A field guide entry

Hyakki Yagyō Emaki

百鬼夜行絵巻

the parade, painted

ChronicleWing · House

The object, plainly

The picture everything else copies

A retelling, not a citation. What follows is the story as the most widely circulated popular accounts tell it — folklore encyclopedias, translations, and the versions that have travelled furthest. It is synthesized rather than keyed line by line to the dated primary texts, and popular versions drift from what those texts actually say. The cited, source-by-source record begins further down this page, and where the two disagree, the record wins.

Everything in this collection is, in a sense, marginalia on one painting that nobody captioned.

The Shinju-an scroll is kept at a subtemple of Daitoku-ji in Kyoto. It's attributed to Tosa Mitsunobu, sixteenth century, Muromachi period, and it is an Important Cultural Property.

It shows the night parade: a long procession of tsukumogami — awakened household objects — with some animals and plants among them, moving along from right to left as you unroll it. Old lutes and cooking pots and umbrellas and bits of armour, with faces and limbs, going somewhere together.

And it has no text.

Most copies of it have no kotobagaki — the explanatory passages a narrative scroll would normally carry — which means the exact meaning is, as the scholarship puts it, unclear.

The founding image of the Japanese monster tradition is a picture with no caption.

Three honest uncertainties

No text. Nobody wrote down what it's about. Every interpretation of the painted night parade is a reading of images.

No fixed order. Copies were made from the Muromachi period through the Meiji, in Japan and abroad — and the yokai appear in different orders. Copyists rearranged the parade. Which means there is no canonical sequence, and the procession you're looking at reflects one copyist's decision about who goes where.

No agreed prototype. Scholars cannot settle whether the Shinju-an scroll is the original or one lineage among several. It may be the headwater; it may be a sibling.

The one with words is in Manhattan

There's an exception, and it's a good one.

A copy in the New York Public Library's Spencer Collection carries narrative text — and that text says the parade is a procession of supernatural beings appearing at an abandoned mansion, after 1177.

An address and a date, for a picture that has neither in Kyoto.

And it gives the whole tsukumogami doctrine its anchor: objects discarded at the end of an era, parading through the ruin of the house they were thrown out of. 1177 sits just before the war that destroyed the old order. Whether the copyist meant that or not, the reading is available.

Two parades, one name

Worth flagging because it clears up a real confusion.

The painted parade — this one — is a carnival of old utensils.

The literary night parade, in the Heian tale collections, is something else: demons that kill courtiers who cross their path, avoided by staying home on the scheduled nights and wearing the right charm.

Those are different institutions sharing a name. One is a warning. One is a joke about the rubbish heap. The scroll tradition and the text tradition are not illustrating each other.

Why this is the deepest stratum

Sekien's printed catalogs — the spine of this whole field — are the Edo printing of this Muromachi painting tradition. A great many creatures here trace back to "a figure in the emaki": the spike-headed thing, the shaggy one over the gate, the badger with the bag, the crawling quiver.

Which means that below the fabrications, below the puns, below the invented citations, the foundation of the canon is an unexplained picture, copied with variations, for four hundred years.

Not a corrupted text. Not a lost original. A painting that was never captioned in the first place, that everyone since has been annotating.

Field guide

At a glance

Behavior
Source-object — the Muromachi scrolls that painted the parade: tool-spirits, animals, and plants processing, famously with no explanatory text.
Oddities
  • The canon's foundation is an unexplained picture copied with variations: no text, no fixed order, no agreed meaning — every copyist reshuffled the marchers.
  • One copy with words lives in the New York Public Library, dating its parade to an abandoned mansion "after 1177."
  • Sekien's catalogs — and this bestiary — are marginalia on a parade nobody captioned.

Overview

The Shinju-an scroll (Daitokuji, Kyoto): attributed to Tosa Mitsunobu, 16th century Muromachi, Important Cultural Property — the night parade painted: a procession of tsukumogami, with animals and plants, and famously no explanatory text — "most scrolls lack kotobagaki, making the exact meaning unclear" [ja-emaki]. The wordless original under everything catalogued here.

The record

  • The one with words: the New York Public Library's Spencer Collection copy carries narrative passages — "a parade of supernatural beings appearing at an abandoned mansion after 1177" [ja-emaki]: the parade given an address and a date, in Manhattan.
  • The copy tradition: versions from Muromachi through Meiji, at home and abroad, with the yokai in different orders [ja-emaki] — the parade reshuffled by every copyist; scholars cannot settle whether Shinju-an is the prototype or one lineage among several [ja-emaki].
  • The doctrinal gap: the painted parade "differs from Heian-period accounts" in texts like the Konjaku [ja-emaki] — the literary hyakki-yagyō (demons that kill courtiers who cross them) and the painted one (a carnival of old utensils) are different institutions sharing a name.

The reading

The catalogs' missing preface. Sekien's books — this project's spine — are the Edo printing of a Muromachi painting tradition, and this entry files the headwater with its three honest uncertainties: no text, no fixed order, no agreed meaning. That's the authenticity-spectrum's deepest stratum — before the fabrications, before the puns, the canon's foundation is already an unexplained picture copied with variations: the whole tradition is marginalia on a parade nobody captioned. The naming-collision with the literary hyakki-yagyō belongs on the doctrine page; the NYPL copy's post-1177 mansion gives the tsukumogami doctrine its narrative anchor (objects discarded at an era's end, parading). And every entry whose ancestor is "a figure in the emaki" (waira, otoroshi, the spike-head behind yamaoroshi, the quiver behind furu-utsubo) now has one place to point.

Two-track labeling

The image

A detail of the night parade scroll: a spiky beast and robed goblins march behind a green banner held on a dragon-headed pole.
Historical plateHyakki Yagyō emaki (Edo-period copy), detail — via Wikimedia Commons
A still-life portrait of the Hyakki Yagyo Emaki, the parade painted: a partially unrolled handscroll rendered in flat ochre and washi-cream, lying at a slight diagonal.
new art, CC0Profile

Citations

The record

  1. [ja-emaki]「百鬼夜行絵巻」ja.wikipedia via WebFetch (retrieved 2026-07-26), Tier C.modern reference Japanese Wikipedia's article on this subject, as retrieved in 2026 — often carries regional detail the English article lacks.All-source record

Where the evidence thins

Open questions

  • Add to 02-HISTORICAL-CATALOGS.md as the pre-Sekien stratum (with the Matsui scroll per umi-zato).
  • The Spencer Collection copy's text (translation status).
  • Tsukumogami-ki (the companion sermon-text tradition) — separate source entry.

Keep walking