Kotoribako: The Cursed Box Born on a Forum

On 3 April 2024, the publisher KADOKAWA released a novel built around a small wooden box — a box that has no physical existence, only a textual one. The book is presented as a fresh interpretation of an urban legend that has circulated in Japan since the middle of the 2000s. The legend is called kotoribako (コトリバコ), and its first appearance can be traced to a single forum post whose archived copy is still readable today.

Quick facts

• Verdict: kotoribako is an internet-born urban legend that first appeared in a 2005 forum post; folklore researchers analyse it as a highly crafted piece of fiction rather than a report of real events.
• Name: kotoribako.
• Type: net kaidan — a scary story that originated and spread on the Japanese internet.
• Earliest known record: 12:57 on 6 June 2005, post 912 in a scary-story thread on the occult board of the forum 2channel, checked directly in the thread’s archive.
• Where it spread: into a dedicated thread opened the same evening, then into Japanese occult blogs, reference books on the uncanny, and a commercially published novel.

What the story says

Summarising kotoribako honestly means summarising it thinly, and that is deliberate. In outline, the story concerns a cursed wooden box. It is said to have been made using parts of the remains of children who died through mabiki (間引き), the historical practice of infanticide, and the curse attached to the object is said to bring harm to children and to women who come near it. That is the core of the premise as it is described in reference summaries: an object, an origin in the deaths of children, and a danger that falls on a specific category of people rather than on anyone at random.

The original telling goes much further, into additional narrative details that this article has chosen not to reproduce. What remains — a sealed container that carries the weight of dead children — is the image that made the legend travel.

Earliest known record

The story first appeared as post 912 in the 99th thread of a long-running 2channel series whose title translates roughly as “Why don’t we collect scary stories that are deadly serious?”, on the forum’s occult board. The post was made at 12:57 on 6 June 2005, under the name Kobako (“small box”). The thread survives on the 5ch past-log server, and this record was checked there directly rather than through later summaries. Yoshida Yuki’s account of the legend gives the same thread and places the post in the early afternoon of that day.

The reaction was immediate. At 17:46 the same day, a different user opened a dedicated thread titled Kotoribako, citing the stir caused by post 912, and the original poster’s first message in it followed at about 22:35 that night. The longer posts that set out the box’s supposed origin came later, on the night of 8 June 2005, in that same dedicated thread; this article does not summarise their content.

Neither the archive nor Yoshida’s account points to any earlier telling. That makes 6 June 2005 a firm earliest date, although an older record cannot be ruled out entirely.

What is documented

What is solidly documented about kotoribako is not the box but the text. The date and time of first posting, the board it appeared on, and a dedicated thread opened within hours can all be read in the archived threads themselves. So can its afterlife in print: Japanese reference works on modern strange phenomena and kaidan — including a 2018 encyclopedia of contemporary Japanese strange phenomena by Asazato Itsuki and a 2022 study of contemporary kaidan by Yoshida Yuki, the same researcher discussed below — treat the story as source material worth cataloguing, which is unusual for a forum post and a good measure of how far it travelled.

Researchers have also examined what the story does. Yoshida, in the 2022 book excerpted by Bunshun Online, places later net legends such as Kyotō-o, Yamanoke and Hasshaku-sama in a lineage that follows kotoribako. He argues that its storyline — horror hidden in a village’s customs — set the direction of Japanese internet ghost stories in the 2000s. In other words, kotoribako’s documented importance is as an influential piece of writing with traceable descendants.

The folklorist Hirota Ryuhei has also written on this territory. His essay on “villages and stations” in internet ghost stories appeared in the magazine Eureka, volume 54, number 11 (2022), pages 222 to 229. Only the bibliographic record was checked for this article, so the essay is noted here as further reading, not as evidence for any particular reading of kotoribako.

What is contested

The story’s claim to be a true account does not survive examination. Yoshida has pointed out inconsistencies in its historical setting. The backstory is set in the first years of the Meiji era, but 1868 was the year midwife regulations prohibited abortion by midwives. In western Japan, moreover, abortion rather than infanticide had been the main practice since the Edo period. The story also requires a scale of hidden killing that its own setting cannot explain. On that basis, Yoshida treats at least the historical part of kotoribako as invented — the work of the poster or of an informant within the story — rather than as a record of something that happened.

That assessment matters for more than accuracy. Yoshida is explicit that being fiction does not lower the story’s value; he is interested in why a historically mistaken picture of child-killing frightened readers so effectively. Treating kotoribako as fiction is not a downgrade — it is the correct category, and the more responsible way to read a story that presents itself as suppressed local history.

What the sources add

The archived threads, Yoshida’s essay and the reference summaries agree on the basics: a single forum post on 6 June 2005, followed by an unusually fast public reaction. Where they differ is in precision and attribution.

The first difference is timing. The reference summary this article first relied on describes a first post on 6 June and a follow-up in a dedicated thread on 8 June, which can suggest the dedicated thread came two days later. The archive shows a tighter sequence. The dedicated thread was opened by someone else the same evening, and the poster joined it that night. The 8 June date belongs specifically to the longer posts about the box’s origin. That separation matters, because it shows the legend’s fame was established by the first day’s posts, before its backstory existed in full.

The second difference is attribution. The argument that kotoribako shaped the later image of a village hiding something, with Hasshaku-sama among its descendants, was credited to Hirota Ryuhei in an earlier version of this article, following a reference summary. In the source checked here, it is Yoshida’s assessment, stated in his own name. Hirota has published on village-themed net kaidan, but his text was not available to confirm a comparable claim.

The third point is what the sources do not settle. The archive confirms when and where the story was posted, not whether anything in it is true. Yoshida’s analysis addresses that question, and his answer is that the historical frame is a construction. None of the sources used here identifies the poster, and this article does not attempt to.

Adaptations

The most visible adaptation is the one this article opened with: Shin’yaku Kotoribako, a novel written by Teshirogi Shotaro from an original concept by Madara Ushi, with illustrations by Yotaro, published by KADOKAWA on 3 April 2024 and described by the publisher as a new interpretation of the urban legend. Alongside it sits the quieter ripple of the reference books, which place a 2005 forum post in the same catalogues as far older material. For a story that began anonymously on a message board, being reinterpreted by a commercial novel nearly two decades later is a considerable second life.