In 1950, the French ethnographers Marcel Griaule and Germaine Dieterlen published an account of Dogon astronomy that described a star invisible to the unaided eye. According to the people they interviewed in what is now Mali, the star was remarkably small and heavy, and it moved around the much brighter Sirius in a cycle of approximately fifty years. Astronomers already knew of a companion with similar properties. Sirius B is a dense white dwarf whose faint light is difficult to separate from that of Sirius A. The resemblance between established astronomy and the account collected by the ethnographers drew attention to their report.
The story became the Dogon Sirius mystery. The puzzle concerned where the reported knowledge came from: how had such a description entered the researchers’ record? It might have been part of an older local tradition, arrived through more recent exchanges with outsiders, or taken shape through misunderstandings and decisions made during the interviews and their publication. The match with astronomy could not tell readers which explanation was right.
The surviving publications allow these possibilities to be examined in detail. Griaule and Dieterlen identified four principal Dogon informants in their 1950 paper and explained how they brought the interviews together. Anthropologists working in the region decades later reported quite different beliefs and questioned the original methods. Their dispute concerns how oral accounts are recorded and transmitted, how much an interviewer may shape an answer, and whether people from different Dogon communities can reasonably be described as holding a single uniform doctrine.
Dogon Country Was Never a Place Outside History
Dogon communities live in central Mali, across the Bandiagara plateau, its sandstone escarpment and the plains below. Their settlements occupy quite different environments, from villages on the heights to those built along or beneath the cliffs. People have farmed, traded and practised crafts there, while family organization and ritual responsibilities have differed among settlements. The cliffs have attracted many visitors, sometimes overshadowing the social and economic histories of their inhabitants.
The area preserves granaries, altars, shelters used for meetings, houses and burial places, along with traces of earlier occupation. Archaeologists have documented remains attributed to populations conventionally known as Tellem, whose history precedes much of the settlement pattern associated with Dogon communities today. Bandiagara entered the UNESCO World Heritage List in 1989. UNESCO’s designation concerns a landscape shaped by long histories of settlement and use. It does not present the region as a place where a single ancient way of life survived without alteration.
Bandiagara has never been cut off from the rest of West Africa. Trading connections linked its inhabitants to neighbouring communities, and people moved between villages, markets and larger towns as political circumstances changed. Islam and Christianity also formed part of the region’s history by the time French researchers were working there. The arrival of tourists and the spread of books about Dogon religion introduced further influences, including on how symbols and ceremonial practices were explained to visitors. Such contacts matter when assessing when an astronomical idea could have entered local accounts.
That history makes statements such as ‘The Dogon believed.’ especially imprecise. What one elder explained to an interviewer could reflect a teaching specific to a village, knowledge held by ritual specialists, an individual interpretation or an answer shaped for the person asking. Even a carefully recorded statement does not show that every Dogon settlement knew the same account. It establishes still less about whether the statement had circulated for centuries in precisely the form later published.
Masks and public masked performances, often connected with funerary ceremonies, are widely associated with Dogon culture. Yet their designs, uses and interpretations differ by place and have responded to changing circumstances, including audiences from outside the region. Anthropologists Anne Doquet and Éric Jolly have examined how tourism and scholarly publication helped establish a standard public image of ‘Dogon tradition’. Their work cautions against using a practice recorded in one village to describe all Dogon communities.
What Sirius B Is, and When Astronomers Found It
Sirius is the brightest star visible in the night sky from Earth. Its brilliance reflects both its luminosity and its distance of about 8.6 light-years. What appears to be a single star is a binary system. The bright component, Sirius A, and the much fainter Sirius B orbit a common centre of mass in approximately fifty years. As they travel around it, the apparent distance between the two points of light changes for observers on Earth.
Sirius B is a white dwarf: the compact remnant of a star that has exhausted the nuclear fuel sustaining its earlier development. Roughly the Sun’s mass is compressed into an object about the size of Earth. This makes it extremely dense, but it is also faint compared with its brilliant neighbour. Even when the two stars appear separated, the light from Sirius A overwhelms Sirius B for an unaided observer. Telescopes and modern instruments allow astronomers to distinguish them and investigate their physical properties.
The companion’s scientific discovery came well before the ethnographic publication. In 1844, Friedrich Bessel used irregularities in the apparent motion of Sirius to infer that an unseen body was pulling on it. Alvan Graham Clark saw the faint companion with a telescope in 1862. During the twentieth century, research into stellar physics explained how a star could be so dense. A NASA account of Hubble observations describes the use of gravitational redshift, the effect of gravity on escaping light, to measure the white dwarf’s mass independently. By 1950, astronomers already knew about the companion’s orbit and unusual physical properties.
The details attributed to the Dogon informants in 1950 still merit examination. A star described as small and exceptionally heavy, revolving around Sirius in about fifty years, invites comparison with Sirius B. But its similarity to Sirius B cannot tell us whether the account circulated widely among Dogon communities, whether it predated contact with European astronomy, or how accurately the French authors reported their interviews.
The Four Informants Named in the 1950 Sirius Study
The primary source is Griaule and Dieterlen’s 1950 article, Un système soudanais de Sirius, printed in volume 20 of the Journal de la Société des Africanistes, pages 273–294. The authors describe research conducted among the Dogon from 1946 to 1950 and refer to related observations among Bambara, Bozo and Minianka people. They placed their Dogon interviews within a comparison involving neighbouring communities. Their mention of neighbouring groups should not be converted into independent proof that each of those peoples possessed the same account of Sirius.
The paper names four main informants rather than an anonymous body of elders. Innekouzou Dolo, an elderly woman, is described as a priestess and diviner at Ogol-du-Bas. Ongnonlou Dolo was an elder linked with Go. Yébéné served as a priest at a binu sanctuary in Ogol-du-Haut, while Manda was a priest living at Orosongo. Griaule and Dieterlen also gave approximate ages and recorded local affiliations and languages. The four names rule out the common claim that the entire account came from one elder interviewed for Griaule’s earlier book.
According to Griaule and Dieterlen, Ongnonlou gave them the most complete version of the system, with the other three contributing substantial elements. The authors then arranged their statements as one continuous account. The result was a coherent narrative, but readers cannot reliably trace every detail to an individual informant or identify differences between their versions. The publication is therefore best treated as an ethnographic synthesis made from several interviews. It is not a word-for-word record of a doctrine spoken in the same form by all four people.
The authors did not pretend to know everything about what they had recorded. On page 274, they described the material as belonging to an initial level of initiation and acknowledged that they could not explain how people without astronomical instruments might know of nearly invisible stars. They chose to publish the account while leaving its origin unresolved. Their uncertainty about the origin of the claim is not evidence for a prehistoric source.
The geographical label in the article’s title has also caused confusion. Its title, ‘A Sudanese System of Sirius’, uses ‘Sudanese’ to refer to the historical Western Sudan, the broad region of West Africa that includes what is now Mali. It does not refer to the modern Republic of Sudan. Nor was the reported material presented simply as the timeless secret of a single isolated ‘tribe’. The article concerned identifiable people interviewed in a region with several communities and a history of contact beyond its borders.
Po Tolo, Sigi Tolo and the Fifty-Year Orbit
Griaule and Dieterlen associated sigi tolo with Sirius and the important Sigui cycle of ceremonies. They also described a smaller star, po tolo, whose name was connected with Digitaria exilis, a tiny grain commonly called fonio. This smaller celestial object appeared in their account as exceptionally heavy despite its minute size. A local name or metaphor cannot be assumed to describe a modern astronomical object. The published account mixes star references with descriptions of plants, religious origins, graphic signs and movements in the heavens.
In one part of the paper, the small “Digitaria” star is also a seed or egg from which the world unfolds. That is an image within an account of creation, with a different function from the scientific statement that a white dwarf is made of densely compressed matter. The ethnographers included diagrams of spirals and compartments and described signs related to elements of the cosmos. Those drawings record symbolic explanations given in interviews. They are not measurements or descriptions of the physical structure of Sirius B.
The detail that attracts the most attention in the 1950 account is the small star’s reported revolution of roughly fifty years, close to the period astronomers calculate for Sirius B. The same article describes a Sigui ceremonial cycle of about sixty years. The reported periods concerned different things. Griaule and Dieterlen discussed the ways people were said to count the longer ceremonial interval, including marks and sequences linked to gatherings for millet beer. The ceremony cannot simply be described as a fifty-year astronomical clock without losing what the researchers actually reported.
The 1950 paper contains descriptions of other stars and their supposed relationships as well. Some popular accounts have identified them with an imagined Sirius C, although a third star matching those descriptions has not been established in the Sirius system. A comparison with astronomy has to consider the matches and mismatches, including names whose identification remains uncertain. Selecting the closest resemblances and ignoring the rest gives a misleading impression of the evidence.
For these reasons, the statement that “the Dogon knew Sirius B” goes beyond what the records establish. What can be documented is a 1950 publication in which two ethnographers brought together statements they associated with the Sirius system. Whether the ideas were old, widespread, independently acquired or consistently interpreted by those who supplied them is a further set of questions. The publication itself does not answer them.
Ogotemmêli and the Earlier Book Were a Different Source
Ogotemmêli is frequently named in retellings of the Sirius story, although he was not one of the four informants in the 1950 study and was involved in a different part of Griaule’s research. The ethnographer interviewed him in a series of meetings in 1946 and published the resulting conversations as Dieu d’eau: entretiens avec Ogotemmêli in 1948. The book later appeared in English as Conversations with Ogotemmêli. It concerned creation, language, ancestors and symbolic relationships, helping introduce European readers to Griaule’s account of Dogon religious thought.
Griaule presented these meetings as an elder gradually instructing a visitor in explanations that were not immediately accessible to outsiders. Earlier research had focused on masks and ritual activity; he now believed he was hearing a deeper account that brought many practices into a single system. His interpretation influenced how subsequent readers approached his work. Readers who accepted the 1948 book as an account of shared Dogon cosmology could easily see the 1950 Sirius material as another part of the same hidden teaching, although it arose from different interviews.
The distinction matters because none of the four principal informants named in the 1950 Sirius paper was Ogotemmêli. There was also a further synthesis, Le Renard pâle, prepared with Dieterlen and published in 1965 after Griaule’s death. It presented another elaborate account of origins and signs. Anthropologists have since disagreed about how these publications relate to one another, how consistent their versions are and how the researchers’ techniques shaped the resulting descriptions. The 1948 interviews, the 1950 Sirius article and the 1965 volume form three different sources, not a single continuous testimony.
Why Walter van Beek Reopened the Case
Walter E. A. van Beek began fieldwork in Dogon communities in the late 1970s, including at Sangha and Tireli. His initial concerns were village life, ecology and religion rather than an investigation devoted to Sirius. As his interviews increasingly diverged from the accounts in Griaule’s books, he began asking informants directly about the published explanations and the practices they were said to describe.
Van Beek found many practices he could recognize, from masks and village altars to accounts of ancestors and rituals. What he repeatedly failed to find was the tightly connected creation system presented in the Griaule publications as a common explanation underlying them. His critique appeared in 1991 as Dogon Restudied: A Field Evaluation of the Work of Marcel Griaule in Current Anthropology, alongside responses by other specialists. Van Beek asked whether an ethnographer had assembled individual accounts into a system that Dogon communities did not generally share.
In Haunting Griaule (2004), van Beek reported hearing Sirius referred to as dana tolo, the hunter’s star. He did not obtain accounts of the invisible companion or the claimed link between Sirius and Sigui. He could identify traditions concerning migration, funerals and masks in his fieldwork, while the comprehensive cosmologies attributed to the community proved harder to corroborate. His challenge was to particular interpretations made in the postwar studies, not to the reality or value of Dogon religious life.
Van Beek’s 2004 article offers an example of how publication could change the answers researchers later heard. A person offered van Beek a symbolic explanation of a mask, but on further questioning said the interpretation came from a tourist guide who had repeated ideas drawn from the Griaule literature. Material first circulated in print could thus return to local conversations as supposed ‘traditional’ knowledge. The person repeating it need not have intended to deceive anyone. For the historian, the problem is that repetition does not provide an independent source for the antiquity of the claim.
Van Beek’s later interviews could not settle what every Dogon specialist had known in the 1940s. He interviewed people decades after Griaule and Dieterlen, in communities with different experiences. An earlier teaching could have been restricted, changed, forgotten, or confined to a group he did not encounter. Still, his findings raised a substantial problem: he found a widespread mismatch between the published cosmological scheme and later accounts, while there was little independent early evidence for the precise Sirius B details. His criticism also made clear how difficult it is to reconstruct shared doctrine from separate interviews.
The Debate Was About Ethnographic Method, Not African Intelligence
Other scholars challenged parts of that criticism. The 1991 discussion in Current Anthropology included responses by specialists, and Geneviève Calame-Griaule, Griaule’s daughter and a researcher of Dogon language and culture, published another reply. Defenders asked whether an outsider working decades later could expect access to restricted knowledge, and whether van Beek had paid enough attention to differences among communities and generations. Those objections qualify the reach of his findings. They do not, without further independent records, establish that the specific astronomical interpretation published in 1950 was historically ancient or widespread.
Secrecy complicates both sides of the debate over the 1950 account. If only a few ritual specialists knew an account, interviews with other villagers might reasonably fail to uncover it. But that makes it particularly important to document who gave each statement, in which language, in what circumstances and with what chance to clarify its meaning. Griaule and Dieterlen supplied names but combined the accounts of their four principal informants in the finished text. The blending makes it harder for later researchers to determine whether different people independently supported the same detail.
The research also took place within French colonial rule. Griaule sought to challenge European prejudices by showing the complexity of African intellectual and religious life. Yet his position, the expectations of his research team, the work of interpreters and the decisions of Dogon collaborators all influenced how material was collected and presented. What reached print depended on those conversations and the researchers’ choices. Recognizing that setting neither reduces Dogon informants to passive subjects nor accuses them of dishonesty when their recorded accounts differ.
A historical examination of the Griaule school by Doquet and Jolly traces how the image of a stable, harmonious and unchanging Dogon society developed through colonial scholarship and was later reused in Malian public life, tourism and local presentations of culture. They examine how people explained practices and how those explanations changed. Their findings caution against assigning one symbolic interpretation to all Dogon masks, village buildings or granaries. A claim should be linked to the people, time and circumstances in which it was recorded.
Could the Astronomy Have Arrived Through Ordinary Contact?
European astronomers observed Sirius B in 1862, and research into its unusual properties circulated long before Griaule and Dieterlen wrote about the Dogon. Information of that kind could have reached the Bandiagara region through officials, missionaries, teachers, merchants, travellers or people who moved between communities. Information could travel without the instruments used to make the original observations. The history of travel and exchange makes such transmission possible, even if the particular route remains unknown.
Possibility must still be distinguished from demonstration. No documented sequence names the person who supposedly carried information about Sirius B to one of the four principal informants, or when such a conversation took place. Nor is there a securely dated precolonial record showing that the same detailed claims had circulated for centuries. The surviving documents establish that the ethnographers published the account in 1950. They cannot tell us how long every element had been known by the people interviewed.
Van Beek proposed a more specific explanation. In his 2004 article he suggested that Griaule’s interests in aviation and astronomy might have influenced the questions he asked, with ideas introduced by the researcher later appearing in interviews as local knowledge. Van Beek offered this as a possible account of the interviews, not as a documented exchange. His observations about leading questions make the suggestion plausible, but the exchanges of the 1940s were not recorded in enough detail to establish exactly what each participant said at every stage.
Misidentification in the process of interpretation offers another possibility. Researchers familiar with European astronomy could have understood local names for stars, ceremonial counting practices and ritual drawings through a framework their informants did not necessarily share. Several processes may have occurred together: ideas could have arrived through contact, been given local meanings and then been arranged into a coherent account for publication. The surviving material is compatible with more than one process. None has been shown to explain the full 1950 description conclusively.
Robert Temple and the Extraterrestrial Version
Robert Temple brought the 1950 Sirius story to a far wider audience with The Sirius Mystery, published in 1976. He combined the 1950 report with material from ancient Mediterranean and Near Eastern traditions, arguing that visitors from the Sirius system had taught people about the stars. Figures in Dogon religious accounts, especially Nommo, became part of his proposed evidence for extraterrestrial visitors. That explanation depended on two assumptions the primary material had not secured: that the astronomical identifications were accurate and that the reported knowledge went back to a remote period.
The 1950 paper was a record and interpretation of what four informants told ethnographers. It did not document the arrival of beings from another star system. Neither that paper nor van Beek’s later investigations produced dated archaeological evidence, independent ancient accounts or astronomical observations that support a spacecraft visiting Mali. The primary reports provide no evidence for the steps required by Temple’s extraterrestrial explanation.
Popular versions of the story usually leave out the details that are most important for judging it. Ogotemmêli was not one of the four Sirius informants named in 1950; the published account blended testimony; the fifty-year stellar orbit and sixty-year ceremony were separate claims; ordinary contact could have transmitted information; and later fieldwork did not recover the same doctrine. These qualifications are part of the record. Ignoring them produces either a sensational claim about impossible knowledge or an equally careless dismissal of the culture discussed in the report.
Dogon Religion and Culture Beyond the Sirius Claim
The astronomical controversy has often overshadowed the region’s documented social history. The Bandiagara region is inhabited by communities whose livelihoods, architecture, art and religious practices can be studied directly, regardless of whether a reported claim about Sirius B is valid. UNESCO describes a landscape containing granaries, homes, sanctuaries and meeting shelters. Those buildings store resources, support daily life and accommodate social activity, alongside whatever ceremonial associations particular communities may give them.
A granary is first a practical building, intended to protect grain and the results of household labour across seasons. Some ethnographers have also interpreted parts of its design through ancestry and creation stories. Those readings cannot be assigned to every Dogon granary without evidence from the people who use it. Doquet and Jolly have shown that some explanations given by guides about doors, shutters and buildings were influenced by expectations created through tourism. A reliable account must identify whose interpretation it is, and where and when it was recorded.
Fox divination is another practice documented in particular Dogon communities. Diviners may mark a sandy surface and later read the tracks left by an animal moving across it. Van Beek studied the practice, and the Smithsonian bibliography records his 2005 study of Dogon fox divination. Doquet and Jolly report that its distribution was limited to certain villages near the cliffs and that neighbouring peoples influenced it. Its occurrence in some villages cannot establish a uniform Dogon religious system, much less knowledge of Sirius B.
The Sigui ceremonial cycle and masked performances also deserve attention as practices with histories of their own. Their timing and meanings must be understood through the communities and periods in which they were observed. Accounts of these ceremonies record their social settings, meanings and changes over time. The debate over astronomy can easily overshadow this more direct record of ceremonies, stories and the choices made by the people who continue to practise them.
What the Dogon Sirius Evidence Establishes
The original evidence allows several findings to be stated securely. Griaule and Dieterlen published a Sirius-related system in 1950, named four principal Dogon informants and combined material from their interviews. Elements of that report resemble the properties of Sirius B, which astronomers had discovered well before the ethnographers published. Later fieldwork, particularly van Beek’s, did not reproduce the comprehensive Sirius account and raised serious questions about the method by which it had been assembled. These findings must be held together rather than allowing any one to stand in for the others.
No one disputes the existence of Sirius B. The historical uncertainty concerns the reported knowledge: when the relevant ideas arose, how far they circulated and how closely the French publication reflected its informants’ own explanations. Suggestive resemblances cannot establish that the complete system was known before colonial contact. On the other hand, the disagreement between researchers is not evidence that every person interviewed in the 1940s fabricated what they said. The limitations of the records are more specific than either accusation.
The sources can be read together without treating them as interchangeable. The 1950 paper documents a particular ethnographic claim; astronomy establishes the properties of the star; and later anthropological research tests how securely that claim can be treated as a shared historical tradition. The enduring mystery in popular retellings often comes from treating those different kinds of evidence as though they proved the same thing. They do not.
Discussion
Comments