The Lost Daughter, and the Sentence a Mother Has Nowhere to Say

Late in this film a woman tells a near-stranger the worst thing she has ever done, and the film declines to punish her for it.

The setting is a Greek island in high summer. Leda Caruso, a professor travelling alone, has taken a rented apartment for a working holiday and finds the beach occupied by a large, loud extended family from Queens. Among them is Nina, a young mother with a small daughter, and Leda cannot stop watching her. A child goes missing and is found. A doll goes missing and is not. Over some days the two women circle each other, and then, in a beach bar, Leda says it.

She says that when her elder daughter was seven and her younger was five she left, that she did not see them for three years, and that when Nina asks what it was like, the answer is that it felt amazing. She goes back afterwards, she says, because she missed them, and then she calls herself a selfish person and an unnatural mother.

What matters is what happens next, which is nothing. The film does not produce a diagnosis. It does not produce a husband whose cruelty explains her, or a depression that absorbs her, or a scene near the end in which grown daughters forgive her and the account is settled. It lets the sentence stand.

That refusal is the subject here. The argument of this essay is that the thing the film exposes is not bad mothering at all. It is the absence of any socially available sentence for I love them and I wanted my life back — a sentence with no register to be said in, no listener who can receive it as information, and therefore nowhere to go except into the ear of somebody one will never see again.

The source is a novel. Elena Ferrante published La figlia oscura with Edizioni e/o in 2006, and Ann Goldstein’s English translation appeared as The Lost Daughter from Europa Editions in 2008. Ferrante is a pseudonym and the author does not appear in public, which is itself relevant to a story about what a woman may put her name to.

Maggie Gyllenhaal acquired the film rights in October 2018. The condition attached to them has been widely reported and is worth stating because it is unusual: Ferrante granted the rights on the understanding that Gyllenhaal would direct the adaptation herself, and that the agreement was void otherwise. Ferrante also published a column in the Guardian in 2018 giving the project her public blessing, saying in substance that it mattered to her, for Gyllenhaal and for women generally, that the work be Gyllenhaal’s own.

The film was shot in Spetses, Greece, between late September and the end of October 2020 — the setting moved from the Italian coast of the novel to a Greek island, and the family from Naples to New York. It premiered at the 78th Venice International Film Festival on 3 September 2021, where Gyllenhaal took the festival’s award for Best Screenplay. It had a limited theatrical release on 17 December 2021 and arrived on Netflix on 31 December, which is how most people saw it, including in Japan, where it is listed under a transliterated title.

On the awards, precision matters more than enthusiasm. At the 94th Academy Awards, held on 27 March 2022, the film received three nominations and won nothing. Olivia Colman was nominated for Best Actress and the award went to Jessica Chastain. Jessie Buckley was nominated for Best Supporting Actress and the award went to Ariana DeBose. Gyllenhaal was nominated for Best Adapted Screenplay and the award went to Siân Heder for CODA. The film did win elsewhere — Best Feature and other prizes at the Gotham Awards in November 2021, and Best Feature, Best Director and Best Screenplay at the Independent Spirit Awards in March 2022 — and those are different bodies with different electorates, and saying so is not a demotion.

One further fact of circulation is worth recording for readers here. The four Neapolitan novels have been published in Japanese by Hayakawa Shobo in translations by Iida Ryosuke. The National Diet Library catalogue records those and does not record a Japanese edition of La figlia oscura. A catalogue absence is not proof of non-existence, but it does suggest that in Japanese, Leda is met mainly through the film rather than through the book.

Consider what a woman who has left is permitted to say afterwards, and notice that every permitted version does the same work.

She may say she was unwell, which converts the wanting into a symptom: something that happened to her, which treatment addresses, which is not a statement about what she preferred. She may say she was very young, which converts it into inexperience that time corrects. She may say her husband gave her nothing, which converts it into a grievance against a third party and makes her the injured one. She may say she regretted it every day, which converts it into a penance and restores the moral order. Each of these has a place. Each can be said at a table with other adults without the room changing temperature.

The sentence that has no place is the one where both halves are true at once and neither cancels the other. I love them. I wanted my life back. I went, and it was not a breakdown, and it was not a misunderstanding, and I was not principally fleeing a man, and the years were not miserable — they were the opposite — and I came back because I missed them, and none of that adds up to a story with a moral in it.

The important claim here is about speech rather than conduct. This essay makes no argument that leaving is fine, or that it is not. It observes that the description has been foreclosed, and that the foreclosure does damage on its own account, independently of anything anybody did. A person can behave impeccably for thirty years and still be carrying a true sentence about herself that there is no occasion to say.

Note also where the judgement actually sits. It is easy and slightly flattering to locate it in other people — the mothers at the school gate, the mother-in-law, the culture. Some of it is there. But the sharper problem is structural rather than interpersonal: even a sympathetic listener has no slot for the sentence, because the available conversational moves are reassurance, alarm, or referral. Reassurance says you do not really mean that. Alarm says we should get you some help. Referral says have you spoken to somebody. All three are kind. None of them is hearing.

Leda does not say it to her daughters, who are adults and reachable by telephone. She does not say it to a friend, a colleague, a former husband or a professional. She says it to a young woman she has known for a few days on a beach, and the film is exact about why that is the only address available.

Two properties make Nina sayable-to. She has no standing to be injured by the disclosure — she is not one of the daughters, not a relative, not anybody whose life is arranged around Leda’s account of herself. And she is temporary. The holiday ends, the island empties, and a sentence said to a person who will disappear does not have to be lived with afterwards. That combination is, for most adults, the entire infrastructure available for saying an unsayable thing: a train, a hotel bar, a hairdresser, a stranger in a waiting room.

But Nina is not a void. She is a young mother in the middle of exactly the condition Leda is describing, tired and cornered and already looking for the exit. So the sentence lands as an instruction when it was offered as information. Leda is reporting a fact about her own past. Nina hears a prophecy about her own future, and worse, a permission slip, and worse still a warning — all at once, from a woman she had decided to admire.

This is why the disclosure detonates rather than relieves. The encounter does not end in recognition and it does not end in comfort. It ends in an act of violence with a hatpin, which the film stages without commentary and without making Nina wrong. A sentence with nowhere to go does not become safe by finding a listener; it becomes dangerous, because the only people positioned to hear it are the ones with the most at stake in what it might mean. When a culture supplies no ordinary channel for a thing, the thing does not stop existing. It surfaces at full pressure in the least suitable place, and the tidy conclusion afterwards is that she should not have said it.

A conventional treatment of this material has a shape, and the beats are worth listing so that the film’s refusals are visible as choices rather than as gaps. It establishes a cause sufficient to the act — the mother ill, the marriage violent, the circumstances extraordinary — so that the audience can hold a stable position toward her. It then requires an arc: remorse, understanding, and either restoration or ruin. And it resolves with an accounting: a reunion, a deathbed, a letter, a scene in which grown children say the words that release her, or a final image confirming that she paid.

This film does none of it. The young Leda is stretched thin, ambitious, brilliant, sensual and frequently exasperated, and the film shows her exasperation with children as ordinary rather than as pathology. The marriage is unhelpful without being monstrous. The affair happens and is not offered as the reason. There is no scene in which she recants. There is no scene in which she is forgiven. The daughters are not brought on at the end to settle the ledger.

The ending instead returns her to a beach, alone, injured, with the sun coming up, peeling an orange and speaking to her daughters on the telephone. What she tells them is that she is alive. Published criticism has made a good deal of the difference between that and the novel’s corresponding formulation, which is reported by critics as considerably bleaker; readers who want to weigh the change should read the novel and the criticism rather than take this essay’s word for it, since the reading offered here is a reading of the film.

The point for our purposes is narrower. By withholding both the sufficient cause and the settlement, the film leaves the audience holding something it is not used to holding: a woman who did a thing that hurt her children, who is not sorry in the way the form requires, who is also not proud, and who is not thereby a monster. That is not evasiveness. It is the substance of the work.

Three frames are available for a mother who says she wanted out, and the film refuses all three.

The first is moral: she is a bad woman, unnatural, and the interest of the story is in her exposure. The second is exculpatory: she is a victim of her circumstances — of an absent husband, of a society that gave her no support, of poverty or of an era — and the interest of the story is in the circumstances rather than in her. The third is clinical, and it is the modern one and the most flattering to everybody: she was unwell, the wanting was a symptom, and the appropriate response is care.

The clinical frame deserves particular attention, because it is offered generously and because it is often correct. Distress is real, treatable, and frequently untreated, and a person who finds herself wishing to disappear may well be describing something that would benefit from a professional’s attention. Nothing in this essay says otherwise, and nothing here is a view about any particular person’s interior. But a frame that is sometimes correct is not therefore always correct, and when it becomes the only available frame it stops being care and starts being a way of not listening. Explaining a person’s preference as a symptom means never having to treat it as a preference.

What the film puts on screen instead is ambivalence as a stable property. Leda at forty-eight is not cured of the thing she was at thirty. She is still delighted and still irritated and still hungry; she still finds children exhausting and still adores her own; she is still the woman who found those three years amazing. The film treats that as a description of a person rather than as a phase she has failed to complete.

The people who made it have said versions of this in public, and their remarks are worth quoting as theirs rather than as authority. Jessie Buckley, who plays the younger Leda, told Refinery29 in December 2021 that she would never judge Leda as a bad mother. Olivia Colman told the British Film Institute that same month that a viewer might watch and think that she had thought of doing that, had not done it, and was not alone. Gyllenhaal told W magazine in March 2022 that people respond to being told the truth, especially about something taboo. Those are the views of three people with a professional stake in the film. They are evidence of intent, not of effect.

The film’s central device is a stolen doll, and it is not decoration.

When Nina’s small daughter goes missing on the beach, Leda finds her. She also takes the child’s doll, and she keeps it, and she does not give it back. She keeps it while the family searches, while a reward is posted, while the child is inconsolable, while Nina becomes desperate. She washes it in her sink. She buys it clothes. When she is finally confronted she says she was just playing.

Consider what that object is doing structurally. It converts an interior state that cannot be spoken into an act that is entirely visible and entirely unreadable. Everybody in the film can see that somebody took the doll. Nobody can construct a motive, because there is no motive in the vocabulary available — not theft, not malice, not madness exactly, not anything with a name. Leda herself cannot supply one, and the honest answer she does give is closer to a child’s than to an adult’s.

So the doll is the sentence in physical form: present, obvious, sitting in plain sight in her apartment, and unspeakable. It also solves a problem of dramaturgy, since a film about a woman who wants to say a thing and cannot would be a monologue, while a film about a woman who commits a small, senseless, cruel act instead and then sits in a room with it for a fortnight is a film.

The temptation to make the doll tidy should be resisted. It is a child’s toy and a stand-in for a daughter and an object drawn from Leda’s own memories of her mother, and a film in which it meant only one of those would be a lesser film. What it does reliably is deny the audience a clean position. Whatever sympathy the confession scene earns, the doll is there in the background, a small girl has been made to suffer for nothing, and the cruelty is not excused by being legible to us as displacement.

This chamber has been here twice already, from two different directions, and this essay is only worth publishing if it goes somewhere the earlier two did not.

The first is our reading of Tsushima Yuko’s Woman Running in the Mountains, which follows a young woman who has a child alone and finds that neither of the two slots available to her — the girl in trouble, then the mother — is issued with a body that wants anything on its own account. That essay’s subject is continuation: desire that never went anywhere, a body that misses being able to move, and a novel that declines to say whether her silence was right. It also makes a point this essay inherits, that a person can be surrounded by accurate information about herself and still be entirely unmet, because the accuracy fills the space where a question would have gone.

The second is our essay on single parenting and a romantic life, which argues that the binding constraint on a single mother’s intimate life is not stigma but arithmetic — hours, childcare, an unbreakable return time, an unpredictable child — and that the two social framings available to her, pity and admiration, both erase the person by promoting her to a function. We have also written on Ferrante before, in the piece on The Days of Abandonment and the self that is larger than the role of wife.

Set those beside Leda and notice what has been removed. She is not short of a body that wants things; she has one and uses it. She is not short of hours; she is a professor alone on a Greek island with a stack of books and nothing to do. She is not short of money, standing, education, or the confidence to hold a room. Every constraint the earlier essays identified has been lifted, and the thing that is still missing is still missing.

That is the new scarcity, and it is the one that cannot be funded. Childcare can be paid for. Hours can be bought, transferred, legislated into being. A vocabulary cannot. There is no policy that issues a woman with a sentence she can say to another adult about loving her children and having wanted her life back, and no amount of material comfort produces one. Which is why the essay before this one could end with arithmetic and this one cannot.

Three objections to the film, and two to this reading of it. They are not straw versions and the responses are narrowings rather than rebuttals.

The first is class, and it is the heaviest. Leda’s ambivalence is affordable. She is a tenured academic on a holiday she chose, in a rented apartment, with the leisure to be interesting about her own past and the vocabulary to make it sound like Yeats. When she left her children she left them with their father and went to a career that was waiting for her. A woman without money who leaves her children does not get a beach bar and a sympathetic listener; she gets a case file, a caseworker and, quite possibly, a court. The film’s charity toward Leda is partly bought with Leda’s advantages, and an identical story about a woman with no resources would be much harder to watch and much less likely to be nominated for anything. That objection is correct and there is no answer to it. What survives is narrower: the film demonstrates that the refusal to resolve a mother into monster or victim is possible at all, which had not been much demonstrated, and a demonstration made under favourable conditions is still a demonstration.

The second is the daughters. Bianca and Martha exist in the film mostly as pressure — noise, need, interruption, a hand pulling at a sleeve while their mother is trying to work — and as adult voices on a telephone. The film spends two hours inside the experience of the person who left and almost none inside the experience of the people who were left. Ambivalence is cheap to depict; consequences are expensive, and they are borne by characters the film does not follow. A film about those two women at thirty would be a harder film and nobody has made it. Concede this fully. The most one can say in mitigation is that Nina and her small daughter are where the cost is made visible, and that the cost arrives in the form of a blade.

The third is aesthetic. Greek light, linen, a beautiful actress being troubled beautifully, a score that tells you when to feel uneasy. Discomfort this well appointed is a kind of comfort, and a viewer may leave feeling that she has confronted something when what she has done is watch something lovely.

Now the case against this essay, and the sharper version comes second. The first objection is that the central claim is close to unfalsifiable. How would anybody establish that no socially available sentence exists? The argument rests on the writer failing to think of one, which is not evidence. And mothers plainly do say this to each other — in private, among friends, at two in the morning. The narrowing is real and is accepted: the claim is about availability in ordinary public registers, not about total absence. It is not what one says to a colleague, a school, a mother-in-law or a paediatrician. That it survives in private speech between trusted women is precisely the point about where it can be said, not a refutation of it.

The second objection is the one that should worry this house most. An essay that praises a film for refusing to judge is itself in the business of not judging, and a reader who arrived wanting permission has been given a long and well-sourced document that declines to withhold it. The essay protects itself by saying it recommends nothing, and that protection is thin, because the persuasion is done by the framing and the conclusion is supplied by the reader. There is no clean answer. The only thing available is to say what is not being said: this essay does not claim that leaving is acceptable, does not claim that it is not, holds no view on what any reader should do about her own life, and would regard a reader who took it as encouragement to act as having been failed by it.

Before anything else in this section, the interest. This house sells private companionship, and an argument that a woman may be carrying a true sentence with no one to say it to is an argument that flatters our position exactly. A reader persuaded by the preceding pages has been persuaded that there is a kind of listener she does not have. We sell hours with a listener. That is the commercial shape of this essay and it should be visible before any of the rest.

It is also worth being uneasy about how neatly the structure fits. The essay argued that the sentence can only be said to somebody with no standing to be injured by it and no permanent presence to hold it. That is a fairly exact description of a paid companion, and an argument that arrives at its author’s product by the front door deserves to be looked at twice.

So the constraints, stated as refusals. An evening here is not therapy and is not offered as a substitute for it. Nothing about it treats, resolves, processes or repairs anything, and a house that told you otherwise would be selling you an unfalsifiable promise, which is the standard mechanism by which people in a vulnerable position are taken advantage of.

Nor can it supply the missing vocabulary. What a bounded evening can offer is a temporary exemption from needing one — a few hours with somebody who has no stake in your account of yourself, and who will not be at the funeral, and to whom a thing can therefore be said without rearranging a life. That is a real thing and it is a small thing, and the distance between it and having somewhere to say the sentence is the whole subject of this essay.

And if this house became the only place in a person’s life where such a thing could be said, that would not be a success for us to celebrate. It would be a description of a gap, and the gap is not one an evening closes. We make no claim about anybody’s children, offer no guidance about anybody’s family, and hold no opinion about what any woman should do.

This essay is film criticism. It is not advice to any reader about her own children, it is not a diagnosis of anybody, and it is not a verdict on any real mother. Nothing here says that leaving is right or that it is wrong, and a reader who takes it as encouragement or as permission has been given something this essay did not offer.

Leda is a fictional character in a novel and a film. No claim is made about Elena Ferrante’s life, about Maggie Gyllenhaal’s, or about any real person’s experience of motherhood. The essay’s reading is a reading of the film. Claims about the novel rest on its publication record and on published criticism rather than on a fresh reading of it here, and readers who want to weigh how the adaptation departs from its source should go to the book.

On the awards, the record is stated exactly and should not be inflated in retelling. The film received three Academy Award nominations at the 94th ceremony and won none of them. It won the Best Screenplay award at the 78th Venice International Film Festival, and Best Feature at the Gotham Awards and Best Feature, Director and Screenplay at the Independent Spirit Awards. These are different awards from different bodies.

Remarks attributed here to Maggie Gyllenhaal, Olivia Colman and Jessie Buckley are taken from published interviews, are dated in the text, and represent what those people said at that time in the course of promoting a film they made. They are cited as statements of intent and reception, not as authority on what the film achieves, and none of them has any connection to this house or knows of it.

Only very short fragments of film dialogue appear above, for the purpose of analysing them. No passage of the novel is quoted anywhere.

The observation about Japanese publication is a statement about what a library catalogue records. The National Diet Library records Japanese translations of the four Neapolitan novels and does not record one of La figlia oscura; a catalogue absence is not proof that no edition exists.

On the clinical question, this essay takes no position about any individual. It argues that the symptom frame is one of several and should not be the only one available in public speech. It does not argue that the frame is wrong, and a reader who is struggling should treat this as an essay about films rather than as a reason to decline help.