Dependency, and the Marriage That Supplied the Thing

There is a Danish word that does three jobs at once, and a book took it for a title. As an adjective, gift means married; it began as the past participle of the verb to marry. As a noun, borrowed from German, gift means poison. And there is a third, now obsolete, inherited from Old Norse: a noun that meant a thing given, a present, which has largely been replaced in modern Danish and survives mainly inside compounds, in the words for a fee or a duty. One spelling. Married. Poison. A thing given.

In 1971 Tove Ditlevsen published a volume of memoir and called it Gift. It was the third and last of the sequence that English-language publishing now sells as The Copenhagen Trilogy, and in English it is titled Dependency, because there was no single English word available that would hold the joke. It is shelved, in almost every bookshop that stocks it and in most of the writing about it, as an addiction memoir.

This essay proposes a narrower reading. The book is about supply. Ditlevsen married four times, and the record is plain about the order and the dates; the third of those husbands was a physician, and what entered her life during that marriage entered it through him, through a procedure he performed and thereafter through his access. Read that way, the marriage is not the setting in which a difficulty developed. One marriage is the delivery mechanism, and the book is an account of what it is to be attached to a person because they are the source.

That is a harder subject than the one the shelf label advertises, and it is harder for a reason that has nothing to do with substances. An attachment funded by supply does not announce itself. From inside, it presents as a marriage — as preference, as fondness, as the ordinary unexamined fact of wanting to be where a particular person is. The book’s achievement is that it declines to pretend otherwise in retrospect. It does not say that she was tricked.

And the prose is the reason to trust it. It is famously level. It records things that most writers would stage, and it stages nothing. That flatness is usually praised as an aesthetic quality, and it is one, but it is also doing evidentiary work: it is the register of a person who is not writing as someone who was deceived. She is writing as someone who chose, repeatedly, with her eyes open, and who is not going to insult the reader by claiming afterwards that she could not see.

This house has a commercial interest in the argument that follows, because it sells an arrangement. That interest is declared in full in a later section, before the counter-reading rather than after it.

Tove Ditlevsen was born in Copenhagen on 14 December 1917 and died there on 7 March 1976. The Danish biographical lexicons give those dates; one English-language library catalogue record gives 1918 for the birth, and that discrepancy is noted here and not resolved.

She married four times. Viggo F. Møller, an editor, in April 1940, divorced 1942. Ebbe Munk in September 1942, divorced 1945. Carl Theodor Ryberg, a physician, in November 1945, divorced 1950. Victor Andreasen, an editor, in August 1951, divorced 1973. The Danish record describes a crisis around 1950 involving substances, a reduced capacity for work, and a hospital admission from which she came back.

The three volumes appeared in Danish as Barndom and Ungdom, both 1967 and later gathered as Det tidlige forår, and Gift in 1971. The English history is more tangled than the current uniform paperbacks suggest. Tiina Nunnally translated the first two volumes and they were published in 1985 by Seal Press, in Seattle, under the single title Early Spring. The third volume did not follow. It waited another thirty-four years: in 2019 Penguin Classics issued all three as separate books in Britain, with Michael Favala Goldman’s translation of Gift appearing under the title Dependency, and the set was thereafter marketed as The Copenhagen Trilogy. Goldman has written that he found the Danish book in an airport in 2016, read it on a plane, and was surprised to discover it had not been translated into English.

In Japanese the three parts arrived together and late, in 2023, as 結婚／毒 コペンハーゲン三部作, translated by Hidani Reiko and published by Misuzu Shobo — a single volume of some four hundred and forty pages containing 子ども時代, 青春時代 and 結婚／毒. The National Diet Library catalogue records it, with the three Danish titles listed in the contents note.

Tove Ditlevsen died by suicide. That fact belongs to the record and is stated here because it belongs to the record. It is not used anywhere in this essay as a key to the book, and a reader should be wary of any account of the work that reaches for it. A writer’s death is not an interpretation of her sentences, and the habit of reading backwards from it has done more damage to this particular body of work than to most.

The double meaning is the sort of fact that circulates easily and is worth checking rather than repeating, so it was checked. The two senses are separate words that happen to have converged on one form. The adjective descends from the Danish verb gifte, to marry, and means married; the antonym ugift, unmarried, is built straight off it. The noun is a loan from German and means poison, and it generates the ordinary vocabulary of toxicity — the adjective for poisonous, the compound for poison gas. They are not a metaphor that Danish invented. They are a collision that Danish is stuck with.

What Ditlevsen did was decline to disambiguate. She put the bare word on the cover of a book about four marriages and about a decade in which something was being administered to her, and left the reader to do the arithmetic. It is not a pun in the sense of a joke. It is closer to a refusal to say which of the two things the book is about, on the grounds that the question is badly formed.

English could not follow. Goldman has described the problem in print and has said there is no English equivalent; the title Dependency was arrived at as a substitute, suggested to him by a friend. It is a good solution and it is also a consequential one, because it replaces a word that names two relations with a word that names a condition. Married and poison are things that happen between people and substances. Dependency is a state of a person. The English title quietly relocates the book from the space between Ditlevsen and her circumstances to the inside of Ditlevsen, and that relocation is a large part of why the book is shelved where it is shelved.

Japanese had a freedom that English did not, and used it. 結婚／毒 does not choose. It sets marriage and poison side by side with a stroke between them and lets the reader hold both, which is what the Danish does by accident and what the English cannot do at all. Whether the Japanese publisher intended that argument is not something this Library can establish, but the effect on the shelf is real: a Japanese reader meets the book already knowing that two things are in play.

And the obsolete third sense is worth keeping in view, even though it is etymology rather than live usage. The same form once meant a thing given. It is not available to a modern Danish reader as a meaning, and nothing in the book depends on it. But for an essay about supply it is a useful accident, because the entire difficulty the book describes lives in the gap between a thing given and a thing supplied.

Here is the reading, stated plainly. The usual account runs: a talented woman developed a dependence, and her marriages provide the domestic scenery against which that happened. Reverse the priority. The book’s subject is an arrangement, and the substance is what the arrangement delivers.

The distinction matters because the two readings predict different sentences, and the book supplies the second kind. An addiction memoir organises itself around a substance and treats the people as circumstances: they help or they hinder, they enable or they intervene. Ditlevsen’s book organises itself around who has access to what, and treats the substance as the thing that flows through a relation. The published accounts of the book, including the translator’s own, describe the beginning of it as a medical event inside a marriage — an injection administered by a doctor during a procedure — and the consequence as a reorganisation of her life around proximity to the person who could administer more.

This Library has not read the Danish, and says so. What is described here comes from the published record, from criticism, and from the translator’s own account of the text he worked on. Nothing in this section should be taken as a report of what any particular page says.

But the structure is not in dispute, and the structure is the argument. To be attached to somebody because they are the source is not a variety of love and not a counterfeit of it. It is a third thing, which runs on the same hardware. The same wish to be near, the same relief at the sound of a key in a door, the same intolerance of their absence, the same willingness to overlook. Every signal that a person would ordinarily read as evidence of attachment is present and is functioning correctly. It is simply being generated by something other than the person.

And that is why it cannot be detected from inside by feeling. Feeling is the instrument that has been captured. The only thing that would settle the question is a controlled comparison — the same life with the same person and without the supply — and that comparison is not available to anyone while they are living it. Ditlevsen appears to have understood this perfectly well, which is the most striking thing about the book and the thing the addiction-memoir shelf cannot accommodate.

The reading above would be a psychological observation and nothing more, except that the Denmark of the 1940s had built its arrangements in a way that made a doctor’s household an unusually concentrated form of access. This is where the essay stops being about a marriage and starts being about a country.

Danish scholarship on the history of drug policy makes the position clear. Before 1955 there was no Danish criminal offence of possessing controlled drugs for personal use. Criminal and administrative sanction was aimed at supply, and it was applied chiefly through the regulation of doctors and chemists. The state had built its entire apparatus of control around the prescribing profession, which means that in the years in question the profession was simultaneously the only lawful channel and the sole object of the law’s attention.

The same scholarship records a second division, and it is the one that does the most work here. The white paper that preceded the 1955 Act distinguished two populations: people who had become dependent in the course of medical treatment, who were registered by the health authorities and handled inside the medical system, and the newer Copenhagen drug scene, registered by the police. The historians summarise the effect without softening it — underclass users were constituted as criminals, and middle- and upper-class users were constituted as patients.

Put the two facts together and the household comes into focus as a structure rather than a story. A woman who entered by way of a doctor, inside a marriage, in a professional home, was in the one channel that the law treated as medicine. There was no black market to enter and no offence to commit. The supply was lawful, domestic and administered by a spouse, and every feature of the surrounding system encouraged everyone concerned to describe what was happening as treatment.

There is a second monopoly in the same hands, and it should be said. Abortion in Denmark was not available on request until the Act of 13 June 1973, which took effect that October. Under the law in force during her marriages, termination was lawful only on narrow grounds — medical, ethical, eugenic, and after 1956 a further ground framed around physical or mental defect. Danish historical accounts estimate that many thousands of illegal terminations were performed annually before 1973, by doctors, by midwives and by people with no qualification at all. What a physician husband held in that period was therefore not one form of privileged access but two, and the published accounts of the book place the first injection at exactly the point where the two met.

None of this is an accusation against any individual. It is a description of what the era had made possible, and of how much discretionary power the law of that time had concentrated in a single professional role inside a private house.

This Library has an essay on Denmark, at denmark-equality-at-home-and-what-still-does-not-balance, and its argument is worth restating accurately before it is built on, because the two pieces are about the same country at two different moments and could easily be confused.

That essay is about infrastructure. Its claim is that Denmark did not ask people to behave better; it built things — subsidised childcare at a scale that converts care from a private negotiation into a utility, parental leave with weeks reserved in a father’s own name, compulsory sex education from 1970, an individual rather than household tax unit, legal recognition for same-sex couples from 1989, a consent-based definition of rape in 2020 — and let behaviour follow the architecture. And its finding is the uncomfortable one: having removed the structural obstacles, Denmark did not thereby deliver intimacy, because intimacy was never downstream of them. What infrastructure removes is an obstacle. What it cannot supply is capacity.

Every date in that list is later than the events of this book. The childcare, the earmarked weeks, the curriculum slot, the individual tax unit, the abortion statute: all of it arrives after the decade Gift describes, and most of it after Ditlevsen was dead. The Copenhagen of these marriages is the same country before the building.

Which gives the two essays a relation worth naming precisely, rather than a resemblance. The Denmark essay observes what is left when the structural problems are gone. This book shows what happens when they are not: the functions that a public system would later perform did not go unperformed, because private people performed them. A physician husband in 1946 was, for the woman married to him, something close to a one-person welfare state. He held the lawful route to a controlled substance and the practical route to a termination, and both of those were, in that decade, discretionary personal favours rather than entitlements.

That is the observation this Library would put to a reader here, and it is the opposite of the admiring reading of the Nordic countries. The point is not that Denmark later got it right. The point is that where an entitlement does not exist, somebody supplies it anyway, and a relation that carries an entitlement is a relation nobody can assess fairly — not the person receiving, and, one should assume, not the person supplying either.

The most frequently praised quality of these books is their levelness. Ditlevsen reports without commentary, and does not pause to explain or excuse. Readers describe the effect as unbearable precisely because nothing in the writing indicates that it should be.

Scholarship on the trilogy treats that levelness as a constructed effect rather than an absence of effort. Torben Jelsbak, writing on the trilogy in the Nordic-studies journal Nordiques in 2022, describes a narrative method that moves freely between the retrospective position and the immediate recreation of past feeling, handling the younger self almost as though she were a character in fiction, and supplying lived experience without the backward-looking explanatory or apologetic remarks a memoirist is expected to provide. He notes that the apparent neutrality is what produces the impression of authenticity, and that the books were written more than a quarter of a century after the events they describe.

Both things are true at once, and an honest essay has to hold them. The flatness is evidence, in the sense that it is the register of a writer declining the two available exits — the exit marked I did not know, and the exit marked I could not help it. Neither appears. A woman who wished to be absolved had a well-worn vocabulary available to her in 1971 and did not use a word of it. That is a choice about self-description, and it is the single strongest support for the reading that she chose with her eyes open.

And the flatness is craft, in the sense that a professional writer of thirty years’ standing, working at a distance of twenty-five years, produced exactly the tone that would make the reader believe there was no tone. Those are not incompatible. But they do mean that the essay’s central piece of evidence is also its most vulnerable one, and the counter-reading below says so without softening it.

This Library has written about a flat register before, at simple-passion-an-account-without-shame, and the comparison is useful mostly for where it fails. That essay argues that Ernaux built a register with no place to put shame, and it makes a good deal of the fact that she had contemporaneous diaries to work from. Ditlevsen had no such record standing behind the book, and a quarter of a century standing in front of it. The flatness is doing different work in each case: in one it is a method held against material that was written down as it happened, and in the other it is a reconstruction performing the absence of a reconstruction.

Two earlier pieces sit close enough to this one that the difference has to be stated rather than assumed.

At acts-of-desperation-the-self-you-shrink, this Library argued that the narrator of Megan Nolan’s novel is not confused and is not suffering from low self-esteem; she is getting something specific out of her subordination, namely relief from the work of being a self, and the shrinking is not the price of that relief but the product she is buying. At days-of-abandonment-the-self-that-was-load-bearing, it argued that a long arrangement can absorb capacities — the ability to be alone, to hold one’s own attention — and return them, if it ends, in worse condition than it borrowed them.

Both of those describe something the relation itself generates. The subordination produces the relief; the arrangement absorbs the capacity. In each case the thing in question is not separable from the relation, which is unpleasant but at least legible: end the relation and the thing goes too.

The case in this book is a third one and, this Library would argue, the harder of the three, because what is being obtained passes through the relation rather than being produced by it. The substance is not the marriage. It exists independently, it has a name, it could in principle be obtained elsewhere, and its separability is exactly what makes the arrangement unassessable. A person in that position can tell herself, accurately, that the two things are distinct — that she loves him and separately needs the other thing — and she will be unable to test the claim, because the only available test would be to remove one and observe the other, and removing one is the thing she is least able to do.

That is not a special property of substances. It is the general shape of any arrangement that supplies something the person in it cannot easily get elsewhere, and the list of such things is long and mostly unremarkable: money, an address, a visa, a name, status, work, a household that functions, an audience, or simply the reliable presence of another body in the evening.

Here is the sentence this Library would take from the book, and it is deliberately not a moral one.

When an arrangement supplies something you need, every instrument you would use to evaluate the arrangement is downstream of the supply. Your judgement about whether you are happy, your account of why you stay, your sense of what he is like, your estimate of what you would do if it ended: all of those are formed inside a system that is currently being fed. This is not a failure of honesty and it is not weakness. It is a measurement problem, and no amount of character solves it.

The second half is the part people find harder to accept. Clarity does not reduce need. It is tempting to believe that seeing the structure of an arrangement is most of the way to being free of it, because that is what insight promises and what a certain kind of writing about relationships sells. Gift is a sustained demonstration that this is false. The levelness of the prose is, among other things, a record of somebody who could describe her own position accurately and whose accurate description changed nothing about what she did next. Understanding is not an exit. It is an inventory.

What does help, as far as this Library can see, is not insight but arithmetic of a duller kind. A second source of the thing, so that the arrangement stops being the only channel. A person outside the arrangement whose view of it is not funded by it. And a willingness to answer a question that is narrower and much more answerable than whether you love somebody: what specifically would stop arriving if this ended, and is there any other way for it to arrive.

That last question does not diagnose anything and does not require anybody to leave anything. It is a description of a supply chain, and most people find it easier to be honest about a supply chain than about a feeling.

This house sells a bounded, scheduled, paid arrangement in which a person receives attention. It has a direct commercial interest in the argument above, and the interest is not adjacent to the argument — it is the same shape as it. An essay whose conclusion is that arrangements supply things, and that the supply is hard to see from inside, sets up rather neatly a house whose own arrangement has written terms, a start time and an end time. The reader should assume that this was noticed here before it was noticed there.

So the declaration is made as constraints rather than as a disclaimer, and the constraints cost something.

First: a paid arrangement is an arrangement that supplies something, and everything said above about the difficulty of evaluating one applies to this one without modification. Its terms being legible does not exempt it. A person who comes here regularly is receiving something, and will develop, exactly as anybody would, a set of judgements about the arrangement that are formed downstream of receiving it. This house is not outside the problem it has just described.

Second: this house is not a second opinion about anybody’s life. The paragraph above recommends a person outside the arrangement whose view is not funded by it, and this house is funded by it. Whatever else an evening here is, it is not that, and anybody who needs that needs someone who is not being paid.

Third: no outcome is claimed. Nothing here is offered as help with dependence of any kind. This house has no competence in that and does not pretend to any; where that is the question, the people to ask are medical, and this is not a referral, an assessment or an opinion about anyone.

Fourth: legibility of terms is not clarity of feeling. The house’s standing argument is that a stated arrangement is easier to think about than an unstated one. That is a claim about thinking, not about feeling, and the two are not the same distance apart for everybody.

Five objections, given at full strength, and the last of them is not answered.

The first is that supply is a frame, and this is a book that refuses frames. The scholarship on the trilogy makes exactly this point about its method: it withholds the retrospective explanation that a memoir conventionally supplies. An essay that arrives thirty years later with the explanation the book declined to give is doing the thing the book was built not to do, and the neatness of the reading should itself be treated as a warning sign. The honest reply is limited. This is a reading, offered as one, and the book is available and does not need it.

The second is the serious one. Carl Theodor Ryberg was a real man, a physician, and treating a marriage as a delivery mechanism denies him any interiority at all. Everything said about him here comes through the account of a woman who divorced him, written a quarter of a century later, and this Library found nothing that he wrote or said about the marriage. The two Danish records consulted do not even agree on the year he died, which is a fair measure of how thin the independent record of him is, and no year is stated here for that reason. He may have believed he was treating his wife. He may have been right about parts of it. He is not available to say, and an essay that converts a person into a mechanism should be read with that in mind — including this one.

The third is memory. Gift was published in 1971 about events beginning in the early 1940s, and it was written, by the Danish record, in the context of psychiatric hospitalisation, by a professional writer with a public, a reputation and children. Twenty-five years is long enough for a life to become a narrative, and the scholarship describes a method that handles the younger self almost as a fictional character. A reader is entitled to treat the whole of it as reconstruction, in which case the essay above is a reading of a construction rather than of a decade.

The fourth is the one that undercuts the central evidence. The essay treats the flat prose as testimony — as the register of someone who chose knowingly. But flatness is a style, it was the property of a writer at the height of her powers, and nothing distinguishes, from the page, an unflinching report from a very well-made performance of one. The strongest version of this objection is not that she was lying. It is that the essay has taken an aesthetic effect and promoted it to a piece of evidence about a woman’s state of mind, which is not a thing prose can establish.

A related point comes from the reception itself. Writing in the Los Angeles Review of Books in June 2021, Nina Renata Aron argued that reviewers met this book with a horror disproportionate to what is actually on its pages, and that in doing so they denied Ditlevsen agency as a writer, expecting from a woman a narrative of shame that male accounts of the same material are not asked to supply. That objection cuts at this essay too, from the other side: an essay insisting that she chose with her eyes open is also an essay deciding what a woman’s sentences prove about her.

The fifth is the plainest. A house that sells a paid arrangement is not a disinterested reader of a book about what arrangements supply. The reading is convenient, the conclusion flatters the product, and no amount of declaring the interest makes the interest go away. There is no answer to this beyond disclosure, and the disclosure has been made above rather than here.

This essay is a reading. It is not clinical guidance, it prescribes nothing, it promises no outcome, and it diagnoses nobody.

It does not diagnose Tove Ditlevsen. No clinical term has been applied to her as a person anywhere in either version of this text, and none should be. She was a writer who published a book, and what can be said about her here is what the published record says and what her sentences do. Her death is recorded once above because it is part of the record, and it is not used as an interpretation of anything she wrote.

It is not a cautionary tale. Nothing here is an argument about substances, about the dangers of anything, or about what anybody should avoid. A reader who takes from this a warning has taken something the essay did not put in.

It makes no claim about Carl Theodor Ryberg beyond the dates of a marriage and a profession, both from Danish biographical records, and it states in the section above that the reading offered here denies him the interiority he would have had.

It is not a statement about Denmark, about Danish people or about national character in any direction. The material on Danish law and drug policy describes what statutes and scholarship record about arrangements in a particular period, and the essay’s reading of what those arrangements made possible in one household is this Library’s inference and is marked as one.

It is not a statement about Japanese readers either, and no inference is drawn from the fact that the Japanese translation arrived in 2023.

No author, translator, publisher, scholar, critic or publication named here has any connection to this house, knows of it, or endorses it. Book and edition details appear so that a reader can find the work; there is no retail arrangement of any kind behind them.

And the argument this house takes from the book is offered as an argument, not a finding: that an arrangement supplying something you need is very hard to evaluate from inside it, and that clarity about the arrangement does not reduce the need. A reader who concludes that this describes nothing in her life has read the essay correctly.