The museum inside the concession
Start with the building, because the building tells you what kind of institution this was. The Museu do Dundo was founded in 1936 in the north-east of Angola, inside the concession of the Companhia de Diamantes de Angola—Diamang—the company that held the exclusive right to prospect for diamonds across the province. This is not a national museum that happened to be in a mining region. It was a corporate museum, an organ of the company’s administrative structure, and it sat at the heart of a concession where the company also held the exclusive right to run commerce and privileged conditions for recruiting local labour. The museum and the mine were the same institution wearing different clothes.
The founding story is worth reading closely, because it is a story the company told about itself. According to the scholarly account published by BEROSE (the encyclopaedia of the history of anthropology), the museum’s origin is narrated as near-serendipity: José Redinha, then an administrative trainee at the Chitato border post, was invited to exhibit his drawings and ethnographic collection at the Diamang staff house in Dundo before returning to Portugal. Instead, the company bought his 496 pieces and installed them in a house. The general director in Lunda, engineer Quirino da Fonseca, framed the purchase as a mission to save aspects of local cultures ‘in rapid extinction’—which the company considered its ação civilizadora, its civilizing action, in which the place of indigenous cultures was in the past.
That sentence is the whole machine in miniature. Extraction in the present; culture assigned to the past; the company as rescuer. The museum did not interrupt the concession. It was one of the concession’s departments.
What the collection actually was
The numbers are not vague. In its first twenty years the collection grew from 496 pieces to 9,535, per the BEROSE account. It was built through a series of in-situ collecting campaigns led by Redinha: Lunda central, north and north-west in 1937; the Sombo region and then the Alto Zambeze in 1938; north-east Lunda and Canzar in 1939; graphic documentation in the Sombo region in 1940; the Alto Tchiboco in 1946. From 1952, the campaigns were executed by ‘natives trained by museum staff,’ and the acquisition of pieces brought in by their local owners became routine.
To the ‘Regional’ collection were added the Dr. Hermann Baumann collection—1,150 pieces collected in 1954 by the German authority on Lunda—and an ‘Africana’ collection of 726 pieces bought from antiquarians and auction houses in Diamang’s European and American offices, sent to Dundo as comparative material. So the museum was simultaneously a site of local collection, a node in a European art market, and a display of the company’s reach. The building was completed in 1949; a folklore music room was added in 1956, based on recordings begun in the 1950s and continued until 1968—through the war for Angola’s independence.
The local Lunda-Cokwe dignitaries who dealt with the museum had their own name for it: Mutenje, a place of sacred or spiritually powerful things. They called Redinha Sa Capuma, ‘great sorcerer.’ That is not ethnographic colour. It is evidence that the institution meant something different to the people whose objects it held than it meant to the company that owned it.
The sound archive: a test case you can actually check
If you want to test the company’s self-presentation against the material record, the sound recordings are the cleanest case, because every step is documented. The CREM-CNRS archive (Research Centre of Ethnomusicology, University of Paris 10) catalogues a collection titled Folclore musical de Angola volume I: recordings made in 1954 in Lunda Norte with Chokwe (Tshokwe) performers, recorded by the Museu do Dundo, published by Diamang in Lisbon in 1961 (with a second volume in 1967), and deposited at the Musée de l’Homme in Paris as a donation from the company’s administrative direction, with a covering note dated 12 February 1964.
The deposit note carries a condition, quoted in the archive record: given the purely cultural nature of the work, the company stipulated that the recorded songs not be lent for commercial purposes. Read that twice. A diamond company that built its fortune on concession labour attached a non-commercial condition to a gift of recorded music—while the recordings themselves had been made by a museum that was a department of the same company. The condition is not hypocrisy in the abstract; it is a documented transfer, with a date, a donor, a recipient, and a clause. That is what a provenance trail looks like when you can actually follow it.
The archive record also notes that the volumes were published in Portuguese and English, and that the Library of Congress holds a related set (AFC 1964/021). So the same corporate-produced documentation of Chokwe music sits in at least two European and North American repositories, catalogued, accessible, and framed by the company’s own publication. The ‘civilizing mission’ did not stay in Angola. It travelled, in boxes, with paperwork.
The dispersal
When the colonial period ended, the collection did not simply stay put. According to BEROSE, in 1975 Diamang transferred 1,324 objects from the ‘Regional’ collections and 681 from the ‘Africana’ collections to Luanda—2,005 pieces that became the foundation of the National Museum of Anthropology of Angola. The same account states that many of these pieces would end up diverted to the international market in ‘African tribal art.’ That is a consequential claim, and it is attributed to the scholarly source, not asserted here as an independent finding. But it matters for how we read the whole enterprise: the corporate museum’s objects did not have a single, stable home. They moved through colonial institutions, national institutions, and markets.
After independence, under the cultural decentralisation policy of Henrique Abranches as director of Angola’s museum services, the Dundo Museum was classified as national. The BEROSE account gives the regional collections at 12,131 pieces and the African collection at 975, with the Baumann collection at 1,150. The museum’s current director, Ilunga André, has described it as multidisciplinary, holding ethnographic, biological and archaeological collections—a framing that keeps the institution’s scientific identity alive while the provenance questions remain open.
Lisbon: what is documented, and what is not
Here the record requires discipline. The Museu Nacional de Etnologia in Lisbon holds major ethnographic collections from Africa, Asia and South America—more than 40,000 objects, according to the museum’s published description—and it has presented an exhibition, ‘Deconstructing Colonialism, Decolonising the Imaginary,’ running from 30 October 2024 to 2 November 2025, framed as part of the museum’s aim to investigate the provenance of its extra-European collections and reflect on the colonial context in which it was founded and its first collections gathered.
What the retrieved record does not establish is a direct transfer of Dundo/Diamang objects to Lisbon. The documented 1975 transfer went to Luanda. If you see a claim that Lisbon’s national ethnology museum ‘still holds the Dundo collection,’ treat it as an open question, not a fact—and ask for the inventory number. The provenance turn at the Lisbon museum is real and documented; the specific Dundo-to-Lisbon link is not, in the sources available here. That distinction is not pedantry. It is the difference between an argument and a slogan.
Why the framing matters
The strongest critique of the Dundo Museum is not that a diamond company built a museum. Companies build things. The critique is structural: the museum was an institute of the company’s administrative body, its founding narrative converted a commercial acquisition into a rescue mission, its collecting campaigns turned local owners into suppliers, and its publications and deposits carried the company’s cultural authority into European institutions with conditions attached. Extraction and patronage were not in tension. They were the same operation, run from the same offices, with the same letterhead.
The provenance reckoning now underway in Lisbon and elsewhere is worth watching precisely because it can be tested against documents like the 1964 deposit note. A museum that wants to decolonise its collections can start by publishing the chain of transfers—donor, date, condition, catalogue number—for every object and recording that came out of a colonial concession. The Dundo record shows what that chain looks like when it survives. The question for Lisbon is whether the chain for its own African holdings is as legible, and whether anyone has asked.
Questions readers actually ask
Did Diamang build the Dundo Museum to stage a civilizing mission?
The documented record shows the company framed the museum within its ação civilizadora, with local cultures placed in the past, and that the museum was an organ of the company’s administrative structure. Whether that was a conscious propaganda plan or an institutional habit is an interpretation; the framing itself is documented in the scholarly account.
Does Lisbon’s Museu Nacional de Etnologia hold the Dundo collection?
No retrieved source establishes a direct Dundo-to-Lisbon transfer. The documented 1975 transfer went to Luanda. The Lisbon museum does hold major African collections and has run a provenance-focused exhibition, but the specific link should be verified against its catalogue before being asserted.
Where can I check the sound recordings?
The CREM-CNRS archive record for collection CNRSMH_E_1964_011_001 lists the 1954 recordings, the 1961 Diamang publication, and the 1964 deposit note with its non-commercial condition. The Library of Congress holds a related set (AFC 1964/021).
What happened to the objects transferred in 1975?
Per the BEROSE account, 2,005 pieces went to Luanda and became the foundation of the National Museum of Anthropology; the same account states many were later diverted to the international market. That claim is attributed to the source and should be read as such.