Nopperabo: The Faceless Encounter

On April 2, 1904, Houghton Mifflin Company published a slim collection of Japanese ghost stories by Koizumi Yakumo, the writer known in English as Lafcadio Hearn. One of its seventeen tales, “Mujina”, turns on an encounter with a figure whose face has no eyes, no nose and no mouth — and then on a second encounter that delivers the same shock all over again. That image is what most readers outside Japan picture when they hear the word noppera-bo, even though the story itself never uses the word.

Quick facts

• Verdict: Not an internet invention. The noppera-bo is traditional folklore, documented in Japanese kaidan collections from 1663 onward — more than two centuries before Hearn, and centuries before online message boards.
• Name: noppera-bo (のっぺらぼう); also recorded as nupperibo in the eighteenth century.
• Type: A faceless apparition described in Edo-period kaidan collections and later illustrated in printed books.
• Earliest known record: 1663 (Kanbun 3), in the collection Sorori monogatari, set in Oike-cho, Kyoto.
• Where it spread: Across Japan under regional names, from Tsugaru Hirosaki in the north to Kumamoto in the south; it reached English-language readers largely through Hearn’s 1904 book Kwaidan.

What the story says

The core of the legend is an absence. The creature appears in a human-like shape — it can be mistaken for an ordinary person until you look directly at it — but the face carries no eyes, no nose and no mouth, only flat, smooth skin. In one regional description the face is likened to a white gourd, with neither nose nor eyes. Its purpose, in the accounts, is essentially the shock itself: the Japanese material stresses that startling people is the main effect, and that anything worse than fright is rare.

The oldest recorded case is short and pointedly incomplete. In the 1663 collection Sorori monogatari, a noppera-bo of seven shaku in height appears in Oike-cho, Kyoto; the account says nothing about what the thing actually was. A century later, the 1767 collection Shinsetsu Hyakumonogatari records a variant called nupperibo making an eerie appearance at Nijogawara in Kyoto, and by 1779 the figure had been given pictorial form in Koikawa Harumachi’s illustrated book Yokai shiuchi hyobanki. Names shift from region to region: zunbera-bo in Tsugaru Hirosaki, and jubako-baba and nopperapon in Kumamoto, among others. What stays constant across the versions is the blank face and the moment of recognition that comes too late.

Earliest known record

The earliest documented reference is the Sorori monogatari entry of 1663. Secondary accounts in English and in Japanese agree on it — same date, same place, same detail that the text declines to explain what the apparition really was.

Two caveats belong here. First, the original text of Sorori monogatari was not examined directly for this article; the 1663 date rests on secondary description, reliable but second-hand. Second, “earliest confirmed” is not the same as “earliest”: an older record cannot be ruled out. Treat 1663 as the oldest firmly dated appearance of the named creature, not as its birth certificate.

A separate question is when the word itself turns up in print, and the answer comes from a dictionary rather than a ghost-story collection. The Japanese dictionary Seisen-ban Nihon Kokugo Daijiten cites as an example of the word a short passage from Tosei kagai dangi, a dangibon printed in 1754 (Horeki 4), where it is written with the characters 混沌房 and read as noppera-bo. That date falls between the 1663 and 1767 records. It is a milestone for the word, not for the apparition, and it does not move the earliest record of the faceless figure away from 1663. Because the Sorori monogatari text was not checked here, this article cannot say what wording the 1663 entry itself uses.

What is documented

Three things are on solid ground. The name and the image are attested in dated Edo-period print: 1663, 1767 and 1779 give a clear textual and then visual chain, with Kyoto as the setting for both of the early written accounts. The physical description is consistent across the sources — human form, featureless face — and so is the modest scale of the threat, which is fright rather than injury. And the regional spread is real: the variant names recorded from Tsugaru Hirosaki to Kumamoto show a motif that travelled and was renamed locally rather than one invented in a single place and copied.

The Hearn side is documented too, as a publishing fact. Kwaidan: Stories and Studies of Strange Things appeared in 1904, published by Houghton Mifflin on April 2 of that year. It gathers seventeen ghost stories, “Mujina” among them; counted with three short essays on insects, the book has twenty pieces in all. Most of the contents were retold from older Japanese texts or from legends Hearn heard from people around him, including his wife Setsuko, with his own interpretation layered on — though some pieces, such as his version of Yuki-Onna, were notably original work.

The 1779 illustrated book is documented at the level of the object itself. The National Diet Library’s bibliographic record for Yokai shiuchi hyobanki lists Koikawa Harumachi (1744–1789) as both author and illustrator, gives the publisher as Urokogataya Magobei, and describes a single volume, 18 cm tall, published in An’ei 8 (1779). The library holds the original, and its digitised images are available online under the permanent identifier DOI 10.11501/10301827. The record confirms the book, its date and its makers; it does not itself describe the page on which the faceless figure appears, but it means the 1779 depiction can be checked against the book rather than against a description of it.

The Kumamoto name jubako-baba also has a paper trail outside the encyclopedias. The folklore database of the International Research Center for Japanese Studies records a folktale from Tamana district, Kumamoto, collected by Noda Taro and printed as “Tamana-gun mukashibanashi (3)” in the journal Mukashibanashi kenkyu (vol. 1, no. 4, Sangensha, August 5, 1935). In that tale a tanuki often appeared transformed into an old woman carrying a jubako, a tiered food box, and was called jubako-baba. A reference answer in the National Diet Library’s Collaborative Reference Database adds that local sources write and read the name in several ways — with and without a long vowel, ending in baba or babaa — and cites among them a 1966 Kumamoto dialect dictionary, Yumoa Kumamoto jiten: Higo kyoku hogen jiten, and issue 57 of the magazine Kabuki. The name is therefore not a one-off: it recurs in Kumamoto material across at least three decades.

What is contested

The contested point is the link between the two halves of this article. “Mujina” is routinely associated with the noppera-bo legend, and it is the reason the faceless figure is recognisable in English at all. But the story never names the creature noppera-bo, and Hearn’s great-grandson, Koizumi Bon, has described the tale as Hearn’s own creation rather than a straight transmission of an existing legend. Given that Hearn’s method in Kwaidan ranged from close retelling to outright invention, the honest position is that “Mujina” belongs beside the Edo-period record, not inside it — a famous literary treatment of the faceless motif whose direct line of descent from the older noppera-bo material is not established.

One further distinction is worth keeping straight, because the title invites confusion. Mujina is an old Japanese word for certain animals, and the yokai called noppera-bo is a separate concept. Japanese reference material treats them as related topics, not as the same thing, and this article does the same.

The regional list raises a smaller question of its own. Jubako-baba appears in encyclopedic lists of noppera-bo variants, but the 1935 folktale record behind the name describes a shape-changing tanuki and makes no explicit link to a missing face. On the records consulted here, jubako-baba is firmly attested as a Kumamoto tradition; its place in the faceless family rests on the encyclopedic grouping, not on the early folklore record itself.

What the sources add

Set beside the popular telling, the institutional and dictionary sources mostly confirm the outline while sharpening the dates. The Edo-period chain of 1663, 1767 and 1779 holds, and its last link is now anchored in a library catalogue: the 1779 book survives, with a named publisher and a digitised copy, rather than existing only as a line in a summary. The word itself gains a dated example from 1754, which shows the name in circulation between the two Kyoto accounts. What the sources do not change is the weakest point: the 1663 entry is still known only second-hand, since no original text of Sorori monogatari was reached.

On regional variants, the picture is less tidy than a list of names suggests. Jubako-baba is well documented as a Kumamoto tradition, from a 1935 field report to a 1966 dialect dictionary, but in the oldest of those records it is a tanuki in disguise, not a blank-faced stranger. The grouping of such figures under noppera-bo is a later act of classification, and that is worth knowing when reading any modern list of variants. No comparable institutional record was located for the Tsugaru name zunbera-bo.

The link to Hearn shows the same pattern from the other direction. A current Japanese dictionary entry describes the faceless creature, notes that similar legends are told in many regions, and mentions that Hearn’s Kwaidan contains an account of it. That is a fair picture of how the figure is understood today, but it is a modern association: “Mujina” never uses the word noppera-bo, and its place in the older tradition remains contested.

Adaptations

The one adaptation the sources support is the one this article opened with: Hearn’s “Mujina”, published in Kwaidan in 1904. Its reach is easy to explain. The book has long since entered the public domain, and its full text circulates freely through Project Gutenberg and Wikisource, alongside a public-domain audiobook recording from LibriVox. For a legend whose Japanese record consists of a handful of terse entries in seventeenth- and eighteenth-century collections, that free availability did a great deal of work: a reader anywhere can encounter the faceless figure in a few hundred words of English prose, which is more than can be said for Sorori monogatari.