The Yonaguni Monument

Summary

Stepped rock terraces lie in about 25 metres of water off Yonaguni, the westernmost island of Japan. A marine geologist has argued since the 1990s that they are a sunken man-made monument; the geologists who have challenged him describe sandstone that splits along its beds and joints.

The formation is real, large and strikingly angular. What does not survive scrutiny is the claim that anyone built it. The case for construction rests on one researcher's surveys and interpretation, his dates for it have moved by thousands of years, and no Japanese authority has recognised it as a cultural property.

What is documented

The find. A local diver noticed the formations in 1986, and a promontory on the island was unofficially renamed Iseki Hanto, or Ruins Point (National Geographic News, 2007). Skeptoid describes the diver as a representative of the island's tourism board scouting for hammerhead shark dive sites. Divers know the site as Iseki Point.

The formation. Masaaki Kimura, a marine geologist at the University of the Ryukyus (now emeritus), surveyed it from 1992 with divers, multibeam sonar and an underwater robot. His figures describe a rough rectangle "about 270 m in length, 120 m in width and 26 m in height", with its base about 25 metres down and its upper terrace about 5 metres below the surface. It has flat terraces, vertical faces and steps.

The rock. The formation is sandstone and mudstone of the Yaeyama Group, laid down in the Miocene, roughly 20 million years ago. The same Miocene sandstone forms the low mountains of the island itself. The geologist Robert Schoch of Boston University, who has made dozens of dives on the site, reported that as far as he could determine it is "composed entirely of solid 'living' bedrock", with no separate blocks placed in position (as quoted by Skeptoid).

Kimura's case. In a paper for the OCEANS '04 conference, Kimura described "artificial topographies constructed about 10,000 years ago": a stepped pyramid resembling a gusuku, the castle and temple complexes of old Okinawa, with a loop road, drainage canals, stairways, a gate and, 300 metres to the south-east, a "stadium". He reported tool scars, two animals "surmised to be turtles" carved in relief on the bedrock, and small scratched marks resembling a cross and a V, and wrote: "No rock fragments can be found at the foot of Iseki Point." In 2007 he told National Geographic News that he had identified ten structures off Yonaguni and five more off the main island of Okinawa, including a castle, a triumphal arch, five temples and a stadium, covering 300 by 150 metres.

Kimura's dates. They have changed:

• 2004 conference paper: about 10,000 years ago, from radiocarbon and beryllium-10 measurements on over 40 samples, with the site drowned by the rise in sea level after the last ice age.
• 2006 report: between about 5,000 and 2,000 years ago, with the site lowered by tectonic movement.
• 2007, in National Geographic News: at least 5,000 years old, on the dates of stalactites in drowned caves, and sunk by an earthquake about 2,000 years ago.

The official position. In 2007 Emiko Ishida, a spokesperson for Japan's Agency for Cultural Affairs, told National Geographic News that neither the agency nor Okinawa Prefecture recognised the remains as an important cultural property, and that neither had carried out research or preservation work on them. On 13 October 2016 the town of Yonaguni began a survey to make an academic assessment with a view to designation. The Yaeyama Mainichi reported that researchers still disagreed over whether the formation was artificial or natural, and that no academic evaluation had yet been made. Takayuki Ogata, an earth scientist at the University of the Ryukyus, saw it for the first time on that survey and said he thought more than 90 per cent of it was natural topography. The town's expert panel, which included Kimura, agreed the next day that it should be protected whether it was artificial or natural.

Leading explanations

Natural sandstone, split along its beds and joints and trimmed by the sea (well supported; disputed by Kimura). Sandstone of this kind is layered into horizontal beds and cut by sets of vertical cracks called joints, and it breaks along both. Schoch told National Geographic News: "It's basic geology and classic stratigraphy for sandstones, which tend to break along planes and give you these very straight edges." A page published by the seismology laboratory of the University of the Ryukyus makes the same argument in detail: the jointed sandstone splits easily at right angles and along flat planes, waves strip it away in blocks and sheets along the joints, and the site was exposed as waves cut back the headland after the sea reached roughly its present level, about 6,000 years ago. It adds a point of logic: rock soft enough for people to work is soft enough for waves to erode. On land, Ogata and colleagues describe landforms at the island's Sanninu-dai geosite that are controlled by faults and joints.

Schoch's own conclusion is that it is "primarily a natural structure". He notes that the island itself has "evidence of very ancient human habitation, including tombs and other structures artificially carved from the bedrock." He does not believe the underwater formation was built, though he allows that parts of it may have been "touched up" by ancient people, who may also have used it.

Construction by people (Kimura's hypothesis; not confirmed by any independent study the registry found). Kimura reads the right angles, flat surfaces, tool scars, carvings and clean floor as evidence of work. He has had support: Toru Ouchi, a seismologist at Kobe University who has dived on the site, told National Geographic News that Kimura's account was "not exaggerated at all." Schoch's reply on the geometry was that "the right angles and symmetry don't add up in many places."

Current status

Debunked. The Yonaguni formation exists and is worth seeing. The claim that it is a monument does not survive examination: it is part of the bedrock, its steps and right angles are what this sandstone does when it breaks, Kimura's evidence of tool marks and carvings has not been confirmed by any independent study, and his dates have shifted from about 10,000 years to between about 5,000 and 2,000 years. As of September 2026 the registry has found no designation of the site as a cultural property and no independent study establishing human work. What would change the finding is a published, independently checked demonstration of worked stone or artefacts in place on the formation.

Last reviewed: September 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.