When a Husband Would Not Grant a Divorce – Help for Women in the Time of the Tokugawa Shogunate

What did divorce look like in Japan during the Edo period, what could a woman do when her husband refused to grant her freedom, and why were there temples to which women fled for refuge? The history of kakekomi-dera is a story of law, violence, custom, and the last doors left open to women within the harsh order of the Tokugawa shogunate.

“The Temple One Runs To”

In the Japan of the Edo period, a divorce could fit on a scrap of paper. That document even had a name: mikudari-han (三行半 – literally, “three and a half lines”), which formally severed the marital bond. And yet between that paper and the real life of a woman there stretched an abyss. For what does a document mean in a world where a person belongs not only to herself, but to the house, to the family name, to the arrangement between families, to the network of neighbors, obligations, reputation, and the expectations of others? What could a woman do if she was denied that paper, and her suffering (perhaps her husband was a drunkard, perhaps he beat her, perhaps he abused her?) mattered less to everyone around her than order itself?

That is precisely why the history of kakekomi-dera moves us so deeply. At the very heart of a society that enclosed a woman within the structure of the house/lineage — ie (家) — there existed special temples to which one could run for rescue: Tōkeiji, Mantokuji, places sometimes called “divorce temples,” and at other times simply “temples one runs into” (駆込寺). They were not a romantic gateway to freedom. Behind their walls awaited interrogations, seals, fees, months of discipline, cut hair, negotiations between families, and a patient struggle with a system that did not wish to yield. And yet for many women they were the last door — narrow, austere, but real. A place where a husband could not touch them, and where legally competent persons might help them perhaps compel a husband to issue consent to divorce, which in the future would make remarriage possible.

That is why this story is more than a mere curiosity from the history of old Japan. It is a question about the true measure of civilization: not how sternly it can defend order, but whether it can recognize that there is a limit beyond which continuing to demand obedience ceases to be a virtue and becomes simple cruelty. In this article, then, we will look not only at the kakekomi-dera themselves, but at the whole world that made them necessary: what marriage meant in the Edo period, what divorce looked like, and what a woman could do when her husband refused to grant permission or turned the home into a place of fear; how the temple procedure worked step by step, and what those two years of suspension looked like for a woman — between her former life and her new fate. For sometimes the maturity of a society is determined not by how beautifully it speaks of family, honor, and duty, but by whether it can admit that there are moments when clinging further to principle ceases to be virtue and becomes cruelty toward the weaker person — and that it is precisely then that even a narrow little gate of rescue must be left open within that order.

If only one could just make it to the gate…

At dawn the road was still dark blue with cold, and the mud at the edge of the track still held the dampness of the night when Oharu at last saw the dark outline of the palisade. She no longer had the strength to run; her straw zori had softened from the dew and had long since ceased to protect her feet. For several ri she had walked in the conviction that at any moment she would hear someone’s shout: the voice of a servant from her husband’s house, who would catch up with her, seize her sleeve, and drag her back to the ie, to that house where everything had a name — duty, obedience, shame — only her pain had none. And yet — she had made it. After a long, exhausting journey, she had reached the temple. Impossible, and yet… could it really succeed?

By the gate of Mantokuji, the famous divorce temple of the Edo period — an enkiri-dera (縁切寺) — there grew a tall keyaki, a Japanese zelkova, an old tree of which it was said that it had seen more women’s tears than many a bodhisattva. The main gate, called the kakekomimon (駆込門 – “escape gate / the gate one runs through”), was still closed, because a ceremonial day was approaching. For women such as her, however, there was a side entrance. It was narrow and severe. It was not for guests, but for those who came here at the threshold moment — between marriage and flight, between a former family name and an uncertain tomorrow.

At Mantokuji, the mere fact of getting behind the gate carried greater weight than many words, and later tradition even repeated that if a woman no longer managed to run across the threshold herself, it was enough to throw a personal object over it — even just a zori sandal — to signal a cry for protection; women were not received automatically, however, but only after the case had been examined by temple officials.

For a moment she stood motionless, with her hand clenched on the end of her sleeve, as though afraid that if she touched the wood of the gate, everything would turn out to be a dream. In Edo, divorce could be simpler on paper than in life: a husband could write out a divorce letter, brief as a blow, those famous three and a half lines — mikudari-han — and send the woman back to her natal home with the formula that she was free to marry again. But if he did not wish to do so, if he preferred to keep his wife as unpaid hands for work, as a vessel for his anger, as part of the household inventory, then the law itself ceased to be a path and became a wall. In such matters a woman could not simply “leave” as a merchant set aside a defective abacus; she needed a document, consent, intermediaries, seals, and above all an institution whose authority was greater than the obstinacy of one man. Divorce temples such as Tōkeiji and Mantokuji were exactly such a fissure in the Tokugawa order — recognized by the authorities, yet limited chiefly to commoners; matters concerning kuge and buke (about the noble-born classes, you can read here: The Birth of Samurai Japan: buke — earth and steel, kuge — poetry and ritual. Who won?) belonged to the state, not to the gates of a monastery.

A man came out from within in plain servant’s dress, not a monk, but one of those lay temple men who dealt with what lay between the world and the monastery: letters, interrogations, escort, accounts, complaints. He looked first at her face, then at her hands. In Edo Japan, hands sometimes spoke more than language — whether a woman worked in the fields, at vats of indigo, at a market stall, at the loom, or whether she came from a house where fingers knew only a powder box and a needle. Oharu was the daughter of the owner of a small oil warehouse in Edo, married off three years earlier to a man from Honjō, taken into the family as a mukoyōshi (adopted son-in-law).

It was meant to be sensible: two houses, one family name, a man of business. Instead, the husband drank, lost money, beat her — not in such a way as to kill her, but in such a way that the next day she could still carry water and bear shame in silence. These were precisely the kinds of reasons most often later written down in petitions: laziness toward work, drunkenness, licentiousness, gambling, violence. But the temple official also knew that not every case was what it seemed on paper; sometimes behind the charge of “wickedness” lay the family’s dislike of a son-in-law, sometimes the desire for another marriage, sometimes the simple war of two houses. That is why at Mantokuji tears at the gate were not enough — there had to be an investigation, a mimotochō (身元帳 – a register of identity and the circumstances of the case, the case file).

“Who stands behind you?” he asked quietly.

The question was cruel and necessary at once. In the Edo period, a woman rarely came truly alone, even if she had made the journey by herself. Behind her there had to be people willing to confirm her story and bear the weight of the case: a father, a brother, an uncle, sometimes a mother, but the signature usually belonged to a man from the natal household. If the temple took a woman under protection, the matter did not end at the moment the wicket gate slammed shut. One had to notify the husband, sometimes his parents, the village head or district headman, the superior of the goningumi (more about the goningumi here: Under the Watchful Eye of the Neighbor: Gonin gumi and Collective Responsibility in the Time of the Shogunate), for the house was not a solitary island, but part of a network of responsibility. The five-household neighborhood groups, the goningumi, watched over taxes, order, and mutual supervision; a person in Edo lived not only under his own roof, but in the eyes of his neighbors.

Divorce, then, was not a private sigh between two people, but an administrative, moral, and social event. At Mantokuji, once the runaway woman had been received, officials sent the husband a notice that the temple had accepted her request for asylum and recommended that he voluntarily issue a divorce letter; if the husband refused to accept the letter or to appear, the matter could be escalated to the officials of the Magistrate of Temples and Shrines.

Oharu bowed her head and whispered the name of her elder brother. He had sold two bales of paper and pawned part of his goods to pay for her journey and the first expenses, though neither of them yet knew how much salvation would ultimately cost. For rescue in Edo Japan had a price. It was not an act of pity, but part of a functioning machine: one had to pay for rice, wood, fuel, service, officials’ time. At Mantokuji, the fees for a woman’s upkeep were not usually crushing, and yet for a merchant or farming family they still constituted a burden; they were paid in installments, depending on the price of rice and the progress of the case. That is why kakekomi-dera were not in practice equally open doors for every woman. In the imagination, the gate stood for all, but in reality it was crossed more often by those whose families had enough means, contacts, and patience to make their way through Edo’s paper jungle.

She was led inside the kami no kata — the enclosed women’s part of the monastery — behind yet another fence, deeper in, where the nuns lived in an order separated from the world. She saw stone lanterns, a lotus pond, wooden footbridges dark with dampness; it was not paradise, but rather a place where beauty and discipline wore the same face. One of the women in gray robes gave her water to wash her hands. Another looked at her hair with such a gaze as though she already saw where to place the scissors.

During her stay in Mantokuji, a woman accepted into a divorce case had to live almost like a nun: her hair was shortened, leaving only a small knot at the back, her conduct was observed, and the slightest insubordination could end her stay without a divorce. Asylum did not mean freedom. It was rather a severe suspension between former dependence and a new status that still had to be won. At Tōkeiji, if the case entered the path of divorce under temple law, a woman usually had to serve a full twenty-four months; Mantokuji could be more flexible and from the beginning of the nineteenth century also allowed earlier departure if, during the stay, it proved possible to reach a private settlement and obtain a divorce letter from the husband.

Before noon she was told to tell everything from the beginning. Not only about the blows. Also about money, about witnesses, about who had arranged the marriage, whether there were debts, whether the husband was truly the son of that house or an adopted son-in-law, whether gifts had been exchanged after the wedding, whether wedding sake had been shared, whether something had been sworn before the local tutelary deity. In Edo, marriage was not a single act, but a series of socially recognizable gestures; likewise, divorce was not a single rupture, but the gradual untying of bonds that joined not only two people, but two houses, two neighborhood groups, sometimes an entire little world.

If the matter could end in a settlement, the temple first pressed for a settlement. Reconciliation was preferred if it seemed possible; only when parents, intermediaries, and admonitions failed did the monastery move on to harsher measures. At Mantokuji, both sides were summoned, heard separately, lodged in separate quarters so as to avoid quarrels, and every stage was written down. At Tōkeiji, when the husband was uncooperative, it was necessary to obtain not only an ordinary rienjō (離縁状, divorce letter), but also a formal document acknowledging a temple divorce, bearing the seals of the husband, the nanushi (community headman), and the leader of the goningumi.

Toward evening Oharu sat by the paper window. Beyond the shōji, the garden murmured softly, as if the wind were moving over the leaves in white tabi. She thought of the mikudari-han (a colloquial name for the divorce letter, 三行半 – literally, “three and a half lines”), of that little piece of paper on which a woman’s future life depended. For a man it could be no more than a form. For her — something like a new breath. Without it, the next marriage could become a source of accusations, disputes, and even shame for the whole family.

She also took comfort in the thought that here, behind the high palisade, even the husband’s letter was no longer the only instrument. If, after the appointed time had passed, he still refused to cooperate, temple officials could refer the matter to the jishabugyō (寺社奉行 – the shogunate magistracy for temples and shrines), and obstinacy could be broken with the threat of imprisonment. At Mantokuji there also existed a form of divorce called okoegakari, issued with the authority of the office, especially when a husband tried by force to take his wife from the temple or ostentatiously disregarded the procedure. Even then, however, the final order had to be maintained: documents were sent to the families, to local superiors, to the structures of authority. Edo liked to have every suffering entered into a register.

Only at night, when the sound of wooden sandals had fallen silent in the corridors, did she allow herself to weep. Not because she was already safe. Safety in such places was never either obvious or immediate. A husband could try to appeal, a family could withdraw its support, the temple could judge her behavior improper, a settlement could collapse in half a sentence. And besides — two years could be as long as an entire lifetime.

Rather, she wept because for the first time in months someone had received her. Someone had not said: suffer, for such is a wife’s duty. Someone had not first asked whether it would not be better for everyone to remain silent. In the hard order of the Tokugawas, where the house was more important than the individual and reputation more precious than tears, these peculiar gates nevertheless existed. Not in order to destroy the world, but in order somehow to help one cooperate with it in such a way that the individual might be saved. Kakekomi-dera did not give a woman freedom in the sense in which we understand it in the twenty-first century. But what they gave — was sometimes the only rescue for a woman whom no one else would help.

What does “kakekomi-dera” actually mean?

The word kakekomi-dera (駆込寺) itself literally means “a temple one runs into” — a place into which someone rushes at a moment of danger, seeking shelter. Very close to it is kakeiri-dera (駈入寺), which emphasizes the same movement: running in, throwing oneself inside, entering under protection. Enkiri-dera (縁切寺), by contrast, shifts the center of gravity: it speaks no longer primarily of the act of flight itself, but of a “temple that severs a bond,” that is, an institution capable of bringing about the legal dissolution of a marriage. Some scholars even propose that the term enkiri-dera should be used for those temples that truly helped obtain a divorce, while kakekomi-dera / kakeiri-dera should be reserved rather for their asylum function; in practice, however, Japanese historians often use these names interchangeably.

It is also worth distinguishing kakekomi from ajiru. They are not synonyms. In Japanese historiography, kakekomu / kakeiru refers rather to the flight of a persecuted person seeking protection, whereas ajiru is connected with the old law of asylum also for people accused of crimes or belonging to groups regarded as dangerous. The place of refuge itself was described by the concept of funyū (不入) — literally “non-entry,” that is, an area into which outside authority could not simply enter. This is a very important nuance: Japanese asylum was based not solely on the “sacredness of the place,” but also on its politically recognized separateness and limited extraterritoriality. That is why kakekomi-dera were something more than a compassionate gate — they were an institutional fissure in the order of power.

Marriage in Edo

To understand what kakekomi-dera really were, one must first set aside the modern notion of marriage as an exclusively private relationship between a woman and a man. In the Edo period, marriage was above all a matter of the house — that is, the ie (家) — but also of interest, succession, labor, prestige, and local order. The question was not merely whether two people “suited one another,” but whether the arrangement between families would work: who would take over the family name, who would inherit the workshop, the field, the shop, who would care for aging parents, who would produce an heir, who would strengthen the position of the house. Even among townspeople, marriage and divorce were perceived as an alliance between families, not merely a matter of the heart.

One must be very careful, however, not to repeat too simple an image: “Edo = uniform patriarchy, woman = powerless victim.” Such an image dominated for a long time, but the historical material itself rather shows that there were nuances here. The model of the rigid ie, subordinated to the man and based solely on an unchanging hierarchy of age, sex, and primogeniture, describes relatively well the ideal of the warrior stratum, not the whole society of the Tokugawas. That pattern was not equally standard among kuge, chōnin, and hyakushō; it was only the Meiji state that imposed the ie model more broadly as a general social norm.

It looked different in a samurai family, different in the house of a wealthy merchant, different in a rural farmstead, and different again in regions where women had a real share in production and cash circulation. In the world of buke (warriors, the privileged class), marriage was more openly political. It required the consent of superiors, had the dimension of an alliance between lineages, and divorce likewise was subject to registration and the control of authorities. Official moral teaching demanded from the samurai wife an obedience almost sacral in nature: “the husband as heaven,” quietness, gentleness, purity, self-control (a day in the life of a samurai wife – you can read here: The Invisible Woman – Life as a Samurai’s Wife).

But even here practice was not as lifeless as textbook slogans suggest. Samurai daughters were taught not only etiquette, calligraphy, the playing of the koto, or the tea ceremony, but also the use of the naginata and what tradition describes as “the way of brush and sword.” Divorces existed in this stratum, remarriages as well, and the position of a woman’s natal house could genuinely protect her from the worst.

Among townspeople and wealthier peasants, the picture was even more fluid. Yes, marriages were arranged, but more often than in the samurai world this took place through miai — meetings of prospective spouses under the supervision of families and intermediaries, rather than through the complete disregard of the future spouses’ own wishes. That fact alone suggests that the female position was not zero. Merchants’ and farmers’ daughters brought into the house not only a body and a reputation, but also labor, money, a dowry, skills, and sometimes education. In sources from the period, one clearly sees that the cash dowries of the wives of chōnin (townspeople) and gōnō (wealthy peasants) remained their own property, constituting a certain form of security also in the event of divorce. This is a very important detail: it shows that a woman was not everywhere and always entirely “defenseless before the system.”

Equally important was the economy of everyday life. In many regions of late Edo, women were not merely part of the “interior of the house,” but an important part of the local economy. They spun, wove, worked with silk, cotton, seasonal labor, auxiliary trade. For example, in Jōshū Province, women’s labor was so significant that later the region gained the name kakaa tenka (嬶天下) — “the land where wives rule.” In the middle of the nineteenth century, in some villages almost all women were involved to some extent in cotton weaving. Such a world produced a different psychology of marriage: a daughter was not merely a burden to be married off, but also a real economic asset like a son, and her family more often had both the interest and the strength to support her in a conflict with her husband.

This can, moreover, be clearly seen in the data concerning rural marriages. In some studied villages, the age at which women married depended on the family’s income: daughters from richer households married earlier, those from poorer ones later, because their labor was more needed at home. Women from lower-income families also more often entered less stable marriages; in research from Mino, as many as 60% of marriages in the poorest income group broke up before twenty years had passed. This is an important corrective to moralizing generalizations: the durability of a marriage depended not only on “virtues,” but also on economics, inheritance, the local labor market, and the structure of the household.

Even the very moment when a union became “official” was not identical everywhere. A marriage had to be reported to the local authorities of both sides, but practice differed depending on the place. Sometimes it was considered that formalization took place upon the move of the wife or adopted son-in-law; sometimes registration was delayed. Sources also remind us that not all unions entered the documents: short-term, unregistered, childless relationships could simply fall out of the records. Again, this shows that Edo society was more fluid than the image of an all-powerful, iron norm would suggest.

The same was true of divorce. Official ethics liked to speak in a hard voice. In the background there still echoed old formulas about the “seven grounds for dismissing a wife,” inherited from earlier legal orders: barrenness, adultery, jealousy, disobedience toward in-laws, talkativeness, theft, “evil illness.” But alongside them there also existed “three circumstances” in which a wife ought not to be dismissed — for example, when she had nowhere to return to, or when the husband had become wealthy only during the marriage. Between moral norm and everyday life there therefore stretched a wide space of negotiation, compromise, and practice. The divorce procedures of commoners, moreover, were frequent and relatively ordinary; the scholar cited in your materials shows that in some villages divorce did not close the way to another marriage for a woman, and many of them returned to the marriage market quite quickly.

It is also important that marriage in Edo never existed in a social vacuum. The house was embedded in a network of neighbors, village or urban superiors, population registers, parish temples, and responsibility groups. The community tracked the movements of people: who had moved, who had entered a union, who had left, who had returned to the parental home. Therefore a marital crisis was not only a drama enclosed behind a paper wall. It was a matter that set documents, relatives, intermediaries, sometimes local officials, and in special cases also temples such as Tōkeiji and Mantokuji in motion. Kakekomi-dera, then, did not appear in a vacuum. They were a response to a world in which marriage was a dense social knot — and precisely for that reason its dissolution required not only emotion, but also form, procedure, and authority.

Perhaps the most interesting thing in all this is that the Edo period was at once harsher and more flexible than we usually suppose. Harsher — because the life of the individual was more deeply inscribed into the obligations of house and community. But also more flexible — because beneath that layer of moral formulas there operated negotiations, compromises, economic calculations, family pressure, and practical solutions. A woman was not equal to a man in the modern sense. Yet neither was she everywhere and always merely a silent shadow. And it was precisely from that fissure between ideal and practice that the temples to which one could “run” grew.

What did an “ordinary” divorce without a temple look like?

Outside Tōkeiji and Mantokuji, most divorces in the Edo period did not unfold behind a monastic gate, but at a low table, over a sheet of paper, in the presence of relatives, neighbors, sometimes the local headman. The best-known form of such a divorce was the mikudari-han (三行半 / 三くだり半), the famous divorce letter “of three and a half lines.”

It was written by the husband — sometimes in his own hand, and sometimes through someone else, if he was illiterate, old, or awkward. It happened that the content was written down by a relative, a neighbor, or even a stranger asked as a favor, and the husband merely placed a mark, a signature, or the impression of a fingernail. The document itself was short, but its significance was immense: it stated that the marital bond had been dissolved, and most importantly — that if the woman later married another man, the former husband “would not say a single word of objection.”

It is precisely this second element that is crucial. A mikudari-han was not merely a slip of paper confirming a separation, but a document securing a woman’s future. Without it, a new marriage could be challenged. The shogunate did not want private conflicts to turn into public disorder, and therefore insisted on the proper documentation of divorces. In the middle of the eighteenth century it was explicitly stressed that a man who had not given his wife a divorce letter and then took another woman should not continue to live in his locality. Likewise, a woman who entered a new union without such a document exposed herself and her family to very unpleasant consequences. One sees clearly, then, that in the world of Edo divorce was not “only an ending,” but also a formal arrangement of the future.

The very text of the mikudari-han could be astonishingly concise and formulaic. There often appeared the formula that the wife “does not suit the house” (ie ni awazu, “does not fit the household custom,” “does not agree with the atmosphere of the house”), or that further life together was impossible. It sounded cold, almost bureaucratic, but that dryness itself was part of custom. In Edo Japan, many of the most important things were said indirectly. Instead of dwelling on guilt, character, or wrong, one wrote down a brief reason, and then passed to the point: from now on the woman may go her own way. At times the document still echoed older ways of thinking about fate and karmic bonds — as if the relationship had not so much “gone bad” as simply exhausted the time allotted to it by destiny.

Very interesting is also the phrase mikudari-han itself: “three and a half lines.” It sounds almost playful, but historians have long disputed where it actually came from. One traditional explanation says that a marriage document once had seven lines, and so a divorce — as it were “cutting the bond in half” — received half that length. The second is more urban and customary: in the world of courtesans there were said to exist short three-and-a-half-line invitations for clients, and over time that format came to be associated more broadly with writings concerning relations between men and women, including their ending. The third explanation is the darkest: according to a Hokuriku custom, three and a half ladles of hot water were used to wash the dead, and because divorce was sometimes symbolically treated as the “death of the union,” such a length was regarded as fitting for a divorce letter as well. We do not know which of these theories is true, but all three are very “Edo” — they join law with ritual, daily life with symbol, and paper with custom.

Importantly, the mikudari-han did not always look ideal and was not always written according to pattern. Earlier, before this format became widespread throughout the country, five-line documents were more common in the Kantō region; only over time did the custom from the Kyoto area spread more broadly. Even later, practice remained flexible. There were longer divorce letters, atypical documents, cases of substitute signatures, and even local sayings suggesting that a person who could not write might sketch a substitute sign, so long as the witnesses understood what it meant. This too says a great deal about Edo culture: form mattered, but more important still was that the community recognized the divorce as accomplished and unlikely to give rise to later dispute.

Put most simply, an “ordinary” divorce in the Edo period was at once simpler and more social than it might seem today. Simpler — because often a short document was enough. More social — because behind that short sheet stood an entire world of witnesses, relatives, local customs, and the memory of the community. The mikudari-han was small, but it carried something important within it: formal permission for a woman not to remain suspended between one house and another. In a society that rested so much on paper, reputation, and the consent of one’s surroundings, those three and a half lines could mean more than a long lament.

Why could a woman not simply “leave”?

In the Edo period, leaving was not a movement in space, but a movement in papers. It was enough to take a few steps beyond the threshold of the house to understand that it was not the woman’s body that was “imprisoned” — but her status. Formally, a marriage did not end at the moment when a wife packed her kosode and left. It ended only when a divorce document appeared: a rienjō (離縁状), usually in the form of the mikudari-han described above. And here lies the essence: in the Tokugawa order, it was the man who was the party issuing that paper, even if he did so at the wife’s request. A woman, if she wished to leave “legally,” had to bring about a situation in which someone — above all her husband — confirmed her freedom in writing.

Why was this so important? Because without paper a woman did not become “free,” but “suspect.” A new marriage could be challenged, and the conflict of two families could very easily turn into a dispute that drew in village headmen, neighborhood responsibility groups, and officials. In theory the bakufu was not interested in the intimacy of commoners, but it was very interested in ensuring that private relations did not explode publicly. Therefore divorce had to be documented — precisely so that there would be no later wrangling over whether “she is still someone’s wife.”

Psychologically, this meant a situation in which a woman might already be at the limit of endurance, might be beaten, humiliated, might simply no longer wish to live in another person’s anger — and yet her leaving without her husband’s consent looked to the world like a violation of order. And not only of moral order, but of practical order: the house was an administrative, economic, and social unit. The woman was part of that system: she worked, gave birth, cared for the elderly, formed a network of obligations. If she disappeared, someone had to explain why. And if no one could or wished to do so, the structure of the house was disturbed.

This did not mean, however, that a woman was powerless. Her agency rarely took the form of “lawsuit, court, divorce” — more often it took the form of strategy and pressure. The sources say this plainly: women sometimes “divorced” simply by fleeing, most often to the parental home. This was a very old, practical path: flight as a fait accompli that forced families into negotiation. In many cases this had nothing to do with romantic rebellion, but was a hard social game: the woman’s father or brother laid down conditions, the husband calculated shame, costs, and the risk of conflict, and local intermediaries tried to extinguish the dispute before it turned into a “case.”

If that did not work, intermediate institutions appeared. And here one sees something significant: even in the hierarchical society of Edo, there existed a social recognition that a woman had rights (unfortunately very unequal in relation to men) — if only the right to protection from brutality and the right not to remain until the end of her life in a state of legal ambiguity. Women could not “directly sue for divorce,” but they could kakekomu — “run for help” to a place that the state sanctioned as a legal broker of divorce: Mantokuji or Tōkeiji. This is very telling: the system did not give women equal tools, but it did leave them an emergency one.

In practice, however, that tool was not for everyone. The myth says: “an open gate for every unhappy wife.” The documents say something else: asylum and help were not automatic, they depended on the decision of temple officials, and above all they required backing — a family capable of bearing the costs, supervising the formalities, and exposing itself to open conflict. In the case of Mantokuji, the procedure included obligatory negotiations between families, full documentation of successive stages, and formal obligations on the part of the woman’s family to pay the fees and to ensure that the runaway woman would observe the temple rules. That meant signatures, seals, witnesses, sometimes the attaching of data concerning temple affiliation within the danka registration system — that is, the whole weight of Edo’s bureaucratic reality.

To this was added economics. Mantokuji charged maintenance fees that could be “not especially burdensome, but not negligible,” and which depended on the current price of rice (why? — about the financial system of Edo we write here: A Country Without Banks – The Monetary System of the Shoguns and Credit Measured in Honor and Shame in Edo Japan); families often paid in installments and — significantly — there was no question of refunds for unused time. Tōkeiji could go further: besides monthly fees, it took large “donations” at entry, and women were divided into ranks on which both their status and the kind of work they performed in the temple depended. In other words: even in asylum there was hierarchy. If the family paid more, the woman had lighter duties; if less, she was assigned heavier, dirtier work.

And it is precisely here that one should say something that sounds like a sociological slap in the face: in women’s dealings with these institutions, class position could matter less than wealth. Mantokuji offered a “window of possibility” to women who could afford its services — regardless of class origin, so long as they had the means. In other words, poverty could close the gate faster than patriarchy.

Therefore the answer to the question “why could she not simply leave?” is this: because leaving was not a private decision (in the eyes of eighteenth-century Japan), but a change of status in a world that lived by form, confirmations, and social consent. And when a woman tried to bypass that world, the world answered her not so much with moral condemnation as with the threat of suspension — without paper, without the right to a peaceful new life, without the possibility of a clean start. Her agency did not disappear. It changed shape: flight, family pressure, the search for intermediaries, and finally the run to the gate where another logic began — the logic of asylum, negotiation, and the document that had the power to cut the bond.

How did the divorce procedure through a temple really work?

It would be easiest to imagine kakekomi-dera as places of miracle: a woman runs through the gate, the monks close the doors, and within a moment the marriage ceases to exist. In reality, it was exactly the opposite. The temple did not “conjure divorce,” but set in motion a long, formalized, and deeply social process in which asylum, mediation, document, custom, and the apparatus of the state met one another. That is precisely why kakekomi-dera were so extraordinary: they did not stand outside the Edo system, but functioned in its fissure — as institutions intermediate, negotiatory, and partly legal.

The first step was, of course, the flight itself. At Mantokuji, a woman initiated the matter by running to the temple and crossing its main gate; if she arrived on a day of high ritual, when the main gate was closed, she was admitted through a side entrance. Once received, she was placed in the enclosed part of the monastery, in the Kami no kata, the deepest female zone of the temple. At Tōkeiji, the matter was from the outset more bureaucratic: upon arrival, the woman had to submit a document of arrival describing her identity and the reasons for her flight, and temple officials initiated a preliminary verification of her story. If discrepancies appeared, criminal accusations, or doubts as to origin, the petition could be rejected. This first stage alone already shows something important: the gate was open not to “despair” as such, but to despair that could be entered into a recognizable procedure.

Then came the notification of the family and the husband. At Mantokuji, once a woman had been received, temple officials sent the husband a formal notice informing him that the temple had recognized her intention to seek divorce through official channels and expressly suggesting that he voluntarily issue a divorce letter. This letter also went to the local headman on the husband’s side. At Tōkeiji, by contrast, the woman’s parents were first summoned, and only afterward was the husband’s family brought into the matter; from the very beginning, therefore, it was clear that this was not a conflict of “her versus him,” but a dispute between two houses, supervised by the temple and visible to local authorities.

Only then did negotiations between the families begin. This is a very important moment, because it helps one understand that kakekomi-dera were not above all “divorce machines,” but rather places that compelled conversation. At Mantokuji, the woman’s parents were summoned together with local superiors; if they appeared, the temple official questioned them and first encouraged reconciliation of the spouses. The woman met her parents at a special consultation place beside the temple office. If she still persisted in the decision to divorce, the family was to begin private negotiations with the husband’s family. At Tōkeiji it was similar: there too private agreement was the preferred solution, and the formal temple path truly began only when settlement failed. This is a very Edo-like feature: the institution does not so much “solve the problem” as first try to restore social order at the lowest possible cost.

If the parties reached agreement, the matter could end in a private divorce or even reconciliation. At Mantokuji this meant drawing up the appropriate confirmations, sometimes also the payment of compensation to the husband’s side by the woman’s family — which is extraordinarily interesting, because it overturns our intuition. It was not always the husband who “dismissed” the wife and gave her something; sometimes the woman’s family also paid the husband’s side in order to close the matter quickly and safely. At Tōkeiji, private agreement could take the form of a private divorce letter, a letter of the parents’ taking their daughter back, or even a document confirming the resumption of marriage. The temple was therefore not only an asylum, but something like an archive and a stage of social agreement, on which a dispute could be rewritten into an acceptable language of documents.

If, however, no settlement was reached, everything became much more formal. At Mantokuji, the woman’s family had to submit a kakae-shōmon, that is, a contract of entering under the temple’s protection. This document confirmed the legality of the steps taken thus far, gave the woman’s temple affiliation — which was connected with the system of obligatory temple registration, the danka seido — and bound the family to bear the costs of her maintenance. At Tōkeiji, the parents signed a document guaranteeing that their daughter would observe the monastery rules, while from the husband there was required not only the ordinary rienjō, but also a separate, much more formal “temple divorce contract,” addressed to Tōkeiji itself and bearing his seal, the village headman’s seal, and that of the leader of the goningumi. Again, this is a very important detail: a temple divorce was not a private agreement between two persons, but a document entered into the local mechanism of control.

Then began the period of residence in the temple. The woman received shelter, but not freedom. At Mantokuji she lived like a half-nun: her hair was cut short, leaving only a small knot at the back, she took part in the daily religious rhythm, the recitation of sutras and meditation, and performed light cleaning and gardening work within the women’s part of the temple. Her conduct was observed, and irreproachable behavior had procedural importance: if she was judged improper, the stay could be interrupted, and the divorce would not come into effect. In the eighteenth century, the very cutting of the hair had an additional symbolic meaning — it signified a temporary withdrawal from the secular world and made the woman “untouchable” in a customary sense. Salvation thus took the form not of liberation, but of entering a stricter discipline.

The length of that stay depended on place and time. At Mantokuji, formerly it could be thirty-six months; later the standard became twenty-five months; for fiancées and concubines there was even a “half term” — twelve and a half months. From the beginning of the nineteenth century, if during the stay it proved possible to secure a private agreement and the husband issued a divorce letter, the woman could leave earlier. Tōkeiji was more rigid in this respect: if the case had already entered the path of formal temple divorce, the woman usually had to complete the full twenty-four months of service, even if the husband eventually yielded. In the nineteenth century, most Tōkeiji cases, moreover, ended quickly in private agreement — on average after eleven days — and only a few women proceeded to the full two-year temple mode. This shows that the very threat of a long procedure could already serve as a tool of pressure.

In the meantime, the temple continually pressed the husband to issue a rienjō. At Mantokuji, once the residence period was over, he was summoned to the temple and distinctly compelled to issue the document. If he cooperated, Mantokuji officials issued their own decree of temple divorce, and the woman returned to her family with a complete set of papers enabling remarriage. Copies of the documents were distributed to the appropriate parties: local officials, representatives of both goningumi, and the officials of the Magistrate of Temples and Shrines. One sees clearly here the logic of the Edo period: freedom becomes real only when it has been confirmed in documents, copied, and entered into the archives of the shogunate.

And what if the husband refused? Then the state apparatus entered the game, along with the possibility of okoegakari rien — a divorce enforced by the authority of government. At Mantokuji, this path was initiated relatively early: if the husband rejected the notice of his wife’s admission to asylum, his resistance was documented and the matter transferred to representatives of the jishabugyō. If, after the residence period ended, he still refused to issue the rienjō, he was threatened with imprisonment, and the divorce was proclaimed officially. Particularly interesting is the fact that if the husband tried by force to “retrieve” his wife from the temple against her will, Mantokuji could obtain an okoegakari rien almost immediately, and the woman would then leave as a free person without the necessity of further residence. At Tōkeiji the mechanism was similar, though more cumbersome: the temple submitted to the jishabugyō an account of the husband’s disobedience, and that office usually threatened the husband with imprisonment if he did not issue the required documents.

It should also be added that the temple’s protection itself was not absolute. There were cases in which husbands tried, through other officials, to obtain the right to remove their wives from asylum. There were also cases in which the woman’s parents tried to withdraw her from the proceedings or failed to appear when summoned by the temple; and then too official procedures could be set in motion. All this shows how dense the web of dependencies was: the woman did not enter the temple as a solitary heroine of a modern drama, but as the point where the interests of husband, father, family, local community, temple, and state crossed.

Perhaps the most poignant thing is precisely that asylum did not mean freedom, but passage into another order. At home the woman was bound by her husband’s will and the rhythm of the family; in the temple — by the rhythm of sutras, documents, and observation. She gained protection, but paid for it with time, obedience, and the necessity of blameless conduct. In this sense, kakekomi-dera say something very important about the Edo period: rescue was possible, but it rarely took the form of freely “stepping outside the system.” More often it was a controlled passage through another, stricter system, which only after many months, seals, and pressures allowed the woman to return to the world as a person legally free.

The Last Door

The longer one looks at the history of kakekomi-dera, the more clearly one sees that it is a story neither of the simple cruelty of old Japan, nor of its hidden humanitarianism. It is rather a story of a society that believed very strongly in order, hierarchy, duty, and the permanence of the house, and at the same time knew that even the most carefully built order can become an instrument of harm. The Edo period did not create a world of equality. It did not give a woman full freedom to decide about herself. It did not recognize marriage as a union of two autonomous individuals who may part at any moment (nor, in those times, did European nations, something we sometimes happen to forget). And yet it left within its structure something extraordinarily meaningful: a narrow, controlled, difficult-to-reach, but real fissure of rescue.

That is precisely where the philosophical paradox of kakekomi-dera lies. Asylum was not a rebellion against order. It was not a revolution. It was not a modern right of the individual against the community. It was part of the Edo order itself — its peculiar safety valve, a place in which the system admitted that not every marriage can be saved and that not every woman can be told to return home in the name of virtue, obedience, and appearances. This is at once very sober and very human. Tokugawa society did not first ask about self-fulfillment. It asked about order. But precisely because order mattered so much to it, it had to leave itself an instrument for extreme situations — for those moments when endurance no longer protects order, but produces only suffering, violence, chaos, and shame.

This recognition does not make the Edo period gentle. On the contrary — it allows us to see its severity more clearly. A woman who ran to a temple gate did not run toward freedom in our contemporary sense. She ran toward procedure, toward discipline, toward months of waiting, toward documents, witnesses, seals, and the decisions of others. She passed from one order into another. Dependence on her husband’s house was replaced by dependence on the monastery, the family, the office, and ritual. Rescue did not take the form of triumph, but rather of a heavy, uneven road toward recovering at least a minimum of agency. This is very important, because it protects us from cheap idealization. Kakekomi-dera were not a paradise for wronged women. They were rather a liminal institution that allowed survival to those for whom the ordinary road had been closed.

And yet that is precisely why they are so moving. Because they speak of something that transcends the history of Japan itself. Every civilization likes to tell stories about its virtues: about morality, harmony, family, duty, respect, order. But the real trial comes not when everything works, but when a weaker, dependent, entangled person is no longer able to bear the role imposed upon them. Then the true measure of a society reveals itself. Not in how beautifully it speaks of virtue, but in whether it can acknowledge a limit to suffering. Whether it leaves at least one place where one may say: no further.

That is why the image of a woman running to the temple gate stays in the memory. Not because it is dramatic. Precisely because there is nothing dramatic about it. There is exhaustion, fear, mud on sandals, shame, uncertainty, and the remnant of hope. There is a world that does not wish to fall apart, and a human being who no longer wishes to be crushed by its weight. Kakekomi-dera were not a symbol of equality. They were not a sign of modern emancipation. They were something more modest, but perhaps for that very reason more true: proof that even in a hard society there existed an awareness that not every marriage ought to be preserved at any cost.

And perhaps that is exactly their deepest meaning. Not the promise of absolute freedom, but the recognition that a human being must sometimes have a last door. Even if it is narrow. Even if it leads not to happiness, but only to salvation. Even if one must run to it with the last of one’s strength.

>> SEE ALSO SIMILAR ARTICLES:

“A Fox Has Settled on Her” – Kitsune-tsuki Possessions of Women in the Japanese Countryside

The Mountain Witch Yamanba – Feminine Wildness That Terrified the Patriarchal Men of Traditional Japan

Ama – Women of the Sea in Japan. They Dive into the Depths While Men Wait on the Shore

In the Beginning, Woman Was the Sun – The Story of Amaterasu

Discover the Story of Umeko Tsuda: Sent Alone to America at the Age of Five – Upon Her Return, She Changed the Fate of Japanese Women