Force Majeure, and the Second They Cannot Agree On

A restaurant terrace at a ski resort in the French Alps. A Swedish family of four is having lunch in the sun. Somewhere above them the resort sets off one of its controlled avalanches, the kind arranged on a schedule so the mountain does not do it by itself later. The diners take out their phones and film it, because it is a spectacle. Then the powder front keeps coming, and arrives at the railing. People scream. The father, Tomas, takes his phone and goes. The mother, Ebba, stays where she is, over the two children. The cloud passes. The tables are still there. Everyone is fine. Tomas comes back and sits down, and lunch, in some technical sense, continues.

This essay is about Force Majeure, released in Swedish as Turist in 2014, written and directed by Ruben Östlund. It describes the premise openly and discusses the rest of the film, including how it ends, though the final sequence is handled at the level of what it accomplishes rather than narrated. A reader who would rather meet the film cold should come back afterwards.

The obvious reading is available immediately, and most of the film’s reputation rests on it: this is a film about male cowardice. A man was handed the one task the culture still reserves for him and failed it in front of his own children, in under five seconds, while taking his telephone. That reading is not wrong. It is simply not what the remaining two hours are doing. If cowardice were the subject, the film would be finished at minute twenty, and it is not.

What the rest of the film is about is this. Tomas says he did not run. Ebba says he did. Nobody is injured, nothing was lost, there is no third party and no money and no lie about sex — the film has swept every other possible object of dispute off the table, deliberately, and left exactly one thing on it, which is the description of an event that both people witnessed. The marriage is not damaged by what Tomas did. It is damaged because the two of them cannot produce a single account of it, and because there is nowhere to take that. This is a far more common injury than cowardice, and almost nobody names it.

Force Majeure was released in 2014 as a co-production of Sweden, Denmark, France and Norway, written and directed by Ruben Östlund, with Johannes Bah Kuhnke as Tomas, Lisa Loven Kongsli as Ebba, Clara Wettergren and Vincent Wettergren as their children Vera and Harry, and Kristofer Hivju and Fanni Metelius as the visiting friend Mats and his girlfriend Fanni. Its Swedish title, Turist, means tourist.

It premiered at Cannes in 2014 in the Un Certain Regard section and won that section’s Jury Prize, which the festival’s own record states as the Jury’s Prize, Un Certain Regard 2014. At Sweden’s 50th Guldbagge Awards, held in January 2015 for the 2014 year, it took ten nominations and six awards, among them Best Film, Best Director, Best Supporting Actor for Kristofer Hivju, Best Cinematography and Best Editing. It was nominated for Best Foreign Language Film at the 72nd Golden Globe Awards.

Sweden entered it for Best Foreign Language Film at the 87th Academy Awards. In December 2014 it was named one of nine films advanced from eighty-three eligible submissions to the shortlist. In January 2015 it was not among the five nominees. That is the whole of what happened with it, in a year when a good deal of the press had treated it as a probable winner.

In Japan it was shown at the 27th Tokyo International Film Festival in October 2014 as Tsurisuto, and released generally on 4 July 2015 as Furenchi Arupusu de Okita Koto, which translates roughly as what happened in the French Alps. An American remake, Downhill, directed by Nat Faxon and Jim Rash and starring Julia Louis-Dreyfus and Will Ferrell, appeared in 2020.

One more item belongs here because it matters later. After the nominations, Östlund and his producer Erik Hemmendorff released a video of themselves watching the announcement and reacting to the omission, in which Östlund performs an extended and extremely undignified collapse. Writing about it in January 2015, Filmmaker Magazine noted that the joke, if it is indeed one, depends on the viewer having seen the film. Whether the distress is real is not decidable from the video, which is the same problem the film spends two hours on, reproduced by its director in public.

The avalanche scene is filmed wide, from a fixed position, in a way that lets the audience see the whole terrace at once. There is no cut to protect anybody and no close-up to tell the viewer what to think. The powder front is beautiful for a long time and then it is not. The scream begins. And in the middle of the frame, plainly, a man gets up and leaves and a woman does not.

So the audience is given something the marriage will never have: an unambiguous view of the event. The film does this on purpose. It wants the viewer holding a settled account while watching two people fail to arrive at one, because the gap between those states is the subject.

Then comes the second piece of construction, the removal of everything else. No one is injured. Nothing is destroyed. There is no infidelity, no debt, no illness, no third person. The avalanche is not even a real disaster; it is a scheduled detonation carried out by the resort for safety, announced by rumbles that punctuate the whole film like a clock. The mountain has a procedure. It is managed. The catastrophe that arrives on the terrace is manufactured, controlled and over in seconds; the one that arrives in the hotel room has no ski patrol and no schedule and does not end.

What is left, with all of that cleared away, is a couple and a description. That is an unusually pure experiment, and the film runs it at length.

The cowardice reading goes like this. A modern father is still, somewhere underneath, graded against a very old job description: he is the one who stands in front. The film puts that job to a five-second test with no preparation, he fails it, and he spends the rest of the week discovering that his standing in his own family was resting on a thing he cannot do. It is a comedy about the collapse of a masculine idea, and it is often very funny. It has support from the director, who has spoken about what research on real disasters shows regarding who actually survives, against what film history has trained audiences to expect of men.

But watch what the film does with its running time. The run occupies seconds. The rest is Tomas saying, calmly and repeatedly, that this is not what happened — that he did not run, that a person cannot run in ski boots, that Ebba experienced it differently and both experiences are valid. And it is Ebba being unable to proceed to any other topic, room or hour of the holiday until he stops saying it.

Notice also what the cowardice reading offers a viewer, which is a verdict. It gives the audience a job: assess the man, find him wanting, enjoy having assessed correctly. The film keeps refusing to let anyone finish that job. Every time the audience is ready to convict, it hands them something that complicates it — the friend with a theory, the wife whose insistence begins to look like something else, the children who are not consulted and are the only genuine casualties in the room. A film that wanted to indict a coward would let the verdict land.

The most useful thing to notice about Ebba is what she is actually after, because it is not what the argument looks like from outside.

She is not asking to be apologised to. She does not seek a punishment or a promise about future avalanches; there is nothing to promise, since the situation will not recur. She is asking him to say that he ran. That is all. A short sentence, describing something that took under five seconds and harmed nobody.

This Library has argued before that an apology matters less than what becomes different after it, and that the confusion is durable because an apology can be witnessed by everybody while a change is visible only to the person who was hurt. That argument holds, and this film sits one step prior to it. Before the question of what will change, there is a question of whether the two people are describing the same event. Where that is unanswered, nothing built on top of it holds, because every later negotiation must begin by re-litigating the earlier one. A shared account is not a nicety; it is the substrate. Two people who agree on what happened last Tuesday can disagree productively about almost anything else. Two people who do not have no common ground to place a disagreement on, so every argument, whatever its declared topic, becomes the same argument in different clothes.

And there is something worse in Tomas’s position than lying, which the film is careful about. He does not behave like a man concealing a known fact. He behaves like a man who has arrived at a version of events he can live inside, and is defending it with the ordinary sincerity of someone defending the truth. A liar can be caught, and catching him ends the matter. A person who has genuinely rewritten the file cannot be caught, because there is nothing to catch. He is not withholding the sentence. He no longer has it.

Here is the observation this essay exists to make, and it is structural rather than psychological.

Every other domain in which two accounts of one event collide has machinery for ending the collision. A court has rules of evidence, an allocation of the burden of proof, a decision-maker with standing, and a judgment that binds both parties whether or not either is satisfied. A workplace has an investigation, a finding and a file. An insurer sends an assessor. A match has a referee and, now, a replay. None of these reliably produces truth. What they reliably produce is closure: a moment after which the matter is, institutionally, over.

A marriage has none of this. There is no neutral party with standing, no rule about who must prove what, no mechanism that ends the matter over one party’s objection. The available outcomes are three: one person concedes, both agree to stop raising it, or it stays open.

It stays open. That is the ordinary outcome, and it is why this is, by a wide margin, the most common unresolved injury in long relationships and almost never the one that gets named. People say they argue about money, or about his mother, or about the housework. Very often what they are doing is running one unsettled description through new material, year after year, because it was never settled and there was no way to settle it. An unsettled description does not sit still; it attaches itself to whatever is available.

This is a claim about how the arrangement is built, not about any reader. It does not say that couples are failing at something more skill would fix. It says that what they are trying to do — reach a binding account of a contested event using only two interested parties — is something no other area of life expects anyone to manage without help, and that the absence of help is invisible precisely because it is total. Nobody notices the missing procedure, because there was never a version of intimacy that had one.

Which brings the film to its most excruciating scenes, and to the reason they are excruciating, which is not the reason usually given.

Ebba tells the story to other people. First to another guest at the resort, then, at greater length and with more damage, over dinner with Mats and Fanni, who have come to visit. She sets out what happened on the terrace. Tomas offers his account beside hers. The room is thereby placed in the position of a jury nobody has sworn in, asked to weigh two versions of an event none of them saw, by people whose marriage they came to have dinner with.

Understand why she does it. Given that there is no procedure, the only instrument left to a person whose account is being overridden is a witness. Saying it in front of others makes a private discrepancy public, and a public discrepancy at least exists; it cannot be waved away in a hotel room at eleven at night. That is the whole mechanism, and it works in the narrow sense that everyone at the table can feel the gap.

And using it is itself a violation. If a marriage has any unwritten term at all, it is that its disputes stay inside it. To take the disagreement to dinner is to conscript guests, to corner a husband in front of a friend, to convert a private failure into a social event. She does it because she has nothing else. That is what makes the scenes hard to sit through. It is not embarrassment on Tomas’s behalf. It is that both people are, at the same moment, entirely right: he is being ambushed, and she has no other tool.

Then the film performs its sharpest structural joke. Eventually Tomas agrees to play the footage from his phone, and it shows a figure running. The evidence a couple never has is produced, on the table, in front of witnesses. And nothing is repaired. Mats immediately supplies an alternative interpretation — perhaps he was going to get help. Tomas says nothing. The dinner does not end in a finding; it ends the way dinners end.

This proves the point better than any argument could. The dispute was never evidentiary. Evidence is what you need when the question is what occurred. The question here was never what occurred. It was whose description of the two of them gets to stand, and there is no footage of that.

This Library has written about another marriage meeting a fact. In an earlier essay on Andrew Haigh’s 45 Years, we argued that nothing in that marriage changes and everything in it changes, because one new item of information reopens forty-five years of accumulated evidence to a second reading; that the wife’s crisis is epistemic rather than romantic, because reassurance about love does not answer a question about description; and that the damage is done by what was never said, which has no date and therefore cannot be apologised for.

The pairing with Force Majeure is exact in one respect and divides sharply in another, and the difference is why both essays exist.

In 45 Years the fact is verified and old. A body in a glacier, a letter from an authority, half a century of distance. Nothing about it is in dispute; what is in dispute is what it means. And meaning is precisely what no witness can supply, because the only competent witness is a man who existed fifty years ago and is not available, including to himself.

In Force Majeure the fact is recent and contested. Both people were there. Both saw it. Within hours they have produced incompatible descriptions of five seconds, and the competent witness is sitting across the table, alive, in good health, and declining. Ebba’s problem is not meaning. It is standing: whether her account of an event she personally witnessed will be permitted to be the account.

That changes what is even theoretically available. Kate cannot be answered. Ebba can be — which is what makes her week unbearable rather than merely sad. The sentence she needs exists, is short, costs nothing material, and is within the gift of a man three feet away. Its unavailability is not tragic the way a dead witness is tragic. It is a refusal, and refusals can be maintained indefinitely at very low cost to the person maintaining them.

This Library has also argued, reading Esther Perel’s The State of Affairs, that after a betrayal the first requirement is naming what happened, and that curiosity about why, offered before the naming is complete, lands as premature empathy for the person who broke the agreement. Force Majeure is the case where the naming never completes at all. It shows that the sequence can stall at its first step indefinitely, without a betrayal, without anybody doing anything dramatic, and with both parties still in the room.

Östlund has said in interviews that part of what set the film going was research he had read on disasters. The claim comes in two parts and they deserve very different treatment.

The first is about hijackings. Speaking to Film Comment in 2014, and again in a filmed conversation archived by the Barbican and published in 2020, he described having read investigations into aircraft hijackings from which, he says, one can tell that the frequency of divorce afterwards is extremely high. He also told PopMatters in November 2014 that one of his goals with the film was to raise the percentage of divorce in society, which is plainly a joke and equally plainly a statement of intent.

We could not source the hijacking claim. We looked for a published study of divorce among couples who survived aircraft hijackings and did not find one. What we did find runs the other way, and is worth reporting because it is checkable. Cohan, Cole and Schoen, in the Journal of Social and Personal Relationships in 2009, examined monthly divorce filings before and after the September 11 attacks in New York City and in four counties at varying distance, and report that filing rates decreased in New York City and in the comparison locales. That is one study of one event, and its subjects are not hijacking survivors, so it does not refute him. It does establish that disaster raises divorce is not a settled finding.

The second part can be traced. In the same Film Comment conversation he refers to work on maritime catastrophes, from the Titanic to the Estonia, showing that men of a certain age are the ones who survive. There is a real paper in that territory. Elinder and Erixson, in the Proceedings of the National Academy of Sciences in 2012, analysed eighteen maritime disasters spanning three centuries and covering more than fifteen thousand people of over thirty nationalities, and report that women have a distinct survival disadvantage compared with men, that captains and crew survive at a significantly higher rate than passengers, and that behaviour in life-and-death situations is best captured by the expression every man for himself. That paper says approximately what he says it says. Whether it is the paper he read, we cannot know.

The general point is worth stating without pointing it at anybody. Statistics of a particular shape — a vivid population, a dramatic rate, no citation attached — circulate because they are good to think with rather than because anyone has checked them, and they are repeated most enthusiastically by people using them as a premise rather than as a finding. A filmmaker is entitled to a premise. What should not follow is the belief that the film then constitutes evidence for it. Force Majeure is not evidence about hijacking survivors, and the statistic is not evidence about the film, which stands or falls on whether its argument is recognisable.

Three objections stand, and they should be built at full strength.

The first is about method: Östlund is cruel to his characters for comic effect. He holds shots past the point of comfort and constructs situations whose function is to trap a person in a humiliation and keep the camera running while an audience laughs. The clearest exhibit is the long sequence in which Tomas finally breaks and weeps — an extended, undignified, physically grotesque collapse in a hotel corridor with his children clinging to him. It can be called unflinching. It can equally be called a director arranging a pillory with very good cinematography and selling tickets to it. The defence available is that the film makes the audience’s laughter part of what it examines, so the cruelty is legible as method rather than taste. That defence is real, and it is not available to a viewer who did not laugh, and for them the objection stands undiminished.

The second is about distribution, and it is the one this Library finds hardest to answer. The injury belongs to Ebba. The interiority belongs to Tomas. He gets the breakdown, the confession, the rescue in the fog, the arc from denial through collapse to some kind of restoration. She gets to be correct, repeatedly, which is not the same as being written. The film’s most sustained access to an inner life is granted to the person who caused the problem — a very old distribution of narrative resources, and pointing out that the film is about a man’s failure does not excuse it, because the failure is only interesting through its effect on her and the film keeps declining to go and look.

The third is aimed at this essay, and it is the serious one. It runs: by reframing this as a problem about incompatible accounts, you have performed exactly the manoeuvre that gets performed on behalf of men. A man abandoned his children and then denied it. Moving from he failed to they disagree about what happened converts a specific and gendered failure into a symmetrical epistemological puzzle between two equally situated parties — and they are not equally situated. Ebba is not merely a person holding a competing account; she is a person whose account is being overridden. That she must call witnesses at dinner is itself evidence that the default distribution of credibility does not favour her. A de-gendered reading of this film is therefore not neutral. It is a reading that happens to be more comfortable for the party who ran.

That objection will not be resolved here, and this essay will not pretend to resolve it. Both things appear to be true at once. The structural point is the more general one and the more useful to a reader whose own unsettled thing has nothing to do with avalanches; it really does operate between two people of any configuration, and naming it gives a person something to do other than assign blame forever. And the structural point is also, in this case, a description that lets a particular man off a particular hook, produced by a reader who found the general question more interesting than the specific verdict — itself a position with a politics.

What can be offered is a limit rather than a resolution. Holding the structural reading does not license anyone to stop noticing who ran, who denied it, whose word carried further at the dinner table, and who had to breach the terms of her own marriage to be heard at all. If the general point can only be had at the cost of those particulars, it is not worth having. The position taken here is that it can be had alongside them, and a reader is entitled to think that convenient.

The interest should be declared before the closing rather than after it.

This house sells private companionship: paid, bounded, scheduled evenings in which a person is listened to with attention. Every step of the argument above therefore runs in our favour. If being accurately heard is scarce, we sell a scarce thing. If there is no procedure between intimates, we are adjacent to an unfilled gap. That we also believe the argument does not cancel that.

So the constraints, stated as refusals. We do not claim that a bought evening settles anything. There is no witness here with standing, no finding, no file, and nothing said in this house adjudicates anything said in anybody’s marriage. We do not offer couples work, mediation or counselling of any kind, and we are not qualified to and do not pretend to be.

And we will not blur the most tempting line. Being paid to accept somebody’s account of an event is not the same as somebody conceding it. The first is a service; the second is a relationship doing something difficult, and anybody who tells you they are interchangeable is selling. The most an evening here can honestly offer on this subject is that a person gets to say the sentence out loud once, in order, without being contradicted while saying it. That is a relief. It is not a settlement, it does not substitute for one, and it should not be sold as though it did.

This is a reading of one film. It is not advice, not therapy, not couples guidance, and not a diagnosis of anybody.

No psychological condition is assigned to either character. Tomas is not described here as suffering from anything, and neither is Ebba. They are figures in a screenplay written by one man, and treating them as case material would be bad criticism as well as bad ethics.

Nothing is claimed about marriages in general, about men in general or about women in general. No proportion, prevalence or rate is asserted anywhere about how often couples fail to agree on a description of an event, because none was consulted and this Library does not believe a reliable one exists. The claim that this injury is common is a claim about what appears recognisable, offered as an argument to be agreed with or rejected, not as a measurement.

The factual statements are limited to what is recorded in this essay’s evidence ledger: production year and countries, writer-director, principal cast, the Cannes award, the Guldbagge outcome, the Golden Globe nomination, the Academy Award submission and shortlist and the absence of a nomination, the Japanese festival screening and general release, and the existence of the American remake. Each is taken from a festival, industry, encyclopaedic or distributor record read for this essay, not from an inspection of any original document.

Where the director’s views are described, they are attributed to published interviews and to what he said in them, and are not presented as the meaning of the film. What a filmmaker intends and what a film does are separate questions, and this essay answers only the second.

The statistic about divorce after hijackings is reported as something the director says he read. This Library could not locate a published study supporting it and therefore does not assert it. The maritime survival study and the September 11 divorce-filing study are cited as what their abstracts report, each being one study with its own sample and its own limits, and neither is offered as evidence about any individual or any marriage.

No passage of the film’s dialogue is reproduced here, and the plot is described rather than transcribed. No filmmaker, performer, production company, distributor, festival, researcher or publication named in this essay is connected to this house, knows of it, or endorses it.