Two traditions and a merge

There is a standard story about where paper folding comes from. Paper is invented in China, travels to Japan, and folding grows up alongside it as a courtly art which becomes a popular one and eventually reaches the West.

Every step in that story is a real event. Assembled into one line, they describe something that did not happen.

Three things called one thing #

The activities the standard story merges are not variants of each other. They differ in purpose, in who does them, and in what counts as doing them well.

Ceremonial wrapping is an etiquette. A gift is wrapped, a fold signifies something about the giver and the occasion, and the correct fold is prescribed. Success is conformity to a code. The practitioners are people who need to get a formal occasion right.

Recreational folding is an amusement. A shape is made because it is pleasing or clever, and success is the object being good. The practitioners are anybody with paper.

The kindergarten syllabus is a teaching method. Folds are performed to demonstrate geometric facts, and success is the child understanding. The practitioners are teachers.

Those are three different activities that happen to share a physical operation.

The evidence for each is different in kind #

The three lineages leave different traces, and the difference explains much of what the record looks like.

Ceremonial wrapping leaves manuals, because a code has to be written down to be enforced. The Ise and Ogasawara schools’ documents survive because they were the instruments of a formal practice, and they are late relative to the practice for exactly that reason: a convention is codified after it exists.

Recreational folding leaves almost nothing, because an amusement has no institution behind it. What survives is incidental — a poem mentioning folded butterflies, a book that circulated as a novelty.

The kindergarten leaves an enormous amount, because it was a curriculum distributed to schools with training materials and inspection.

Reading the asymmetry wrongly #

That asymmetry produces a specific error, and it is worth naming because it is the opposite of the usual one.

If the record is read at face value, the European educational lineage looks better established than the Japanese recreational one — more sources, better dated, institutionally documented. A naive reading concludes that folding is better attested in Europe.

It concludes that because schools generate paperwork and households do not. The asymmetry is a fact about documentation, not about activity, and it points in the opposite direction from the folklore, which over-dates the Japanese lineage. Both errors come from treating the record as a sample.

What each lineage did not have #

The clearest way to see that these are separate is to notice what each one lacks.

Ceremonial wrapping has no notion of difficulty. A wrapping is correct or incorrect; nobody attempts a harder one, and there is no progression. The whole apparatus of technique, of models that are beyond a beginner, of a folder getting better — none of it applies.

Recreational folding has no notion of correctness in the ceremonial sense. There is no authority to be wrong against, and a folder who invents a new bird has not violated anything.

The kindergarten has neither. Its folds are neither correct-by-code nor good-by-taste; they are demonstrative, and a fold that shows the geometric point has done its job whatever it looks like.

Three activities with three success conditions is about as strong a sign of separate origins as this kind of evidence offers. A single practice that had diversified would be expected to share at least one of them.

When they met #

The merge has a date, and it is more recent than almost anybody assumes.

Kindergartens spread through Europe and America from the 1840s. They reached Japan in the 1870s, as part of a deliberate importation of Western educational method during the Meiji period, and the folding exercises came with them.

So from about 1880 there is a European systematisation of paper folding being taught in Japanese schools, in a country with its own centuries-old folding practices. From that point the two are not separable, and anybody attempting to date a fold by which tradition it belongs to is working in a period when the traditions had already met.

The pajarita and the loose end #

The European lineage has its own unresolved question and it deserves stating rather than smoothing over.

Froebel assembled his syllabus from folds already current in German-speaking Europe. Where those came from is not answered by anything he wrote, because he was not asking. Spain has the pajarita, a folded bird with its own name and its own long presence in popular culture, and the earliest thing anybody can point at for it is an engraving of 1793.

The usual explanation is transmission through al-Andalus with papermaking, which is plausible, is centuries earlier than the evidence, and is exactly the kind of claim the record’s checker refuses to let pass unmarked.

The word arrives late too #

A small piece of evidence that the merge is recent: the vocabulary.

Origami as a general term for the practice is not old in Japanese, and it displaced other words — orikata, orisue — which had narrower and partly ceremonial senses. The 1797 book is orikata, not origami. The generalisation of one word to cover the whole activity is roughly contemporary with the kindergarten’s arrival and with the emergence of folding as a subject rather than as several unrelated things people did.

In English there is no native word at all. Paper folding is a description rather than a name, and origami was borrowed only in the twentieth century, which is why English-language accounts frame the whole subject as Japanese: the word arrived from Japan, so the practice was assumed to have.

That is a nice example of a general hazard. A borrowed name is evidence about where a word travelled and is routinely read as evidence about where a practice originated.

Convergent folds are not evidence of contact #

A tempting piece of evidence needs to be handled carefully, because it looks stronger than it is.

Several folds are common to the Japanese and European traditions: the same boat, the same box, the same preliminary arrangement of creases. That looks like proof of contact.

It is not, and the reason is the subject of a whole other rung. A square has a small number of natural first moves, the flat-folding conditions are restrictive, and two people exploring systematically cover the same ground. Shared specifics — the same sector angles, the same idiosyncratic sequence — would be evidence. Shared structure is what constraint produces.

What the merge did to the subject #

The consequence is not only bookkeeping. The merge changed what folding was.

A practice that is simultaneously an etiquette, an amusement and a teaching method has three constituencies with three standards, and what emerged in the twentieth century took something from each: the codification of the first, the ambition of the second, and the systematic ordering of the third.

The diagram notation is a good example of the mixture. It is a code, in the ceremonial sense — a fixed set of marks with agreed meanings. It exists to transmit recreational objects. And it works because the audience had been through a syllabus.

The rule that came out of the merge #

One artefact of the merge is the rule this site works under, and its history is now legible.

One sheet, no cuts is not attested in the oldest sources — the 1797 book cuts freely. It is a twentieth-century codification, and codification is the ceremonial lineage’s contribution: the instinct that there should be a correct way to do this.

Applied to a recreational practice, that instinct produces a rule about materials. Applied to a teaching practice it would produce a syllabus. The rule is the etiquette tradition’s habit operating on the amusement tradition’s material, which is what a merge looks like from the inside.

Napkins, and the case for a fourth #

The awkward case deserves its paragraph, since leaving it out is how the three-way division stays tidy.

Elaborate napkin folding was a courtly European practice in the seventeenth century, with printed manuals, named forms, and professional practitioners. It is folding, it is documented, it is earlier than Froebel by two centuries, and it fits none of the three lineages: it is ceremonial in purpose like Japanese wrapping, decorative in ambition like recreational folding, and connected to neither.

It is usually omitted from accounts of origami’s history because the material is cloth rather than paper, which is a reasonable line to draw and is a line about substrate rather than about practice. Drawing it differently gives four lineages, and there is no principled reason the count should stop there.

What is genuinely shared #

Something is common to all three lineages, and it is worth identifying because it is the only thing that is.

The material. All three depend on paper of a particular kind: thin enough to fold many times, strong enough to survive being creased, and cheap enough that ruining one is not a loss. That constraint is arithmetic and it is checkable, and it is the one thing genuinely upstream of every folding tradition anywhere.

So the standard story’s first step — paper comes first — is right. Everything after it is three stories.

What a merged chronology does to a claim #

The practical damage is best seen on a specific sentence, and the commonest one will do: origami has been practised in Japan for over a thousand years.

Unpacked against three lineages, that sentence is assembled from parts of two of them. The thousand years comes from paper’s arrival and from ceremonial practice inferred to be old. The word origami refers to the recreational lineage, whose earliest attestation is 1680. The implied continuity comes from nowhere at all.

Each component has some support. The conjunction has none, and it is the conjunction that gets repeated.

That is the mechanism by which a merged story generates claims stronger than any of its sources: dates migrate between lineages, because once the lineages are one thing there is nothing to stop them.

The thousand years, decomposed #

The disputed sentence can be taken apart arithmetically rather than rhetorically, and the arithmetic is more damaging than the objection.

A thousand years of practice, counted back from now, begins around 1026. Two dates cut that span into three pieces, and both are in the record above.

1680, the earliest attestation of recreational folding — Saikaku’s line about folded butterflies. The 1870s, when the kindergarten syllabus arrived in Japan and the two traditions stopped being separable.

So the thousand years is:

Six hundred and fifty years of inference. From 1026 to 1680 there is nothing attesting recreational folding at all. The practice may well have existed; the interval carries no evidence either way, and it is sixty-five per cent of the claim.

Two hundred years of attested, unmixed practice. From 1680 to the 1870s the recreational lineage is documented and is not yet entangled with anything European. This is the part of the sentence that is true as stated, and it is a fifth of it.

A hundred and fifty years of mixed practice. From the 1870s to now there is a Japanese recreational tradition and an imported European syllabus in the same schools, and no claim about either can be made without the other.

Which is the useful form of the objection #

Written that way the sentence does not have to be refuted. It has to be apportioned, and the apportionment is a stronger statement than a denial.

Nobody needs to argue that folding did not happen in Japan before 1680; the three-lineage account does not say it did not, and the ceremonial lineage certainly did. What the decomposition says is that the sentence’s authority comes from its longest component, and its longest component is the one with nothing under it.

Two hundred years is what the record supports, and two hundred years is a long time. A recreational practice attested continuously from before the industrial revolution is a genuinely old tradition by any ordinary standard, and it is what the evidence buys.

The reason the shorter number is never quoted is the one the next section gives, and the arithmetic sharpens it: the difference between the honest figure and the popular one is not a correction of a few decades. It is a factor of five, and four-fifths of the popular figure is either undocumented or documented as something else.

Why the merged story is attractive #

It is worth asking why the single-lineage version is so durable, because the answer is not stupidity.

A single continuous tradition is a better story. It has one origin, one line of descent, and a satisfying arc from ancient ceremony to modern art. Three separate practices converging by accident in the nineteenth century is a worse story and a more accurate one.

It also flatters everybody. A folder in the single-lineage version is the inheritor of a thousand years; in the three-lineage version they are the inheritor of about a hundred and fifty, plus some older material that was doing something else. Nobody has ever gained from the second version.

What would settle it #

The three-lineage account makes predictions, and it is worth saying which evidence would move it.

If the lineages were really one, there should exist Japanese recreational folding materially earlier than 1680 — some depiction, inventory or reference showing amusement rather than ceremony. There is not, and its absence is weak evidence rather than none, given how badly household practice survives.

If the European lineage descended from the Japanese one, there should be a transmission route with something on it: a traveller, a trade, an object. The Iberian route through al-Andalus is the standard proposal and has, so far, nothing dated on it earlier than papermaking itself.

And if the kindergarten’s folds were Japanese in origin, Froebel’s sources would show it. They do not; they show a German-speaking practice he drew on without comment.

None of that is proof. All of it is the sort of thing that could turn up, and the account is arranged so that it would be visible if it did.

The idealisation, named #

The three-way division is itself a simplification and should not be treated as the final word.

There are certainly more than three lineages. Ceremonial folding in Japan is not one practice but several — wrapping, noshi, the folded documents of formal correspondence — with different origins. European folding before Froebel is not one thing either. Napkin folding, which is an elaborate seventeenth-century European practice with its own printed manuals, sits awkwardly in every account and is usually left out.

Three is the number of lineages the record distinguishes clearly. It is not a claim that there were exactly three.

The subject the merge produced #

A closing observation about what the three lineages jointly made possible, since the essay is otherwise about separating them.

The modern mathematical subject needs all three contributions and would not exist on any one. It needs the recreational lineage’s objects — the bases, the classical models, the accumulated repertoire that gives the theorems something to be about. It needs the ceremonial lineage’s codification, without which there is no shared standard and no agreement about what counts as the same model. And it needs the educational lineage’s systematic ordering, which is where the idea of a sequence, and eventually of a pattern, comes from.

The crease pattern as this site treats it is all three at once: an object, held to a standard, described in an order. Nothing in any single lineage would have produced it.

Where this goes next #

The paper had to arrive first takes the one genuinely common ancestor and prices it, and the kindergarten was a geometry class follows the third lineage forward into the subject’s own methods.

That is the argument for taking the merge seriously rather than treating it as a bookkeeping correction: it is the event that made the subject possible.

The surprising connection to end on. The merge is usually presented, when it is presented at all, as the West learning origami from Japan. The record shows traffic in both directions and the better-documented flow going the other way: a German syllabus, taught in Japanese schools, within a generation of its invention. Two of the three lineages had already exchanged a great deal before anybody in either wrote the word origami in a Western language.