In October 2023 a team of Indonesian scientists published a paper saying that a terraced hill in West Java hides the oldest pyramid on Earth, its deepest layers shaped by human hands more than 20,000 years ago. Five months later the journal took it back. To the authors and their supporters, led by geologist Danny Hilman Natawidjaja and championed by Graham Hancock, what happened in between was suppression: a finding that would rewrite human history was pulled not because it was wrong but because it was unwelcome. What was retracted was the paper's dating. What was denied, in the authors' telling, was a fair hearing. How it happened is on the record: third-party complaints, a publisher's investigation, and a retraction notice the authors refused to accept. Why it happened is the argument.
The authors' own statement, "The Unjust Retraction of Groundbreaking Research: A Call for Academic Integrity," was published on Hancock's website on 21 March 2024 (https://grahamhancock.com/natawidjajadh1/). It says the paper "provides strong evidence that Gunung Padang is home to the most ancient pyramidal edifice yet discovered anywhere in the world, with parts of its man-made interior structure being more than 20,000 years old," a finding that "shatters established historical timelines." It says the team's years of work "have been undermined and suppressed by conventional voices who find this discovery unacceptable." The authors appended the original article, the critiques sent to the publisher, their correspondence with Wiley, extra data and a reply to the notice. A central technical point is what they call Unit 2, a body of columnar rock under the visible terraces that, they argue, earlier Indonesian archaeologists failed to recognise as built. Calling those rocks natural columnar joints, they write, does not rule out that they were carved before being covered. Twelve authors signed it.
Scientific American's report on the paper (https://www.scientificamerican.com/article/experts-doubt-claims-that-worlds-oldest-pyramid-was-discovered-in-indonesia/) sets out what the team claimed and how. Gunung Padang has five stepped stone terraces with retaining walls and stairways on top of an extinct volcano. Between 2011 and 2014 Natawidjaja and colleagues surveyed beneath it with several ground-penetrating methods and identified four layers, which they read as separate phases of building. The innermost is a hardened lava core that, according to the paper, had been "meticulously sculpted." Over it were layers of rock "arranged like bricks." They dated the layers using soil lodged between rocks in drill cores. The first stage, they said, came between 27,000 and 16,000 years ago, further work between 8,000 and 7,500 years ago, and the visible terraces between 4,000 and 3,100 years ago. "The pyramid has become a symbol of advanced civilization," Natawidjaja told the magazine. "It's not easy to build pyramids. You need high masonry skills."
Nature's news team broke the next step on 28 November 2023, in "A 27,000-year-old pyramid? Controversy hits an extraordinary archaeological claim" (https://www.nature.com/articles/d41586-023-03546-w). The structures at Gunung Padang, the subhead said, would be much older than Egypt's pyramids "if they're even human constructions at all," and the paper had raised the eyebrows of some archaeologists and "prompted an investigation by the journal that published it." That was the first public sign that the publisher was looking again.
The Guardian followed on 16 December 2023 with "'Really, really weak': experts attack claim that Indonesia site is 'world's oldest building'" (https://www.theguardian.com/science/2023/dec/16/really-really-weak-experts-attack-claim-that-indonesia-site-is-worlds-oldest-building), the headline quoting critics. Its account adds the detail that changed the story's temperature: the paper had been proofread by Graham Hancock, whose Netflix series Ancient Apocalypse had suggested the site could be the work of a lost culture destroyed in a cosmic event at the end of the Ice Age. Archaeologists quoted said the pottery on the surface pointed to a site a few thousand years old. Natawidjaja himself confirmed the investigation, saying it "addresses concerns raised by third parties regarding the scientific content of our paper."
Retraction Watch reported the outcome on 20 March 2024 in "Controversial pyramid paper retracted when authors turn out to have radiocarbon-dated nearby dirt" (https://retractionwatch.com/2024/03/20/controversial-pyramid-paper-retracted-when-authors-turn-out-to-have-radiocarbon-dated-nearby-dirt/). It quotes the notice in full. Concerns had been "raised by third parties with expertise in geophysics, archaeology, and radiocarbon dating." The publisher and the editors concluded there was "a major error," not identified in peer review: "the radiocarbon dating was applied to soil samples that were not associated with any artifacts or features that could be reliably interpreted as anthropogenic or 'man-made.'" Therefore, the notice says, "the interpretation that the site is an ancient pyramid built 9000 or more years ago is incorrect." All the authors disagreed. Retraction Watch also notes that Natawidjaja had tried to link the site to Atlantis.
Times Higher Education's report the same day, "Paper claiming 'world's oldest pyramid' retracted over 'error'" (https://www.timeshighereducation.com/news/paper-claiming-worlds-oldest-pyramid-retracted-over-error), carries both sides. Natawidjaja said the retraction was "solely based on unfounded claims raised by third parties who hold differing opinions and disbelieve in the evidence, analysis, and conclusions." Critics said the structure should be dated from artefacts found at the site, which suggest it is about 2,000 years old. The report adds that Hancock had visited for his series and argued that a prehistoric civilisation might have spread its science and technology before the Ice Age ended.
Gunung Padang's terraces are real and old. The retracted 2023 paper dated soil from drill cores to claim a buried pyramid more than 20,000 years old; the journal pulled it after third-party complaints, saying the soil was not tied to anything man-made. The authors and Graham Hancock call that suppression of an unwelcome discovery, and the dispute remains unresolved.
What Stays Contested
Whether the deep units under the terraces were shaped by people has not been settled by excavation; the argument still rests on remote sensing and cores. The authors say the retraction judged a method rather than the data. Nobody has yet dug down through Unit 2 and published what lies there.