Tokyo Sonata, and the Suit He Keeps Putting On

A man in a suit leaves a house at the same hour every morning. The suit is pressed, the bag is in his hand, and nothing visible in the frame has changed. He has been made redundant. Rather than say so, he has decided to keep the mornings intact, which means that he now performs, at considerable daily expense, the appearance of the thing he no longer has.

That is the premise of Tokyo Sonata, directed by Kurosawa Kiyoshi and released in Japan in September 2008, and it is almost always described as his premise: the story of a father. The father, Sasaki Ryuhei, is played by Kagawa Teruyuki. His wife Megumi is played by Koizumi Kyoko. There are two sons, one old enough to leave and one still at primary school, and each of them is running a concealment of his own. The film lasts a hundred and nineteen minutes and spends them on four people who share a house and do not, in any working sense, share a story.

This reading puts the wife at the centre. It does so knowingly against the grain of most accounts of the film, and against the director’s own stated emphasis, which is given its due later and is not softened when it arrives. The argument is that the father’s deception is the film’s engine rather than its subject; that the subject is the person who keeps the arrangement running while the deception runs inside it; and that the film’s least explicable passage, the night Megumi spends away from the house, is the one place where a want that belongs to her is permitted to surface, and is given nowhere at all to go.

The claim underneath is not about a family and not about a country. It is about a structure. An arrangement can be held together by somebody who has no power to change its terms, and competence at holding it is precisely what makes that person invisible inside it. That is a claim about how a household can be organised. It is not a claim about Japanese households, about Japanese women, or about any national disposition whatsoever, and nothing that follows should be read as one.

Tokyo Sonata is a co-production of Japan, the Netherlands and Hong Kong, running a hundred and nineteen minutes, written by Max Mannix, Kurosawa Kiyoshi and Tanaka Sachiko, photographed by Ashizawa Akiko, and released in Japan on 27 September 2008 through Pix. Alongside Kagawa and Koizumi, Koyanagi Yu plays the elder son Takashi, Inowaki Kai plays the younger son Kenji, Igawa Haruka plays the piano teacher, and Yakusho Koji plays the burglar who appears late and changes the shape of the last movement.

It is worth being exact about the prizes, because the film is often introduced in a way that inflates them. At the sixty-first Cannes Film Festival in 2008 it received the Jury Prize in Un Certain Regard. Un Certain Regard is a parallel section rather than the main Competition, and the Jury Prize is not that section’s top award, which went to Tulpan. This is a real distinction from a serious festival and it is not a Palme d’Or, and a publication that blurs the two is selling something.

At the third Asian Film Awards, held in Hong Kong in March 2009, the film won Best Film, and Kurosawa, Mannix and Tanaka won Best Screenwriter. Those are two wins, and they are the film’s largest awards.

Against that, a fact that is rarely mentioned. At the thirty-second Japan Academy Film Prize, the ceremony covering films released in that window, Tokyo Sonata appears nowhere: not among the winners, and not among the excellence-award nominees in any category, acting, directing, screenplay or craft. The domestic academy did not take it up at all. This is an absence in a published list rather than a positive finding, and it is stated as such, but it is a striking absence for a film that had already taken a prize at Cannes.

Koizumi Kyoko was recognised that season, and the shape of the recognition matters. The eighty-second Kinema Junpo Best Ten gave her its best actress award, the thirty-third Hochi Film Award did the same, the thirty-second Yamaji Fumiko Film Award gave her its actress prize, and she received the fifty-ninth Minister of Education Award for Fine Arts in the film category. Every one of those was awarded for two films together, Tokyo Sonata and another release of the same year. Nobody handed her a prize for this performance alone. That is not a slight on the work. It is a precise description of what the record says, and this essay is going to make a large claim about that performance, so the record should be visible before the claim arrives.

Ryuhei’s department is moved to Dalian because the labour is cheaper there, and he is let go. What he does next is not a lie so much as a job. He leaves at the usual hour. He returns at the usual hour. He queues at the employment office, sits in parks, eats at a soup kitchen, takes a humiliating interview, and eventually finds work cleaning a shopping centre, where he changes out of his overalls and back into the suit before going home.

Consider what that programme actually requires. It requires that the household continue to produce mornings. Somebody has to keep the house running to a timetable that has become purely ceremonial. Somebody has to make the meals that the schedule implies, keep the accounts that the vanished salary is supposed to be feeding, absorb the school’s complaints, and register nothing. The deception is his. The infrastructure the deception runs on is hers.

This is the first thing the usual reading misses. A concealment of this kind is not a solo performance. It is a two-person production in which only one person has been told there is a play on, and the other one is doing the lighting, the catering and the front of house without a script. The father’s secret consumes the wife’s labour whether or not she knows about it, which means that from the very first morning she is inside the plot as a resource rather than as a party.

And the film is careful to show that his authority is the point of the exercise. When he later insists that his younger son give up the piano, he explains his position to Megumi in terms of parental authority needing to be free of wobble. He is not protecting the family from bad news. He is protecting a standing, and the standing is the thing the household is arranged around.

Then the film does something that reorganises everything before it. In the middle of the argument about the piano, with her husband raising his hand to their son and then turning on her for the way she is looking at him, Megumi says that she knows. She has seen him standing in the queue for food. She has known for some time and has said nothing.

What she does next is the most important gesture in the film, and it is not a confrontation. She asks him whether he really wants this said, because saying it will demolish his authority completely. That is not a threat and it is not mercy. It is an accurate statement of the mechanism she is standing in. The single lever available to her is the one that destroys the thing she is charged with maintaining. She can be the person who keeps the household intact, or she can be the person who names the fact, and the arrangement does not permit both.

So knowledge, here, is not power. It is an additional load. Before she knew, she was maintaining a household. After she knows, she is maintaining a household and a fiction, and she is doing it silently, and there is no one to whom the second job can be reported, because reporting it is the act that ends it.

This is why the reading that treats her as passive, or as lacking the will to hold the family together, gets the picture exactly inverted. She is not failing to act. She has correctly identified that the available actions all cost more than they buy. What looks like inertia from outside is a person who has run the calculation and found no move on the board.

Here is the structural claim in full, and it is the reason this film belongs in this Library rather than only in a film column.

The household has assigned Megumi total responsibility for its functioning. Meals, school, money in and money out, the emotional weather of four people, the anticipation of what will be needed before anybody asks. It has assigned her no corresponding standing. She does not set the terms. She did not set them when they were set and she cannot reopen them now, because there is no procedure for reopening them and no forum in which to try. When the elder son suggests, in the scene before he leaves, that she could simply divorce, her reply is not about love or duty. She asks who would then do the mother. She is describing a post that has to be staffed, and she is the staff.

Two research findings are worth putting beside this, both stated as what they are. Allison Daminger, writing in the American Sociological Review in 2019, interviewed seventy people from thirty-five couples and argued that household labour has a cognitive dimension distinct from its physical tasks: anticipating needs, identifying options, deciding, and monitoring whether things are actually happening. She reports that in her sample women did more of this cognitive work overall, and more of the anticipation and monitoring in particular. She also reports something this essay must not skip, because it cuts the other way: participation in decision-making, which she calls arguably the component most closely linked to power and influence, was roughly equal between men and women in that sample. Anticipation and monitoring are not the same as authority, and her data say so.

The second is older and blunter. Bittman, England, Sayer, Folbre and Matheson, writing in the American Journal of Sociology in 2003 using Australian and United States data, found that a woman’s hours of housework fall as her share of household earnings rises, but only up to the point where the two partners earn about the same; that the baseline level of her housework is far higher throughout; and that in the small group of couples where the woman supplies most of the income, the pattern reverses and the division becomes more traditional again. Earning, in other words, does not straightforwardly purchase standing. Something other than the ledger is setting the terms.

Neither study is about this film, this country or this household, and neither is offered as proof of anything about them. What they establish is narrower and sufficient: that a person can carry the anticipation and the monitoring of a shared life without that carriage converting into any authority over its terms, and that researchers can measure the gap. Megumi is a character in a film. The position she occupies is not a fiction.

One national figure belongs here, fenced on both sides. Japan’s Statistics Bureau publishes the Survey on Time Use and Leisure Activities; the 2021 round, released in August 2022, reports for households of a couple and children with a child under six, as a weekly average per day, one hour and fifty-four minutes of housework-related time for husbands against seven hours and twenty-eight minutes for wives, and it reports the same table back to 2001, where the figures were forty-eight minutes and seven hours and forty-one minutes. The Sasaki household has no child under six and is therefore not in that category at all, and an average describes a distribution rather than any home in it, and nothing about a national average licenses a sentence beginning with the words families in this country. What the table does establish is that the asymmetry the film dramatises is not an invention of the screenplay, and that it was being measured, and has been measured for two decades, and has narrowed mostly on one side of the column.

Kurosawa Kiyoshi arrived at this material from somewhere specific. Cure in 1997, and a run of films through the following decade, established him internationally as a maker of dread. Loft in 2006 and Retribution in 2007 immediately precede Tokyo Sonata, and all three were photographed by the same cinematographer, Ashizawa Akiko, who won the Yokohama Film Festival’s cinematography prize for her work that year. The camera that had been looking at Kurosawa’s hauntings turned, with no change of operator, to look at a family at dinner.

The formal consequence is visible throughout, and it is the reason the film unsettles people who would not describe it as frightening. The Sasaki interior is not a location. Kurosawa has said that the inside of the house was a built set, with the ceilings constructed, lit so that artificial light would read as sun; that he does not generally like shooting on sets because a set cannot surprise you; and that he therefore placed the camera in awkward positions, behind pillars and beyond windows, where obstacles made the shot difficult, on the view that the constraint would make the image richer. So the domestic space is watched from outside its own furniture. Doorframes intervene. The family is composed, again and again, as something observed from a position in the room that nobody in the room occupies.

He is explicit that this was not a horror programme. Asked about the film’s opening, in which a storm blows through the house and riffles the papers as though the objects were alive, he said that he does not set out to make things frightening at particular moments, and wondered whether it comes out that way from habit. He said that because he had made so many horror films, empty streets had been eerie and effective, and that in this film he wanted to erase the horror colour as far as possible, which is why he filled the employment office and the food queue with more people than he had ever put in a frame. He said that what he did try to produce was the precariousness of a family of four: that they are holding on to being a family by a margin, and that a small trigger could scatter them, and that an ordinary conversation or an ordinary room could tip into a ruin.

Take him at his word and the point stands differently, and better. The dread in Tokyo Sonata is not atmosphere laid over a drama. It is the drama. The arrangement itself is the thing that could collapse at a touch, and the camera is positioned to keep reminding you that it is being held up by somebody. That is a structural dread rather than a psychological one. Nobody in this house is haunted. The house is the apparatus, and the apparatus is load-bearing on one person, and the film is shot from the angle at which that becomes visible.

Late in the film a burglar breaks into the house, takes Megumi hostage, makes her drive, and keeps her out overnight. They end up at a hut by the sea. He is a failed locksmith, defeated by his own life, and at some point he turns violence on himself and she stops him. In the morning she goes outside and there is nothing there but tyre tracks running down the sand to the water.

This sequence is where most readings falter, and it is easy to see why. It is tonally unlike everything around it. It arrives without preparation, borrows from a different kind of picture, and resolves into an image that will not quite convert into a meaning. Any account of it is contestable, including this one, and the honest thing is to say that at the start.

Here is what can be said with a source. Kurosawa has described inventing the character, in an early redraft, as somebody who would drag the mother out of the house willy-nilly so that she would recognise that she is a human being before she is a mother or a wife. That is his account of the function, given in an interview in December 2008. Read it carefully and it contains the essay’s argument and the essay’s problem in the same sentence. The film cannot get her out of the house by her own hand. There is no plausible action available to this character that would remove her from the arrangement, so a stranger with a weapon has to be written in to do it for her. The eruption is real and it is not hers. It is done to her, and the film seems to know that this is the only door it has.

That is why the sequence reads, to this Library, as a want with nowhere to go rather than as a liberation. Something surfaces in that night that has no name in the household vocabulary and no route back into the household at all. And then it evaporates. The tracks go into the sea. Nothing is carried home, nothing is told, and the next morning she is back in the arrangement with one more thing she cannot say.

Note what she is not given. The father is given an accident, a stretch unconscious in the road, a found envelope of money and a decision about returning it. The elder son is given a war and a letter. The younger son is given a piano and an audition. Each of the men is given an action with a consequence that lands. She is given a night that leaves tyre tracks.

Kurosawa has said that he aimed to portray the four evenly, and that in a sense the film is four self-contained stories interlocking, with four people of different ages and sexes so that any viewer will find somebody to recognise. That description is accurate and it explains the film’s strange rhythm, in which characters disappear for long stretches and return having changed off-screen.

But four parallel lines do not fall on the four people equally, and the asymmetry is the whole argument. Ryuhei’s line is private to him. Takashi’s line takes him out of the country. Kenji’s line is conducted in secret, funded by diverted school money. Megumi does not have a line of her own in that sense. She is the intersection. She is the only character who appears inside all three of the others: taking the school’s complaint about the lunch fees, covering for Kenji, being asked to sign the elder son’s consent form, absorbing the husband’s temper about authority. Her screen time is not spent on a strand. It is spent servicing everyone else’s.

Which is exactly what makes the argument hard to see. A character who has no separate plot looks, to a first reading, like a minor character. The film’s structure encodes her invisibility so faithfully that the structure itself is then taken as evidence that she is not the subject. This essay reads it the other way round. The reason she has no thread is the reason she is the subject: the position she occupies is defined by having no separate business of its own.

Whether that is the film making an argument or this essay making one is a fair question, and it is taken up directly further down.

This is not the first time this house has written about the terms of a household, and the earlier essays should be credited rather than quietly recycled.

This Library has argued, writing about a Japanese television drama in which a marriage begins as an employment contract, that an ordinary marriage already has terms — who cooks, who earns, whose career yields — and that what it lacks is a document, so the terms end up enforced by assumption rather than by agreement. It has argued, writing about the gap between being a good wife and being a happy woman, that the first has a specification with items and a grade while the second has none, and that a measurable goal beats an unmeasurable one for the same hours regardless of which a person values more. It has argued, writing about holidays and rituals, that a role is valued for appearing effortless, which means the labour producing it has to be concealed, and concealed labour cannot be thanked without exposing the machinery. And writing about Force Majeure, it argued that intimates have no procedure for settling a disagreement about what happened.

What this essay adds is a further turn of the same screw, and it is the reason the film was worth a commission. Those pieces described terms that were unwritten, goals that were unmeasurable, labour that was unthankable. This one is about standing. Megumi’s problem is not that the terms are undocumented, nor that her work goes unrecorded, nor that she cannot prove what happened. Her problem is that she has the facts, has the competence, has the whole arrangement resting on her, and has no position from which to propose a change to it. She is not short of information. She is short of a chair at the table where the thing is decided.

And the sharper half follows from that. Her competence is not incidental to her invisibility. It is the cause of it. An arrangement that is being held together well produces no evidence of being held. Nothing breaks, nothing is late, nothing is missing, and so there is nothing for anybody to notice. The better she is at the job, the less the job exists.

Before the argument is allowed to feel finished, the interest behind it has to be visible.

This house sells private companionship: paid, scheduled, bounded evenings in which a person has somebody’s attention. The reader most likely to be moved by this essay is a woman who recognises the position it describes, and that same woman is the person this house would like to have as a client. An essay arguing that such a woman is structurally unseen, and that her competence is the reason, is an essay that flatters our commercial position from beginning to end. That is true whether or not the argument is correct, and it stays true even if the argument is correct.

So the declaration turns into constraints, stated as refusals rather than as a disclaimer.

We do not claim that a reader in this position is unwell, deficient or in need of anything. Nothing here is a diagnosis. A person can recognise every paragraph above, want nothing at all, and close the page with the recognition and no purchase attached.

We do not claim that anything we sell alters a household’s terms. It does not. An evening of attention is not standing, and it does not become standing by being repeated. If somebody’s only hours of being attended to are hours she has bought, that is a description of a gap in an arrangement, and it is not a result this house should be pleased about.

We do not claim any outcome of any kind. Nothing here says that being seen for an evening relieves, repairs or improves anything. That would be the easy sentence to write and it is the sentence this house will not write, because it could not be supported and because unfalsifiable promises are the usual instrument for taking advantage of people who are tired.

And we do not claim that this film endorses anything. Neither the director, the writers, the cast, the production companies nor the distributor knows of this house or has any connection to it. The film was made for its own reasons, none of which are ours.

Four objections, and the first is the one that could sink the essay outright.

The first is that centring Megumi is an imposition. The film keeps her peripheral, and a reading that promotes a peripheral figure to the centre can always be constructed for any film by any critic with a thesis. The strongest form of this objection does not come from a rival critic; it comes from the director. Kurosawa has said that the elder son is the character he was most attached to, that the son is in fact the one pulling the entire family, that he cut the son’s material back because it was growing large enough to take over the picture, and that he abandoned an early plan to shoot the boy’s war. That is the filmmaker naming a centre, and the centre he names is not the wife. This essay cannot answer that by out-arguing him about his own film. What it can say is narrower: a director’s account of his intentions is evidence about the intentions and not a ruling about what the finished work turns out to be organised around, and the fact that the son had to be cut back because he would otherwise have taken over is itself an admission that the picture’s balance was engineered rather than discovered. That is a real reply, and it is weaker than the objection.

The second is that the night sequence cannot bear this or any weight. It is tonally so strange — a hostage-taking that turns into an almost comic confession by the sea and then into an image of tracks in the sand — that every available reading of it, including the one offered here, is a reading imposed on an unreadable passage. A viewer could reasonably call it a misjudgement, or a genre reflex the film could not suppress, or a joke. This essay uses it as the film’s single eruption of a want with nowhere to go, and that use is a choice, not a finding. Only one piece of external support exists for it, which is the director’s own statement of what the burglar was invented to do, and a statement of function is not a guarantee of effect.

The third is about the ending, and it is the objection this essay finds hardest to answer. The family reassembles at a table. The stolen money is returned. A letter arrives. The younger son plays Debussy at an audition and the room, and the film, is briefly at peace. Kurosawa has said that he wanted to infuse the ending with real hope rather than false hope, that the family is attempting to start over with all of its problems still in place, and that the father returns the money because he needs to be at zero. He has also said, of the final passage, that a kind of blessing arrives on this family in a form that goes beyond reason, and that the form is the music. That is a candid description of a consolation that is not argued for. By the film’s own logic nothing has been settled: the man has no position he can name, the elder son is gone, the terms of the household are exactly where they were, and Megumi has one more unspoken night in her. The music does the work that the plot has not done. A reader who concludes that the ending is unearned has the director half on their side.

The fourth is aimed at this house rather than at the film, and it is the one that should be hardest to hear. A publication that sells attention to women in Megumi’s position has an obvious incentive to find her invisibility beautiful. There is a real risk in an essay like this one of savouring the condition it describes: of turning an unpaid, unacknowledged and unchosen load into a theme with a pleasing shape, praising the woman for bearing it, and thereby doing on the page precisely what the household does at the table, which is to admire the smoothness while leaving the terms untouched. The elegance of the argument is not evidence that the argument is kind. The only defence available is not to claim that we have avoided this, because we may not have, but to say plainly that nothing in the position described here is admirable, that competence exercised without standing is a cost and not a virtue, and that a reader who wants out of an arrangement is not obliged to find anything in it moving.

This essay is a reading of a film. It is not clinical guidance, it diagnoses nobody, it promises no outcome and it prescribes no action.

Nothing here is a statement about Japanese families, Japanese marriages, Japanese women or any national character. The film is one work of fiction about one invented household, and the structural argument is about how a household can be arranged, not about where it is. The one national statistic used is a published government survey of reported time use; it describes averages across a population, it does not describe any individual home, and the household in the film does not fall within the category the cited table covers.

The awards record is stated exactly. The Cannes distinction is the Un Certain Regard Jury Prize in a parallel section, not the section’s top prize and not a Competition award. The Asian Film Awards recognition is two wins at a regional ceremony. The film received nothing at the thirty-second Japan Academy Film Prize, which is an absence in a published list rather than a positive judgement by anyone. Koizumi Kyoko’s acting prizes that season were awarded for two films together rather than for this performance alone.

The director’s views are quoted as his own and are not treated as settling the meaning of the work, and where they cut against this essay they are reproduced at their strongest rather than paraphrased into harmlessness. No passage of the screenplay is reproduced. Descriptions of what happens on screen are descriptions, corroborated against published synopses and festival records rather than against a frame-by-frame inspection, and dialogue is summarised rather than quoted.

The research cited establishes associations and definitions in particular samples and nothing more. Daminger’s study is seventy interviews with thirty-five couples and reports, against the drift of this essay, that decision-making participation was roughly equal. The Bittman study uses Australian and United States survey data from the 1990s and does not describe Japan or the present day. Neither is evidence about the film, and no causal claim is made from either.

Megumi is a character. Nothing here presents the inner life of any real person as established, and no claim is made about the private lives of the performers, the writers or the director. This house’s commercial interest in readers who recognise the position described is stated in the body of the essay rather than only here, because an interest disclosed after the argument has done its work is not a disclosure.