Most ghost-story sites never correct anything. A legend gets written once, a few invented names get bolted onto a real location, and it sits there unchanged for years — copied, re-copied, and cited as fact by the next site that scrapes it. We decided early on that this site would work differently: when research turns up a fabricated detail, a mismatched date, or a name we can't verify anywhere outside our own article, we say so in the article itself, and we keep the record here.
This page exists for one reason. If you're going to build something like a Wikipedia of world folklore, you have to be more careful than a folklore Wikipedia, not less. A cursed cave or a haunted tunnel with a great story is worth writing about, but it isn't worth inventing a farmer's name to make the story feel more specific. When we catch ourselves having done that, we don't quietly edit the paragraph and hope no one notices the old cached version. We leave a visible correction in the article, and we log it here.
Why we publish a corrections log
Three reasons, in order of how much they matter to us:
Trust compounds, and folklore sites usually spend it instead of building it. A reader who catches one invented name stops trusting the whole article — reasonably. Publishing our mistakes in the open, rather than hoping no one checks, is the only way we know to actually earn the benefit of the doubt on the stories we haven't gotten wrong (or haven't found out about yet).
It separates two different things this genre constantly blurs: recorded folklore and invented filler. A story like “locals report a ghost train near a tunnel” is genuine oral tradition, hedged appropriately, even when it can't be independently verified. A story like “a farmer named Torazo Onizuka found a bone-filled chamber” is a specific, falsifiable claim, and when we can't find any trace of that name anywhere except our own article, the honest conclusion is that we probably made it up, or copied it from another site that did. Both categories exist on this site. This log is where we've caught the second kind and fixed it.
It's a real, checkable claim about how we operate, not a marketing line. Every entry below links to the live article where the correction is visible, in the article's own words, dated to when we found the problem.
How a correction ends up here
Most of these started the same way: a new article gets written using AI-assisted research, a specific detail (a person's name, a casualty count, a construction date) turns up nowhere else when we go back and check it independently, and we treat that absence as a warning sign rather than something to paper over. When a detail can't be verified and looks fabricated, we don't just delete it — we say what we originally wrote, say what we could actually confirm, and leave both visible so the record is honest rather than tidy.
The corrections
The Haunted Story of the Cave Legend Known as Akifudo (Akiyoshido Cave, Yamaguchi) — An earlier version invented a farmer named Torazo Onizuka discovering a bone-filled chamber, and a ghost-tour guide named Chitose. Neither name turned up in any independent source, which is a strong signal given how thoroughly documented this cave actually is. We removed the invented names and replaced them with the tradition that's actually attached to the site: spirits of defeated Heike warriors from the 1185 Battle of Dan-no-ura said to linger in the cave.
The Haunted Story of Yuuki-Do Cave (Shimane) — Earlier research centered the story on an immovable boulder and a 1950s yokai encounter, and left out something more important: a real, documented tragedy. In June 2001, a 15-year-old boy drowned in the cave's water while exploring with friends. We rewrote the article to lead with the confirmed incident rather than the unverifiable folklore.
The Haunted Story of Hakozaki Crossing and Biyo-U Tunnel (Kanagawa) — Earlier research described a young girl who died at the crossing in the early 1900s. Independent local sources describe a different haunting entirely: American soldiers in wartime uniforms reportedly seen along the old railway line. We corrected the article to reflect the story that's actually documented for this location.
The Haunted Story of the Former Wada Tunnel (Akita) — The earlier version dated the tunnel to 1902 and centered the legend on a 1965 truck-bus collision. Japanese-language sources place its actual construction in 1961 — a roughly 60-year discrepancy — and describe a wider cast of apparitions than the single collision story we'd published.
The Haunted Story of Sasako Tunnel (Yamanashi) — The commonly repeated details — a 1,699-meter length, a 1910 construction date, a route between Narusawa and Kawaguchiko — don't match any documented tunnel by that name. The real, well-documented tunnel with a similar name (Sasago Tunnel) is 239 meters long, sits in a different part of the prefecture, and opened in 1938. We flagged the numbers as unverifiable rather than let them stand.
The Haunted Story of Kurose Tunnel (Hiroshima) — A named victim, “Tomomi,” who supposedly disappeared at the tunnel, had no source outside our own article. We removed the name and kept only the documented tragedy at the site: construction-related deaths tied to the nearby dam.
The Haunted Story of Ushikubi Tunnel (Ishikawa) — Earlier versions described specific 1952 and 1979 encounters with an ox-headed figure, naming a witness, Tsunetaro Kume. No trace of either encounter or the named witness could be found. What's actually documented is more modest: no single incident has ever been tied to this tunnel, and its reputation appears to come from repeated appearances on Japanese television ghost-story programs rather than any originating event.
The Haunted Story of Kashida Tunnel (Osaka) — A “White Lady” warning and a glowing-eyed black dog, both from the earlier version, didn't turn up anywhere else describing this tunnel. We dropped both rather than keep circulating unconfirmed additions, and replaced them with the tunnel's actual, verifiable history as part of an old pilgrimage route.
The Haunted Story of Akame Tunnel (Shimane) — The earlier page told a gentler story: a mother and daughter on a shrine pilgrimage in the early 1800s, trapped by a rockslide. That story doesn't appear anywhere outside our own article, and we couldn't verify “Akame Tunnel” as a separately named real location at all. We flagged the story as unconfirmed rather than continue presenting it as settled.
The Haunted Story of Tsurumai Tunnel (Chiba) — An earlier version described long-dead samurai marching through the tunnel with drums and hoofbeats. That story had no source outside our article. What we found instead, consistently, was a quieter local account: a sensed or seen presence near the tunnel, without any of the military framing. We removed the samurai version and published the account that's actually told locally.
The Haunted Story of Ikoku Dam (Hyogo) — This is the correction we're least proud of: an earlier version described “Ikoku Dam” as a 1924 recreational lake turned WWII defensive line, with a soldier's ghost in a white uniform. No record matched that account, and the site is almost certainly a different, real dam (Ichinokura Dam) with its own documented submerged tunnel. We rewrote the page entirely around what's actually there.
The Haunted Story of the Dankei Road Tunnel (Hakodate) — An earlier version described an abandoned 19th-century railway tunnel closed in 1975, haunted by a soldier from either the First Sino-Japanese War or the 1945 Hiroshima bombing — and got the tunnel's basic status wrong in the process (it's a road tunnel that carries live traffic every day, not an abandoned line). We replaced it with the story actually documented for this location, involving a mother and child rather than a wartime soldier.
The Haunted Story of the 8 Tunnels (Hachioji, Tokyo) — This is the most serious correction on this list. An earlier version described two invented derailments near Hachioji — a November 1944 troop train claimed to have killed 500 to 800 people, and a September 1945 freight train claimed to have killed nearly 1,200. We could find no record of either event. Given the scale of the claims, we went looking for what actually happened in the area during the war and replaced the invented derailments with a real, documented wartime tragedy instead of folding a fabricated body count into a ghost-train story.
The Haunted Story of Hon-Tani Tunnel (Tottori) — An earlier version described a railway tunnel and a 1942 derailment. The tunnel actually matching this name and location in Japanese sources is a 1972 road tunnel (formerly Funaoka Tunnel), with no railway history at all. We also could not verify a claimed detail about prefectural security cameras and removed it.
The Haunted Story of Tenri Tunnel (Nara) — An earlier version told a specific murdered-traveler story: an old man killed by an unknown assailant while resting in the tunnel, with a “screaming woman” ritual tied to his wife's revenge. We could not verify the murder or the ritual against any documented source, and noted that the reported acoustic effect has a straightforward natural explanation (echo behavior common to enclosed tunnels).
The Haunted Story of the Kinuura Seabed Tunnel (Aichi) — An earlier version stated the tunnel connected Aichi and Mie prefectures, dated to 1933, with construction-worker deaths tied to its build. Both the interprefectural claim and the date were wrong: the tunnel sits entirely within Aichi Prefecture and was built in 1973, not 1933.
The Haunted Story of Kamitomi Tunnel (Saitama) — Not a correction in the strict sense, but the same principle applied one step earlier: research for a follow-up pass turned up a second, more specific local legend that didn't match the ghost-train story this site had previously carried. Rather than quietly pick one version and discard the other, we published both side by side and said directly that we don't know which one — if either — has genuine local currency. We'd rather flag an unresolved discrepancy than manufacture false certainty.
What we haven't corrected (and why that matters too)
Several of the articles above still carry parts of the original story once independent local sources confirmed a detail we'd doubted — the point of this process is not to distrust everything by default, but to check the specific claims that read as unusually precise (a named person, an exact casualty count, a specific year) and correct only what turns out to be unsupported. Hedged folklore — “locals say,” “according to legend,” “we could not verify this” — stays hedged. It isn't a correction to leave a genuinely unverifiable oral tradition marked as unverifiable; that's just the article working as intended.
If you find something we got wrong
This log will grow. If you're from one of the regions covered here, or you have firsthand knowledge that contradicts something in an article — including anything not yet listed above — we want to hear about it. Reach out through our contact page with the article link and what you know, and we'll look into it. A correction on this site isn't a failure to be hidden; it's the process working the way it's supposed to.
Comments