Graham Hancock on the Joe Rogan Experience #2215, fact-checked

What the evidence shows: The Dokos shipwreck, off the Greek island of Dokos in the Argo-Saronic Gulf, is generally recognized as the oldest known shipwreck site discovered by archaeologists. Wood and pottery from the site date it to the Early Helladic II period, roughly 2200 to 2250 BCE, which makes it about 4,200 years old, close to Rogan's initial figure of about 4,000 years. His revised figure of about 6,000 years is incorrect: that would place the wreck near 4000 BCE, well before the Early Bronze Age date the cargo supports. Because everything biodegradable dissolved, no hull survives, so it is more precisely the oldest known nautical trade site rather than a preserved ship.

What the evidence shows: Hancock's figures track a 2024 peer-reviewed study in the Proceedings of the National Academy of Sciences (Bradshaw, Reepmeyer, Moutsiou and colleagues), whose demographic models predict the first human arrival on Cyprus between 14,257 and 13,182 calendar years ago, near the end of the last Ice Age. That study infers 2 to 3 large, organized migration events within under 100 years, each bringing roughly 1,000 to 1,375 people, and concludes the Mediterranean was not a barrier to early seafarers, closely matching his claim of planned migrations of a thousand or so people. Cyprus is a deep-water oceanic island that was never connected to the mainland, so it could only be reached by sea, as a separate 2024 palaeogenomic paper reiterates in describing Epipaleolithic seafaring visits over 12,500 years ago at Akrotiri-Aetokremnos followed by early Neolithic settlement. The main caveats are that the 14,000 to 12,500 window and the group sizes come from demographic modelling rather than direct census, the timing and permanence of the earliest occupation remain debated, and the figures are best read as roughly accurate estimates.

What the evidence shows: Both parts of the claim align with the peer-reviewed record. Optical dating of stone artefacts at the Madjedbebe rock shelter in northern Australia set a minimum age of about 65,000 years for human occupation, so a figure of 50,000 years ago is a conservative lower bound rather than an overstatement (some researchers argue for a date closer to 50,000, but none younger). On the crossing, modeling of the peopling of Sahul (Australia, New Guinea and adjacent islands joined at low sea level) finds that reaching the continent required at least one open-ocean crossing of roughly 100 km plus several shorter crossings of 20 to 30 km, and that this water gap persisted even at the maximum glacial lowstand of about 125 m below present sea level. The requirement for genuine open-water voyaging at the lowest sea level is therefore well established.

What the evidence shows: Fossil human footprints at White Sands National Park in New Mexico were first radiocarbon dated in a 2021 Science study, using Ruppia cirrhosa seeds, to between about 21,000 and 23,000 years ago. After skeptics questioned the aquatic-seed dating, a 2023 follow-up study led by USGS geologists used two additional independent methods, radiocarbon dating of terrestrial conifer pollen and optically stimulated luminescence of quartz grains, and both reconfirmed the 21,000 to 23,000 year range (OSL gave a minimum age of about 21,500 years). The footprints therefore date to at least roughly 21,000 to 23,000 years ago, placing humans in North America during the Last Glacial Maximum, which is the substance of Hancock's point. Saying the prints are dated to more than 23,000 years slightly overstates the published upper bound: 23,000 years is the top of the range, not a floor, and the dates are not accepted by all researchers.

What the evidence shows: Hancock accurately reports the published claim: in 2017, a team led by Steven Holen and San Diego Natural History Museum paleontologist Thomas Demere published a Nature paper arguing that broken mastodon bones and stones (interpreted as hammerstones and anvils) at the Cerutti Mastodon site reflect marrow extraction by hominins about 130,000 years ago, more than 100,000 years before the widely accepted arrival of humans in the Americas. However, the claim is a contested outlier, not established science. Most leading archaeologists reject it, noting the absence of stone tools or debitage, no human remains, fracture patterns (including a shattered tooth with no marrow) consistent with natural or construction-equipment damage, and uranium-series dates that do not by themselves prove human agency. As of 2026 the mainstream consensus still places human arrival in the Americas at roughly 13,000 to 26,000 years ago, so Hancock reports the finding faithfully but understates that it is a widely disputed interpretation rather than a settled discovery.

What the evidence shows: Archaeologist Tom Dillehay led excavations at Monte Verde in southern Chile, and his peer-reviewed work dates human occupation there to at least roughly 14,500 calibrated years before present, with a lead author paper reporting ephemeral human activity between about 18,500 and 14,500 cal BP. That is older than the Clovis-first model, which held that people first entered the Americas around 13,000 years ago, and the Monte Verde findings are widely credited with overturning that model. Dillehay's own publications state that the old Clovis-first model of entry around 13,000 years ago no longer explains the peopling of the New World. A 2025 study by Surovell and colleagues has recently disputed the site's antiquity, but Dillehay and much of the field continue to accept the pre-Clovis dating, so Hancock's characterization of the discovery is accurate.

What the evidence shows: The archaeological consensus places the anthropogenic dark earth (terra preta) itself in the Late Holocene, with most deposits dated to roughly 2,500 to 500 years before present and widespread formation peaking around 1,000 years ago. The oldest figure near Hancock's number comes from Silva et al. (2021), who reported microcharcoal and mineral inputs beginning about 7,630 years ago at the Caldeirão site, arguing these predate soil management for cultivation. That ~7,600 year date reflects a contested natural-origin hypothesis, not a settled date for terra preta formation: Lombardo et al. (2022) counter that Middle Holocene charcoal is commonly stratified in Amazonian soils and that such deep, older carbon does not date when the dark earth actually formed. So about 8,000 years rounds up the oldest reported (disputed) date and presents a minority interpretation of naturally deposited carbon as if it dated the terra preta itself.

What the evidence shows: The scale is well supported: demographic-history reviews place the post-1492 decline of Indigenous populations in the Americas at roughly 75 to 95 percent, so the 90 percent figure sits squarely within the accepted range. One peer-reviewed review states that during the initial phase of European colonization infectious diseases were the primary killer among Native American communities, with an estimate that about 95 percent of the deaths were caused by introduced pathogens (smallpox, measles, influenza and others) to which populations had no prior immunity, the so-called virgin-soil effect. The caveat is that the framing of disease versus genocide is contested: scholars increasingly note that colonial violence, enslavement, forced labor and social disruption also drove mortality and often created the conditions in which epidemics became so lethal, and whether disease was the sole primary driver remains a matter of active debate. In short, the 90 percent disease figure reflects mainstream estimates, but treating disease and colonial violence as cleanly separable causes oversimplifies the current scholarship.

What the evidence shows: The Turin King List (Royal Canon of Turin), the principal ancient Egyptian source for the length of Khufu's reign, assigns him 23 years, and the traditional construction estimate for the Great Pyramid is roughly 20 years, a figure originating with Herodotus. A 2025 peer-reviewed paper states that the construction duration of about 20 years matches roughly Khufu's 23-year reign per the Turin King List, and that the Wadi el-Jarf papyri (the Diary of Merer) indicate pyramid construction lasted until the very end of his reign. Hancock's framing, that the pyramid would have had to be built within Khufu's reign of about 23 years if it is his tomb, matches this mainstream Egyptological reasoning. The one caveat is that contemporary evidence (graffiti dated to Year 28-29 and the Diary of Merer at Year 27) leads some Egyptologists to argue his actual reign, and thus the possible build window, was somewhat longer than the Turin figure of 23 years.

What the evidence shows: The core claim matches the published record. Hublin and colleagues (Nature, 2017) reported Homo sapiens fossils from Jebel Irhoud, Morocco, dated by thermoluminescence to 315,000 years (plus or minus 34,000), making it the oldest securely dated evidence of our species and pushing the emergence of anatomically modern humans back to roughly 300,000 years. Hancock's figures of about 300,000 years for modern humans and 310,000 for Jebel Irhoud fall squarely within that dating range. His account of earlier views is looser: through the 1990s and 2000s the accepted oldest H. sapiens fossils were the Ethiopian Omo Kibish (about 195,000 years) and Herto (about 160,000) remains, with Skhul and Qafzeh in the Levant around 90,000 to 120,000 years, so the specific 50,000 and 110,000 year milestones he cites are approximate rather than exact. Overall the central scientific point, that the date was revised sharply older by the Jebel Irhoud finds, is correct.

What the evidence shows: The broad claim that most North American mounds and earthworks were plowed under and turned into farmland is well documented. The National Park Service records that of some 600 Hopewell earthworks once present in Ohio, most were leveled by settlers to flatten fields, and that nearly all of the many renowned earthwork complexes of southern Ohio were reduced by two centuries of plowing and later mechanized agriculture. The specific numbers Hancock cites, however, are not mainstream archaeology: the one-million total (and the paired figure of about 100,000 surviving) originates with independent researcher Gregory L. Little, who extrapolated far beyond the roughly 100,000 mound sites cataloged in the Smithsonian Bureau of Ethnology surveys of the 1880s to 1890s. Peer-reviewed and agency sources describe the original mound-building record as numbering in the thousands to tens of thousands regionally, with no continent-wide census supporting a pre-1500 total of one million. In short, the loss narrative is accurate while the one-million figure is a speculative estimate, not an established archaeological finding.

What the evidence shows: Danny Hilman Natawidjaja and colleagues did publish a peer-reviewed paper in the journal Archaeological Prospection (online 20 October 2023) arguing that Gunung Padang is a buried pyramidal structure built in multiple phases, with the earliest construction dated as far back as roughly 25,000 to 27,000 years ago, deep in the last ice age. Outside experts in archaeology, geophysics, and radiocarbon dating quickly challenged the conclusions, noting that the ancient dates came from soil samples not demonstrably associated with human-made artifacts, so the radiocarbon results could not reliably show human construction at those depths. On 18 March 2024 the journal retracted the paper, over the authors' objections, citing this dating flaw. The paper also acknowledged Graham Hancock for reviewing the manuscript, and the site featured in the first episode of his Netflix series Ancient Apocalypse. So Hancock is correct that the paper was published and peer reviewed, but its central 27,000-year pyramid claim was rejected by specialists and formally retracted.

What the evidence shows: The 2023 paper in Archaeological Prospection (Wiley) claiming Gunung Padang is a pyramid built up to 27,000 years ago was retracted in March 2024, but not, as Hancock says, without good reasons. The publisher and co-editors-in-chief cited a specific, substantive methodological error: the radiocarbon dates came from soil samples not associated with any artifacts or features that could be reliably interpreted as man-made, so the conclusion that the site is an ancient human-built pyramid was not supported by the evidence. Experts quoted by Nature and Scientific American (including archaeologist Flint Dibble) noted there was no clear evidence the buried layers were built by humans rather than shaped by natural processes, and that dating old soil does not date human construction. The retraction followed an editorial investigation prompted by expert concerns, which is standard practice, so the characterization of a baseless retraction driven purely by a hostile ganging-up is not accurate.

What the evidence shows: The account in 1 Samuel 5-6 conflates two distinct events. At Ashdod (1 Samuel 5) the Ark afflicts the Philistines with tumors, but the text gives no death toll and does not say 50,000 died there. The figure of about 50,000 comes from a later scene at Beth-shemesh (1 Samuel 6:19), where the victims are Israelites struck for looking into the Ark, not Philistines. That number is itself a well-known textual crux: the Hebrew reads awkwardly as fifty thousand and seventy, and many modern translations and scholars treat the fifty thousand as a scribal corruption and read only seventy. The affliction is described in Hebrew as tumors or swellings (ophalim), which scholars have interpreted as bubonic plague buboes, hemorrhoids, or tularemia, not as cancer.

What the evidence shows: The 3,000-year-old banana finding does not come from Easter Island (Rapa Nui). It comes from the Teouma site on Efate Island, Vanuatu, roughly 6,000 km away, where researchers identified banana (Musa) microparticles in the dental calculus of Lapita settlers dated to about 3,000 years ago (published in Nature Human Behaviour, 2020). On Rapa Nui itself, banana and other crop microfossils are associated with the post-settlement period (roughly the 13th to 15th centuries AD), not 3,000 years ago. Peer-reviewed work on Rapa Nui dates the earliest human occupation to about AD 1000 to 1300, consistent with the conventional archaeological view of settlement around AD 1200. No published study reports 3,000-year-old banana phytoliths from an Easter Island crater, so the claim conflates a Vanuatu discovery with Rapa Nui and does not overturn the accepted Rapa Nui settlement chronology.

What the evidence shows: The historical low point for Rapa Nui (Easter Island) was about 110 to 111 people, not 11, roughly a hundredfold understatement. A 2024 genomic study published in Nature reconstructs this collapse and states that in the 1860s Peruvian slave raiders kidnapped about a third of the population, taking captives to work in Peru, and that after a smallpox outbreak the population fell to an estimated 110 individuals (commonly dated to the 1877 census). The slave raids were therefore real and central, but the low point resulted from the raids plus a subsequent smallpox epidemic (carried back by repatriated captives) and colonial-era emigration, not from the raids alone. The directional claim that Peruvian slave raids devastated the island is accurate, but the specific number of 11 and the framing of a single cause are wrong.

What the evidence shows: Earth's axial precession is real, and Polaris is indeed the current North Star because the rotation axis presently points almost exactly at it. The rate of precession is about 50.3 arcseconds per year, which works out to one degree every roughly 72 years, matching Hancock's figure. The 'roughly 26,000 year' cycle is also correct, but the accepted modern value based on the measured rate is about 25,772 years, not the 25,920 he cites 'to be exact.' The 25,920-year figure is the idealized 'Great Year' (or Platonic year) derived from a round 50 arcseconds per year and 360 degrees, so the general framing is accurate while the precise number he offers is a traditional idealization rather than the actual measured value.

What the evidence shows: This restates the Orion Correlation Theory advanced by Robert Bauval and popularized by Hancock, which mainstream Egyptology and archaeoastronomy reject. Physical and textual evidence dates the three Giza pyramids to the Fourth Dynasty of the Old Kingdom, roughly 2550 to 2490 BC (about 4,500 years ago), built for Khufu, Khafre, and Menkaure, not to 10,500 BC. Professional astronomers who examined the alignment (notably Ed Krupp of the Griffith Observatory and Tony Fairall of the University of Cape Town) found the layout of the two outer pyramids sits about 38 degrees from north while Orion's Belt in 10,500 BC sat closer to 47 to 50 degrees, so the claimed fit for that epoch is not exact and requires inverting the constellation to force a match. Archaeologists more broadly characterize Hancock's lost Ice Age civilization thesis, which the 10,500 BC dating supports, as pseudoarchaeology that cherry-picks evidence.

What the evidence shows: The underlying astronomy is roughly sound: the Sphinx does face due east toward the equinox sunrise, and because the equinox precesses westward through the zodiac over a roughly 26,000 year cycle, the vernal equinox sun would have risen in Leo about half a cycle ago, near 10,500 BC (about 12,500 years ago). The problem is the implication that this dates the monument. Mainstream Egyptology, backed by the associated temple layout and building program, places the Sphinx in the reign of Khafre around 2500 BC, and the eastward gaze fits the Old Kingdom solar cult rather than a Leo alignment (the Sphinx faces due east at every era, so any Leo match is not distinctive). Hancock's broader thesis of an advanced civilization roughly 12,000 years ago is rejected by archaeologists and scientists for lack of any physical, textual, or archaeological evidence. The Leo-at-sunrise calculation is correct; using it to date the Sphinx to that epoch is not supported.

What the evidence shows: The water-erosion hypothesis, advanced by Robert Schoch and John Anthony West in the early 1990s, is a minority view rejected by mainstream Egyptology and most geologists, who date the Sphinx to Khafre's reign around 2500 BCE. Schoch himself argued for a construction date of roughly 7000 to 5000 BCE, not the 12,000-plus years the quote asserts, so pinning the erosion to the Younger Dryas (about 9700 BCE) overstates even the fringe claim. The premise that only Younger Dryas rainfall was heavy enough also fails: radiocarbon-dated climate records for the Eastern Sahara show a humid African Humid Period long after the Younger Dryas, with gradual desiccation of the Egyptian Sahara beginning only around 5300 BCE, meaning wet conditions capable of weathering limestone persisted for millennia after the Younger Dryas ended. Geologists including K. Lal Gauri and James Harrell further attribute the enclosure weathering to salt exfoliation, poor limestone quality, and wetting of the buried monument by sand and floodwater rather than direct rainfall over 12,000 years. There is no accepted evidence that the Sphinx is more than 12,000 years old.