Note 049

Ikiryō and Shiryō Are Not a Matched Pair

By KWAIDANOTE · Published September 28, 2026

A Pair That Looks Ancient

English explanations of Japanese ghosts usually open the same way: there are shiryō, the spirits of the dead, and ikiryō, the spirits of the living, and the second is the unusual one. Two halves of a system, presented as though the language had always had both.

The written record does not support the symmetry, and the way it fails to support it is worth setting out, because the words arrive at very different times and for very different jobs.

The Dead One Is Older by Three Hundred Years

The standard historical dictionary of Japanese gives, as its earliest citation for shiryō, a passage from the collection of Buddhist miracle tales compiled between 810 and 824 — a line arguing that if even the dead spirit and the bare bones behave in this way, how much more should a living person not forget a kindness.

For ikiryō in its familiar sense — the spirit of a person still alive — the earliest citation is the Konjaku monogatarishū, around 1120, in the twenty-seventh book: a woman of the province of Ōmi whose living spirit has entered someone.

That is a gap of three centuries. The dead spirit was available as a written word before the living one existed on paper at all.

It is worth looking at how that first ikiryō is established, because it is not established at all. In the tale, nobody examines the afflicted person and announces a diagnosis. There is no exorcist who puts the question and gets an answer. The identification arrives afterwards, in the form of a report about what people were saying — that it was the living spirit of a certain woman in Ōmi. The earliest written instance of the term is a piece of gossip, attached to an event after the event was over, which is a fair description of how most of these identifications work in the older material.

And the First “Seirei” Is Not a Spirit

The gap is wider than it looks, because the characters now read ikiryō do turn up earlier — in a legal commentary of 833 — and there they mean something else entirely.

Read in the Chinese way, 生霊 is an ordinary term for the living people, the population, the subjects of a ruler. That is how the 833 text uses it. The dictionary’s earliest citation for those characters read in the supernatural sense is a puppet play of 1699.

The characters we now translate as “living spirit” spent their first nine hundred years in Japanese writing meaning “the general public.”

What the Most Famous Case Writes Instead

Our note on the Lady of the Sixth Ward sets out what the Genji chapter actually has on the page: not ikiryō but ikisudama, an older and differently built word, and mononoke written in plain kana rather than characters.

So at the moment this idea reaches its most famous expression, in the early eleventh century, the vocabulary we file it under is not yet in use. The example came first and the label was fitted later.

When They Do Become a Pair

They do eventually pair, and the earliest pairing found for this note is in a thirteenth-century text of the Heike material: the Chancellor’s living spirit and dead spirit alike should be pacified. Both words in one clause, treated as two cases of one problem.

That is four hundred years after the first shiryō and a century after the first ikiryō, which is roughly how long it takes for two words to notice each other.

Notice also what the sentence is doing with them. It is not distinguishing the two; it is disposing of both at once. Whatever is attached to that man, living or dead, is to be pacified by the same means. The first time the pair appears in one clause, the point of putting them together is that for practical purposes it does not matter which of the two you have.

Where the Distinction Actually Does Work

The neat opposition is at its clearest not in classical literature but in modern practice.

A history of Japanese female mediums published in 1930 sets out the work of the itako of the northeast as three kinds of calling: the dead mouth, the living mouth, and the god mouth — summoning the spirit of someone who has died, the spirit of someone alive but far away, and the voice of a deity. The ritual handling differs between them, down to which way the beads are turned.

The tidy pair is sharpest where the evidence is most recent. That is the reverse of the usual presentation, in which the distinction is ancient and modern practice is a survival of it.

The Word That Does Not Fit Either

There is also a third term, and it belongs to neither column.

Rihonbyō, the detached-soul sickness, appears in a story of 1925 with a gloss explaining it as a popular name of Chinese origin for the state in which the soul leaves a living body — and attached to the belief that a person who sees their own shape will shortly die. In that story a character says flatly that no such thing can exist, and the narrator uses the word anyway.

It is not jealousy and not revenge. It is a symptom, with a prognosis. The living spirit of the classical tales and the detached soul of the sickbed are the same event described by people with entirely different problems.

What the Sequence Shows

One reading is that Japanese always distinguished the living spirit from the dead one and simply took a while to write it down.

The dates suggest something plainer. The words were picked up at intervals of centuries, by Buddhist compilers, by court lawyers using a Chinese idiom, by tale collectors, by playwrights, and by twentieth-century mediums with a working procedure, and each of them needed the word for a different reason. Read backwards it resolves into a taxonomy. Read forwards it is a series of people reaching for whatever term was to hand.

There is a practical consequence for anyone reading this material in translation. When an English version says “living spirit,” it may be rendering any of several Japanese words, of different ages and different registers, and the tidy pairing may have been supplied by the translator rather than found in the text. That is not carelessness on the translator’s part — there is often no other English available — but it does mean that the symmetry a reader sees in English is frequently the last thing added rather than the first thing there.

Sources

Nihon ryōiki 日本霊異記, compiled between 810 and 824; cited in the standard historical dictionary of Japanese as the earliest attestation of shiryō 死霊.
Konjaku monogatarishū 今昔物語集, c. 1120, book twenty-seven; earliest attestation of ikiryō in the sense of a living person’s spirit. Wikipedia
Ryō no gige 令義解, 833, for the characters 生霊 used in the sense of the populace; and a puppet play of 1699 for the earliest supernatural reading of the same characters.
Heike monogatari 平家物語, Kōya manuscript line, early 13th century, for the pairing of the two terms in one clause.
Nakayama Tarō 中山太郎, history of Japanese female mediums, 1930; the three kinds of calling practised by the itako.
Okamoto Kidō 岡本綺堂, “Rihonbyō” 離魂病, first published 1925.