芸 and 性: What Japan Has Sold, and For How Long

A question that sounds modern is not. If someone is paid to keep you company for an evening, attentively and skilfully, and sex is not part of what was arranged — what exactly is being sold, and is that distinction real or is it a polite fiction? In most countries this is a new argument with no settled vocabulary. In Japan it is about four hundred years old, was settled by trade regulation rather than by philosophy, and had a slogan.

This page tells that history without the costume drama, which means telling the ugly half first. A version of this story that begins with geisha and elegance is not history; it is tourism.

From 1617 the Tokugawa authorities licensed prostitution and confined it to designated quarters, of which 吉原 in Edo is the famous one; it was rebuilt outside the city as 新吉原 after the great fire of 1657. The quarters were walled and gated, and the gates worked in one direction more than the other.

The women inside were, overwhelmingly, not there by choice. They were sold by impoverished families under indenture contracts — 年季奉公 — with terms of years, and the debt structure was arranged so that the term rarely ended when it was supposed to: clothing, bedding, food and medicine were charged against the woman, so working harder could increase what she owed. Movement was restricted. Disease was endemic and mortality was high; many did not survive their contracts.

The aesthetic apparatus around the highest-ranking courtesans — the processions, the robes, the poetry, the near-mythological status of a 花魁 — was real and was also a marketing surface on top of that. Both things are true at once, and a history that keeps only the woodblock prints has kept the advertising and thrown away the ledger.

The geisha profession emerged separately and later. Male 幇間 came first; female 芸者 appear around the middle of the eighteenth century. From early on they were a distinct trade, and — this is the part that surprises people — the distinction was enforced from the outside, by the authorities and by the licensed quarters themselves, for a commercial reason rather than a moral one.

The 遊郭 held a licensed monopoly on selling sex and did not want geisha competing with it. So geisha were registered through the 検番 system, their engagements were booked and logged through it, and they were barred from doing the courtesans’ business. The trade’s own phrase for the rule was 芸は売っても身は売らぬ — we sell art, not our bodies.

Two honest qualifications belong here. The rule was a rule, not a description of every life lived under it; individuals departed from it, and 旦那 patronage arrangements existed that were their own complicated thing. And geisha, too, were often placed as children under indenture, so the distinction was drawn inside a system of bonded labour rather than outside one. But the boundary itself was institutional and consequential. It had a registry, an enforcement mechanism, and a commercial constituency that wanted it kept.

In 1872, in the aftermath of an international incident involving a Peruvian ship and its indentured passengers, the government issued the 芸娼妓解放令, nominally releasing indentured prostitutes and geisha from their contracts. The announcement was more decisive than the outcome: the system continued in a modified legal form, with the debt reappearing in new arrangements.

The real break came with the 売春防止法 of 1956, in full force from 1958, which abolished the licensed quarters. Alongside it the 風営法 of 1948 supplies the taxonomy that still governs adult businesses. The combination produced the structure Japan has now, and it is worth stating precisely because it is so often described vaguely: a wide range of paid adult services is lawful and regulated, while one specific act sits outside what the law permits to be sold.

The consequence is a culture in which the question this page opened with is not abstract. Large numbers of people, on both sides of the counter, have had to hold a clear working line between what is being sold and what is not, for decades, as an ordinary professional matter. Whether every business holds that line is a separate question and the honest answer is no. But the line is legible here in a way it is not in most places, and it was legible long before the current law drew it.

The section on the 遊郭 refused the romance about that half. Leaving it there would sell the same costume drama at half price, with the 芸者 standing in as the clean side of a dirty trade, and the page would then be doing on its own account what it accused others of.

So: girls entered 芸者 houses as children, under 年季奉公 contracts, frequently sold by families in difficulty, and carried the cost of their training, clothing and keep as a debt that had to be worked off. A woman whose art was genuinely her profession could still be, in the same years, a person who could not leave. The training was real, the skill was real, the indenture was real, and none of the three cancels the others.

And the practice usually named 水揚げ sits at the centre of the difficulty. In the accounts that describe it, a young woman’s first sexual encounter was arranged with a patron for a sum that went toward her debt. How general it was, how it varied between districts and periods, and how much of the standard account comes from post-war popular writing rather than from the record are all genuinely contested among historians, and this page is not in a position to settle any of it. What is not contested is that the 芸/色 boundary, whatever it held at the level of the trade, did not reliably protect an individual girl inside a house that held her debt.

Which does not weaken the argument of this page. It sharpens it into the only form worth having. The distinction between paid art and paid sex was a real institutional fact, enforced by registries and trade interest — and it protected the category rather than the person, because that is what institutional enforcement does. A boundary can be entirely real at the level of a trade and entirely unavailable to a woman who is inside it and in debt. Both of those are true of 芸 and 色, and the second is the half that a modern reader has the most direct use for, because it names what a boundary has to be attached to before it protects anybody: not a tradition, and not a reputation, but her own ability to leave.

The shape of the law is worth one section on its own, because almost everything visible in this market today is a consequence of its omissions rather than of its prohibitions. The 売春防止法 prohibits 売春, and it reaches solicitation, procurement, the provision of premises and the business of managing it. What it does not attach is a criminal penalty to the two people in the transaction itself.

A prohibition constructed that way does not remove a market. It determines the form the market takes. Everything must be sold as something else, and the something else has to be specific enough to be described in an advertisement and defensible enough to survive a licensing inspection — so the industry that grew up under this law is organised, with great precision, around naming exactly what is on offer and excluding one act by name. The 風営法 of 1948 supplies the categories that make those descriptions official.

Which produces an irony worth sitting with rather than scoring a point off. A law written to suppress a trade has made this one of the few countries where paid intimacy is routinely advertised with an explicit, written statement of what is and is not included — not out of candour, but because the licensing regime requires the description and the criminal law requires the exclusion. The honesty is a by-product of regulation, and by-products are worth exactly as much as deliberate virtues when what you need is a description you can read in advance.

And the limit belongs here rather than in a footnote. None of that makes any individual business honest, several of the categories are widely understood to be observed loosely, and a written description is a description rather than a guarantee. What it establishes is narrower and still useful: in this country, asking for the scope in writing before booking is a completely ordinary request, made constantly, that no legitimate business finds strange.

The most useful question this history answers is not whether the 芸/色 line was real. It is what was holding it, because the answer changes by period and the changes are instructive.

In the licensed period the line was held by registration, by trade organisation, by the commercial interest of houses whose product was the distinction, and by the fact that crossing it was visible to a whole district. Not by anyone’s virtue. In the modern licensed trades it is held by category, inspection and the criminal exclusion described above — again structural, again not virtue. And in the unlicensed wellness and relaxation trade, which is where a great deal of body-centred work now happens, it is held by nothing at all except what an individual practitioner decided privately and what an individual customer will tolerate.

Which is this library’s standing finding about safety, arriving here as economic history rather than as advice. Where holding a line is structural, it holds on ordinary days and with ordinary people. Where it depends on somebody being principled in a room with the door shut, it holds exactly as often as people are principled, which is often and not always — and the person who discovers the exception discovers it alone.

So the checkable question for a reader is not which tradition a place invokes. It is which of those three things is holding its line — a licence and a category, a written scope she has in her hand before she arrives, or an assurance from somebody she has not met. Those are different amounts of protection and they are distinguishable before booking, which is the only point at which the distinction is any use.

The useful conclusion is not about lineage. It is that the most common objection to a service like this one — that a boundary of this kind is obviously a fiction, since everybody knows what is really being bought — is an argument that has already been run, at scale, for centuries, in a country that built registries to enforce the answer.

That does not settle whether any particular house is honest. Nothing settles that from the outside, and history confers no character reference. What it does establish is that the category is a real category rather than an evasion invented last year, and that a reader here does not need it explained from first principles. She has grown up in a country where the distinction has an address.

Moonlight is not a 花街 and claims no descent from one. It has no 検番, no district, no lineage, no inherited repertoire, and it would be dishonest to borrow the aesthetic of institutions whose labour arrangements it would refuse. The history above is reported, not claimed.

And the section above put a question to the reader that this house has to answer about itself, or the question was rhetorical and worth nothing. Of the three things that have held this line in three periods — a licence and a category, a written scope in her hand before she arrives, or an assurance from somebody she has not met — the first does not apply to us. We are not a licensed adult business and do not present as one.

So what is on offer here is the second, and the second alone, and a reader is entitled to know that it is the weaker of the two real ones. A written scope agreed in advance binds nobody but us, is enforced by nothing but our wanting to remain a business people return to, and would not survive contact with somebody who had decided to disregard it. That is the true shape of it. Anybody who tells you otherwise about an unlicensed practice of any kind — ours included — is describing an intention and calling it a protection.

What it does share with that history is one structural feature and it is the one this chamber keeps returning to: the scope is settled before the evening, in writing, and it is settled by the person receiving rather than by the house. The 検番 held the line from outside because the quarters wanted it held. Here there is no registry, so the written scope does that job — and it is written by the one person whose interest is unambiguous.

So the answer to the opening question is that the distinction is real, that Japan has held it longer and more formally than almost anywhere, and that holding it has never depended on anybody being naive about what else was for sale next door. It was a trade rule with a registry behind it. Today it is a written scope with a person behind it, which is a smaller instrument and, for once, one that belongs to the right side of the table.

And the second answer, which the history only gives up once the romance is off it: the rule protected the category, not the woman. A trade can hold a line perfectly while a girl inside it and in debt has no access to it at all. So the useful inheritance here is not the boundary. It is the knowledge of what a boundary has to be fastened to before it does anybody any good, and the short version fits in a sentence — not a tradition, not a reputation, and not somebody’s good character, but whether the person it is supposed to protect can leave at any point without paying for it.

Which is the one thing a reader here can check before anything else is decided, about us or about anybody. It needs no history at all. The history is only how this page knows the question is worth asking, and that four hundred years of people asking it did not make it unnecessary to ask again.