azania paper restricts itself to chronology and funerary inference; Zangato's 1991 thesis and 1999 monograph treat orientation as a typological trait but do not advance an astronomical interpretation.The methodological standard for confirming an alignment is well-developed (the work of J. McKim Malville and Fred Wendorf at Nabta Playa, of Ruggles at Stonehenge and Callanish, of Anthony Aveni at Mesoamerican sites). Applying it to Bouar would require a season of azimuth measurement at calibrated instruments across the seventy clusters, with corrections for atmospheric refraction and the precession of the equinoxes back to c. 1000 BCE. That work has not been done and remains the most consequential open project for the site.

Who built the Bouar megaliths?

The builders' cultural and linguistic identity cannot be recovered with current evidence. The principal building horizon at Bouar (c. 1000–500 BCE on the David and Zangato chronology) predates the main rainforest-route pulse of the Bantu expansion (Grollemund et al., PNAS 112(43), 2015) and falls before the arrival of the present-day Adamawa-Ubangian-speaking Gbaya, who entered the region in the 16th century CE. The builders were therefore neither Bantu nor Gbaya. The most plausible candidates are pre-Bantu agricultural populations of Adamawa-Ubangian or earlier substrate affiliation, but no linguistic reconstruction can connect them directly to a known language family.Material culture from the excavations is consistent with Late-Stone-Age to early-Iron-Age technology. Etienne Zangato's 1999 BAR monograph documents iron metallurgy in the regional sequence from the 9th century BCE onward, overlapping the megalithic florescence. The economy was agricultural (the rainforest-perturbation phase documented by Bostoen et al. 2015 opened savanna mosaics that supported cereal and tuber cultivation), and the labour mobilization required for individual monument construction is compatible with a village-cluster scale rather than a state-level organization.The absence of human remains in the excavated tumuli (David 1982; Zangato 1991, 1999) further constrains demographic inference: bone is the principal direct evidence for the builders' biological affiliation, and at Bouar that evidence is missing. Genetic work on present-day populations cannot stand in for it because of the intervening Gbaya arrival and the broader Bantu-expansion demographic restructuring documented by the Nature 2023 ancient-DNA synthesis on the Bantu expansion.

Why are there no burials inside the tumuli?

Both Pierre Vidal's 1962–1966 prospections (published 1969) and Nicholas David's two-tumulus excavation programme (published in Azania 17(1), 1982) failed to recover human skeletal material inside the monuments excavated. Etienne Zangato's later sondages and full excavations across multiple Bouar sites repeated that result. The pattern is not an artifact of preservation: skeletal material survives elsewhere in the region's archaeological record in comparable lateritic soils.Several explanations are entertained in the literature. The first is that the monuments are cenotaphs — memorial constructions commemorating individuals or families whose actual remains were interred or processed elsewhere. This is the interpretation Zangato's 1999 BAR monograph favours, on the grounds that the monument density correlates with what would be expected for individual or family-status markers but the contents do not match a primary-burial function. The second explanation, less developed in the literature, is that excarnation or other mortuary practices removed the remains before or shortly after the monument construction, leaving an empty container.The third possibility, that the monuments were never funerary at all — that they functioned as ritual, calendrical, or territorial markers without a mortuary component — is not directly contradicted by the evidence. Without ethnographic continuity to the builders or a corpus of human remains to study, the choice between these interpretations is underdetermined. The cautious phrasing in current Central-African archaeology is "memorial or commemorative," leaving the question of what is being commemorated open.

How does Bouar compare to the Senegambian stone circles?

The Senegambian stone-circle complex (Wassu, Sine Ngayene, Kerbatch, Wanar, inscribed on the UNESCO World Heritage List in 2006) is the most natural comparative case for Bouar within sub-Saharan Africa. Both corpora are large — Senegambia contains roughly 1,000 individual monuments across approximately 33,000 km² of the lower Gambia and Salloum river basins, while Bouar contains a smaller area (c. 7,500 km² per UNESCO's 2006 dossier) but several thousand individual stones across roughly seventy clusters. Both lack human remains in many of the excavated monuments and use locally quarried stone (laterite pillars at Wassu, granite blocks at Bouar).The chronologies differ. Senegambian dating, summarised in works such as John Sutton's 1982 work and the more recent radiocarbon series from Wanar (Holl et al.), places the principal building horizon between c. 300 BCE and CE 1500. On the David and Zangato chronology, Bouar's principal horizon (c. 1000–500 BCE) is significantly earlier, making it the older corpus despite the Senegambian circles' greater popular profile and UNESCO World Heritage status.The architectural form differs too. Senegambian monuments are typically circles of dressed laterite columns 1–3 m tall, with a single capstone or lyre-shaped pair often present; Bouar tumuli are mounds of laterite rubble with granite uprights set around the perimeter or crowning the apex. The shared traits — absence of bodies, locally sourced stone, association with watercourses — are striking, but the form-language is distinct, and no direct cultural connection between the two builder traditions has been established.

What did the rainforest perturbation phase have to do with the monuments?

The palaeoclimatic synthesis assembled by Koen Bostoen, Bernard Clist, Charles Doumenge, Rebecca Grollemund, Jean-Marie Hombert, Joseph Koni Muluwa and Jean Maley in Current Anthropology 56(3), 354–384 (2015), drawing on pollen records from Lebamba (Gabon), Lake Barombi Mbo (Cameroon), and Mopo Bai (Congo), documents a substantial perturbation of the central-African rainforest beginning around 4,000 BP on the forest periphery and reaching the core around 2,500 BP. The perturbation produced vegetation mosaics with patches of open forest interspersed with wooded or grassland savanna.This window corresponds almost exactly to the principal building horizon at Bouar on the David 1982 and Zangato 1999 chronology. The ecological consequences are direct: savanna and savanna-mosaic landscapes support the cereal and tuber cultivation systems documented archaeologically in the Bouar sequence, and the population densities and labour-mobilization capacities those systems can sustain are larger than what closed rainforest can support. The perturbation phase opened corridors across the central-African belt that allowed both the Bantu expansion (further south and east) and, in the Adamawa highland setting, the village-cluster economies that built the Bouar monuments.The monuments are therefore not simply a tradition that happened to exist at Bouar; they are likely a tradition that became possible at Bouar precisely when the rainforest opened. This is a strong correlation rather than a demonstrated causal link, but it is one of the more robust environmental-archaeological articulations available for any sub-Saharan megalithic corpus, and it tightens the chronological framework considerably.

What would a rigorous archaeoastronomical study of Bouar require?

The methodology is well-developed and has been applied to comparable sites in Africa and beyond. A defensible archaeoastronomical study of Bouar would require, at minimum, the following components. First, a systematic per-site azimuth survey across a representative subset of the seventy clusters, with calibrated theodolite or differential-GNSS measurements of the principal axis of each monument and of any internal alignments visible after the laterite cover is mapped. Second, horizon-profile measurement at each surveyed site, accounting for the local topography and any obstructions that affect the observed rising or setting azimuth of a celestial body as distinct from the purely geometric value.Third, correction for atmospheric refraction (typically 0.5° at the horizon at low latitudes) and for the precession of the equinoxes back to the relevant epoch — c. 1000 BCE on the David and Zangato chronology, c. 3000 BCE if Vidal's older dating is given weight. Stellar declinations have shifted measurably over those intervals and any star-alignment hypothesis has to use the back-calculated position. Fourth, statistical testing of the observed azimuth distribution against a null hypothesis of random orientation, using methods comparable to those Clive Ruggles applied at Stonehenge and at the Carnac alignments in Brittany.Fifth, comparison with the broader regional record, particularly the standing-stone sites in north-western Cameroon documented by Scott MacEachern (Antiquity 68, 1994), which are the closest neighbours and may represent the same builder tradition. A study meeting these criteria has not been published for Bouar. Until one is, statements that the complex "is" an astronomical observatory remain hypotheses, and the eastward orientation of the corpus is most accurately treated as a typological observation — striking, suggestive, and untested.