Teller of Tales: Mariano Llinás on La Flor

By Jordan Cronk

Some will say I reinvented the wheel. Yes, I’d say, I reinvented the wheel.”La Flor, Episode 4

To begin, a question: What exactly is La Flor?

It’s a pertinent query, albeit one with no easy answer, so let’s break it down. The first thing to know about La Flor is that, yes, it’s a 14-hour-long film—868 minutes to be exact, including intermissions and what must be a record breaking 40-minute end-credits sequence. But these talking points also obscure much of what makes Mariano Llinás’ ten-year passion project such an approachable—dare I say accessible—feat of populist filmmaking. For starters, La Flor is not a film in the traditional sense: rather, it’s a number of films, all but one of which happens to be rather traditional. Comprised of six discrete episodes, La Flor is what we might call an anthology film, in that a number of individual narratives, each working in different genre registers, have been compiled under one banner. Except that, unlike more familiar multi-director efforts, just one person is responsible for La Flor’s extraordinary stories. Add to that the actual conceit behind the endeavour—that the same four actresses would appear in different roles in each episode—and you have one of contemporary cinema’s great collaborative undertakings. Needless to say, following La Flor’s prize-winning bow in BAFICI and its bold placement in the International Competition at Locarno, its reputation precedes it.


The first person we meet in La Flor is Llinás himself. Seated outdoors at a picnic table, leafing through a notebook and sketching the film’s flower-like symbol, he begins to describe to the viewer the film’s six-part structure and the concept behind its construction. His description sets the template in matter-of-fact fashion: over the course of 14 hours we’ll be seeing a B movie, a musical (“with a touch of mystery”), a spy movie, a remake of an old French film, and a parable about captive women in 19th-century South America. The other is difficult to describe, he says, but for the sake of brevity let’s call it a reflexive fiction (also with a touch of mystery, as it happens).

Llinás has been candid about approaching La Flor as an opportunity to learn and rediscover bygone tools and techniques, so it makes sense that he would look to genre to help facilitate the process. Key to the success of the project is his commitment to each genre—not their aesthetics, per se, but the way they operate on a formal level, and how they unconsciously reverberate in a sociohistorical context. These aren’t satirical revamps of old-timey trends: the only suggestion of irony in the film occurs when Llinás periodically appears between episodes to offer self-aware commentary. Which may explain some critics’ baffled reaction to the tone and full-tilt action-adventure machinations of much of the film’s first three episodes. After all, for all the playful digressions of Historias extraordinarias (2008), that film maintained a literary scope and imagination that signalled a clear thematic integrity. How can a mummy movie featuring a cannibalistic cat possibly be sincere?    


The four actresses enlisted to anchor this epic undertaking come from the world of Argentine theatre, and together comprise the troupe Piel de Lava. Most are relatively unknown internationally, though one, Laura Paredes, has, over the course of the film’s extended gestation, made a name for herself in the films of Matías Piñeiro. (Another Piñeiro regular, Agustina Muñoz, plays a prominent role in the spy episode.) The others, Elisa Carricajo, Pilar Gamboa, and Valeria Correa, exhibit a range and versatility to confirm them, too, as unique talents. Together, the four navigate each episode with a verve and authority rarely glimpsed in the more forwardly formal and rigorous cinema of Argentina’s most visible arthouse exports.

The sheer variety of roles on display—as well as the sometimes dramatic transformation each actress undergoes from episode to episode—is invigorating. Witness, for instance, Carricajo’s shift from an emotionally vexed scientist in Episode 1 to a steely-eyed spy leader in Episode 3, or Gamboa’s turn in consecutive chapters from talkative torch singer (whose songs are performed in the film by producer Laura Citarella) to mute assassin. In the ’80s-set Episode 3, their most dazzling collective showcase, the four make the most of the extended runtime (a full 342 minutes on its own), cross-continental espionage plot, and nesting-doll narrative with an array of fine-tuned Cold War characterizations. In his opening remarks, Llinás declares La Flor a film by and for his actresses; in a beguiling coda that ends Episode 4, featuring all four women as “themselves,” candidly playing for the camera, these words have once and for all been made flesh.


Beginning in Episode 4, the actresses play a far less concrete role in the proceedings. By the end of the episode, during which they’ve variously played Canadian Mounties, witches, and Casanova’s concubines, they’ve become something like ciphers, and by the final chapter literal abstractions. (They don’t appear in Episode 5.) Following a trio of stories that worked to update classic narrative tropes, Llinás casts a wider net for the remaining three. In Episode 4, a meandering metafiction following a wayward filmmaker as he drags his exasperated crew from Québec’s [sic] Kashwakamak Lake to the Caribbean islands off South America in search of the perfect pink trumpet tree to film, Llinás skewers a certain de rigueur trend in documentary filmmaking wherein perspectives shift, reality intrudes, and art ultimately prevails; in Episode 5, he blatantly pillages Partie de campagne (1936), forsaking his actresses for a silent, black-and-white remake of Jean Renoir’s lyrical country romance; and in Episode 6, the diary of an Englishwoman held captive by Indians is paired with deeply impressionistic compositions, shot with a camera obscura, that shift the film’s final moments into the realm of feverish allegory.

But more than any one narrative tradition it might employ, La Flor is above all a film about the art of storytelling itself, its lineage and its ongoing utility. Llinás understands how fictions evolve and how they are disseminated, how they operate and how they can be translated across cultures. Only the grand finale has a traditional ending; the others stories are variously abridged or outright abandoned. La Flor is not an attempt to redraw the boundaries of storytelling, but rather a creative experiment in rethinking, on a moment-by-moment basis, the nature and capacity of storytelling in a contemporary context.

Language, Lineage, Legacy

Inasmuch as La Flor is about traditions related to cinema and storytelling, it’s equally about the formal grammar and linguistic idioms of each: in short, language. Beyond Llinás’ command of genre lies a flair for words, speech, dialect, and vernacular that lends the film an uncanny internal rhythm and propulsion. Likewise, his inventive use of framing devices—for instance, the extended rounds of multilingual voiceover narration (a callback to Historias extraordinarias) that mark Episode 3, the missives that bridge Episode 4’s many forking paths, or the transparent intertitles that quietly punctuate Episode 6—each present unique new avenues to receive, explore, and think through the nuances of language.

Like many films made under the El Pampero Cine banner, a collective that has worked for over 15 years to rethink the bounds of independently produced Argentine cinema, La Flor—which by any measure must be considered the group’s grandest achievement to date—is a work in constant dialogue with history. Its lifeblood is the cinema. To hear Llinás tell it, the film is nothing less than an attempt to remake the medium in its own image. But it’s also something more precious than that: it’s a document of five lives told through cinema. Time itself is bound up in the very idea of La Flor, and it generously extends the freedom and possibilities afforded by giving time back to the viewer. Rare is the film that so casually and sublimely carries the weight of history and accumulated experience.

Cinema Scope: Take us back ten years. You finish Historias extraordinarias—what’s next in your mind? I can’t imagine it’s a 14-hour film.

Mariano Llinás: You’re right, it was not a 14-hour film. In the beginning it was the four actresses. I’m not sure if it’s right to call it an idea, but I had the will to work with these four women. I had met them a couple of years before the shooting of Historias extraordinarias, and at that time I was quite concerned about fiction. I was preoccupied by it—completely obsessed with the subject of fiction itself. It was around this time, in 2006, that I saw the girls in a play called Neblina—“mist” or “fog”—and I was astonished. I remember thinking that these four can take fiction to new places.

This was still the beginning of the new century, and fiction was in trouble—it was endangered. During that time everyone was talking about the boundaries between documentary and fiction, and people still are. But nobody was thinking about or finding the pleasure in fiction, nor in the real. Everyone would say, “I’m not interested in telling regular or traditional stories,” which is an idea I believe in, because I do think storytelling is plagued. But the cure for storytelling wasn’t the one that people were trying, which was a…I don’t want to say slow, but a dispassionate form of cinema lacking in intensity. I was thinking that, if used correctly, nothing can be as emotional as fiction. We all know and try to follow the Samuel Fuller quote in Pierrot le fou (1965): “Film is like a battleground.” But I thought that leaving fiction out of the cinema was not the right way to go against the obligations of storytelling. I was thinking that maybe fiction and storytelling are not the same things, and you can make strong, rich, extreme fantasies without falling into cheap storytelling, or toward moralizing or wherever storytelling often takes us.

So that was my thesis. And when I saw the girls, I knew they could do that—they could take us places we’ve never been, by just being a tool for fiction. They had the same questions, the same obsessions: the questions of fiction were just as strong in their work as they were in mine. And they came to me with an idea to do a film based on this same play that they were in, and I said yes without thinking, just because I wanted to work with them. We tried to make something out of this idea for two years, meeting every Thursday. It was then that we fell in love, let’s say, speaking of fiction. If you saw one of those meetings you would look at us and think these people are in love. And though the film based on this play wasn’t working out, we kept on because we wanted to be together. So we tried out different ideas, different stories, and they all seemed fine to me—I thought even the most clichéd stories could work if they were doing them. It was then that I realized that what I wanted was not one picture: I want to make hundreds of pictures with them! I want to make pictures with them a genre in itself. So that’s when I had the idea to make one picture that would be all the pictures—every picture.

And then came the drawing that you see in the film, with the four arrows going up, a circle in the middle, and another arrow going down. For me that meant four beginnings, one complete story, and one ending. This symbol became the scheme. There’s a Leonard Cohen song that says, “A scheme is not a vision.” In this case the scheme was a vision! I thought, “Well, this is new, isn’t it? This could work.” Four beginnings with no endings that would lead one to the other, followed by a complete story and an ending. Why not? And in each story the actresses would be playing different roles. This could be a way to give each of them a proper showcase. Each of these actresses has many skills—one picture would not suffice. We needed to see and understand the switch between fictions and how they turn from one emotional state to another, working non-psychologically.

Once I realized the six stories could work, I wrote the first one, a dumb story with mummies, very quickly. It was like throwing the dice. But when the stories came to me, I immediately knew which ones would work. As you probably know, there aren’t many mummy pictures in Argentina. We don’t have a mummy tradition like Hollywood. So it would be an idiotic, or not very serious, story. I know I’m talking about Pierrot le fou and all that, but that’s what first came to me: mummies!

Scope: At one point did you realize that each episode would also be in a different genre?

Llinás: From the very beginning. That was part of the game! I like classic cinema, Hollywood pictures, French pictures—I don’t see many new pictures. I don’t really know contemporary cinema—or rather I see many popular films, but not many of the good ones. I’m very ignorant in that regard. Most of what I see nowadays is made before 1970. These are the films I watch and the films that I teach.

Scope: So the idea would be to work with older traditions familiar from these classic films?

Llinás: Yes, that would be the idea, but genre would be a way of reaching the older traditions. With all due respect, when critics think of old pictures, they sometimes only think of the genres. I think genre was and still is a way for filmmakers to get to something we like to do. It’s not that we like working with genres, which might entail a scene with horses, or a car chase, or a pursuit. For most filmmakers—the good ones at least—it’s not that they want to make, for example, a Western, and then have to shoot a horse scene or a battle scene to fit the genre. Instead, we have the idea to shoot a horse scene, and that requires the Western genre. It’s the opposite: first we have the idea for the image, then we write.

So I wanted to learn some old tricks that are not so current in contemporary cinema: ways of cutting a picture, with regards to continuity, for example, or different ways of working with sound. So many filmmakers work with sound in the same way. That droning bass sound you hear in so many current films—you’d think David Lynch invented that sound. Every picture sounds the same. For older filmmakers, the sound of their films was very precise and distinct. There’s an Argentine filmmaker, Hugo Santiago, a master to me; he was like my father of cinema. He was Bresson’s apprentice: the way he and Bresson thought of sound, the very precise, delicate way they worked with sound, this was not how I was working with sound. I wanted to learn, and the genres would take us there. Also, costumes: I wanted to make a picture with costumes and colour.

Scope: So you were setting challenges for yourself?

Llinás: Challenges, maybe, but not exactly. It was more like I wanted to give a gift to myself—pleasure, something restorative. It was more like that—exploring regions that have been forsaken by the cinema. And of course each of these stories would help the actresses develop and showcase what they can do. The genres were a way of finding new images.

Scope: How involved were the actresses in the conception of their various characters? Were you writing everything and giving the scripts to the actresses, or was it more collaborative?

Llinás: It wasn’t one way. Early on, I was collaborating with one of the actresses on some dialogue and she eventually said, “C’mon, you should be writing the dialogue.” The second episode has these extremely long speeches—so long that one of them said, looking back, “I can’t believe I did that, remembering all that precise dialogue. I was so young.” And she’s right: we were young, we still had the energy to sell that extreme soap-opera register. So that of course was all very written, and I wasn’t flexible with the words. I had an ear for how I wanted that dialogue to sound. But the following parts were dictated by the conditions. The spy film took very long to shoot. We started that in 2011 and finished in 2016. It was very difficult. One scene could take many years.

Scope: Were you finishing one entire episode before you even began writing the next one, or was there overlap?

Llinás: Never. Maybe we would reshoot a little after, but otherwise there was no overlap. And we didn’t begin editing the episodes until much later. For the spy film, the script was like a novel; the voiceover is almost literary. But since it was shot over such a long time, that allowed me the opportunity to be more loose with the players. The actresses are speaking French, which they didn’t know, so they’d be repeating the lines back and forth with someone who spoke French. It was like a Fellini film: they were reciting words that they didn’t always understand.

Scope: One of the primary concerns of the film as a whole seems to be language, and not just linguistic language but cinematic language: how each can be conveyed and interpreted.

Llinás: I’d say yes, that’s correct. There are things in certain episodes where I’m not sure that people who don’t speak the language can understand. At one point in the first episode, Laura’s character, the scientist, stops speaking Catalan completely. She starts speaking very fast, like a Spanish B-horror movie dialect. It’s like a whole other voice. It’s frustrated some Argentine viewers. Some of the voices of the American characters, the soldiers, are taken from YouTube. And later another character’s voice was made using Google Translate, so it’s a kind of mechanical voice. When you see Argentine pictures, for example, you can get carried away by the images, which can be interesting, but then the characters speak, and they all sound the same. A lot of filmmakers can deal with images, but in the sound department they just obey the rules: how the players should act, how they should sound, how their voices should be mixed by the sound engineers. Maybe this care for language is a tool to find new sounds. I want my pictures to sound a particular way, to work with the music. I wanted the spy film to sound like a ’70s French thriller, like those Alain Delon films—not the Melville ones, which are very good, but the cheap, B-grade Melville pictures like those by Henri Verneuil.

Scope: As far as writing and casting the individual parts, were you having a bit of fun shuffling the actresses between drastically different roles?

Llinás: Yes, but I’d also say we were not working in a psychological way: we hate psychology. For us the characters are an appearance, a sound, and a way of looking. They are exterior things. And I know saying that is heretical to some people. Nowadays there’s so much talk—people like to talk about their characters, but they talk too much before making the picture. For us it was as easy as, “Her character was good in the first picture, and now she’ll be bad, and now she’ll be sad, and now she’ll be romantic, and now she’ll be the leader, and now she’ll be English!”

But it was all part of a greater composition. I don’t paint, but as an example I imagine the masters saying to themselves, “I used some red up here, and now I need some down here.” Every decision informs things in the rest of the composition. If one of the characters would be more decisive or strong in one episode, it means that the others cannot be as strong, but also that they can be strong in the next episode. It became a joke for us—and maybe a punishment for her— that Pilar spoke so much in Episode 2, but then played a mute in Episode 3. Pilar is so skilled, but we were like, “Maybe it’s time for you to shut up.” And she was like, “Well, I’ll make the best mute ever!” So there’s an equilibrium.

As Argentine filmmakers, we live amongst players from the world of theatre. In Argentina the boundary between cinema and theatre is slight, both activities are closely related, maybe because many filmmakers started working so closely with theatre people.

Scope: Like Matías Piñeiro, for instance?

Llinás: Right. That’s really what his films are about. In fact, almost every actress from Matias’ pictures appears in La Flor. María Villar wanted to work on the picture, so we had to invent a character for María! Which is a good example of how we would invent characters for people, not the other way around. Flaubert’s notion of the right word, finding the right word for one idea: this idea translated to cinema would result in the “right image.” Humbly, I would disagree. I don’t think we need to find the right image for each idea. We have to find the right story for each image. The images are there. We need to invent the stories and make them fit. These games are one of the reasons the film is so long!

Scope: Once each actress had her character for a specific episode, was there ever a need to move one into another role for any reason?

Llinás: No, but that’s because we conceived of the characters specifically for each actor, not just the actresses. Now that I have finished this picture, and having felt the making of this picture as a kind of school, where I learned a lot—constantly thinking with images—that I feel like my next picture will be the same, as far as the application of what I’ve learned, but also something different. One of the things that I’ve been constantly considering over these ten years is how pictures are written before they are shot. Of course, I’ve done that—everyone does it. But I can’t find a reason why pictures are written and then shot. It should be the opposite: the images should come first, and then be organized. Why are we writing words for images we don’t have? It’s ridiculous.

Scope: Is Episode 4 an attempt to work out this issue cinematically?

Llinás: Episode 4 was the switch.

Scope: Was there actually no script?

Llinás: Some parts I had. But we did wander around as you see in the film, and the pink trees, the trumpet trees, were things that just appeared to us, documentary-style—our own version of what we were talking about before, the thin boundary between documentary and fiction. It was the same with the spy episode. Of course I knew how the shoot was going to go—it was not improvised—but why am I supposed to write a generic line or a scenario beforehand? It’s ridiculous to write those things if I have the people there with me on set. The sad truth is that it’s because you need it to convince producers, to get the money; that’s the only reason. In Histoire(s) du cinéma (1988-1998), Godard says that he wishes to make a film with Dolores del Río and Tyrone Power that would take place in the South Seas and that would end with them jumping into a volcano. That’s it—and that’s the point. Nobody’s going to remember anything except Dolores del Río jumping into the volcano. So that’s what I came to learn and I tried to work that way. We were just looking at the world and shooting.

Scope: How did you fund such an epic project?

Llinás: Our way [at El Pampero Cine] is really not so different than a lot of Argentine pictures. I think what’s maybe different is what we’ve done with those funds. We believe you can do anything you want if you work outside the industry. The thing that limits you is the industry—the system. When we noticed that, even before Historias extraordinarias, we started working not only away from the industry, but against it. I don’t know how it works in the US, but here we have the “weeks” system. This applies to the shooting, which is one aspect of the production. Once the scheduled weeks of shooting and budget are set, you can’t have extra days. And that of course would not work for a film like this. We shot this film whenever we could—we didn’t have to buy people’s time, to keep them from earning money elsewhere. We approached it like a hobby, something we do because we want to. So once you remove the shooting from a set schedule and budget, that changes everything: you can search, try new things, experiment. Howard Hawks worked the other way, as did Hitchcock. So it can’t be all bad. But the point is, it’s not necessary. Nowadays it’s not that you need a Mitchell camera that requires six people to haul from one place to another. I can carry the camera I shot La Flor on in a bag, by my side. Why are we still working the same way we did in the ’50s, when pictures had to be made by professionals? Now there are these elaborate colour-correction programs. You shoot a shitty image with no colour, and you have to buy special drives, or discs and discs and discs to correct and store the files. And in the end it’s more expensive than shooting on film itself—more expensive than 35mm used to be, using practical optical tricks. There’s a phrase I like very much by Guillaume Apollinaire: “When we have time, we have freedom.” And it’s true. Time is money, they used to say. But time is also freedom.

Scope: It is a very free film.

Llinás: Because of time. It would not be free if we had to get to something or somewhere at all times. Everyone made other pictures in between. I understand if you have to pay a big star who has to stay on schedule. But there are many potential pictures that do not require big stars that can be shot with cheap cameras, that can be shot in a very humble, free…I don’t know, simple…

Scope: Modest.

Llinás: Yes, modest. And maybe the images would then be more honest, too.

Scope: There seems to be a shift in the second half away from popular genres toward more specialized and reflexive traditions. The fourth episode in particular seems almost like the key to unlocking the movie. Following the prolonged production of the third episode, was there a concerted effort to switch gears going forward?

Llinás: For the fourth episode, when I had the idea for the film crew to shoot trees, I thought why not have them be shooting for a film about trees that rebel against humans. And then I was told that there actually is a picture about that, made by M. Night Shyamalan [The Happening, 2008]! So I was a little discouraged. I didn’t want to use Shyamalan’s idea. After the third episode, where I knew exactly what I needed to do, I became worried about the fourth episode. I didn’t know what it was or what it would be about.

Scope: You say something similar in your introduction to the film, that you have no clear idea what that episode’s about.

Llinás: And it’s true. All I knew is it would be something about a film crew and trees. At that point I didn’t even know it would be a fictional film crew, and not our own crew. But so many filmmakers have done that: Miguel Gomes, even one of our pictures from Alejo Moguillansky, did that—a film about the crew, the crew, the crew! All that said, I knew this story had to have a crew, but I needed it to come from the world of La Flor, to be part of the same picture. But after finishing Episode 3 I really wanted to go back to the world of Historias extraordinarias—I wanted to shoot Historias extraordinarias again! I wanted to get away from the Cold War and the ’80s. I wanted to head out on the road again with that ridiculous crew. And that was the very first step: going back to the road. The director character speaks of a hatred between him and the actresses—which did not occur in real life—and how he needs to go back out and shoot trees. It would help bring us—as it would him—back to the beginning. For us that meant having the girls acting in Spanish again, and bringing back stuff like the Lubitsch scene with the Canadian police, where they attack the director. It was all to feel free. And it was done very quickly. If it took six or seven years to make Episode 3, then it took two months to make Episode 4, five days to make Episode 5, and four days to make Episode 6. So maybe that’s why you sense a turn with those episodes, because it was a turn not just in the time we spent making them, but also a turn in freedom, a turn toward fresh air. We were coming out of winter—the Cold War—and into spring. It was very important to me to shoot the springtime in the next episode. It was a relief after so many years.

Scope: If Episode 4 is the spring, then would it be safe to say that Episode 5, as an homage to Partie de campagne, represents summer?

Llinás: What do you think about the idea of an homage? Who are we to make homages? Stealing would be more appropriate. I wanted to remake that picture—I wanted to be there. I worship Jean Renoir. But if I’m going to make an homage to him, the best thing wouldn’t be to remake one of this pictures. It’s more of an insult, actually, to make one of his pictures again. So I don’t see it as an homage. I’m nobody to make an homage to Jean Renoir. It’s like painting: great painters don’t do homages, they just paint the same pictures in their own style. When Picasso—sorry for the example, I’m not saying I’m Picasso—looks to Manet, or when Manet paints Olympia, it’s not an homage, they’re trying things for themselves that have been tried by the masters. Before painters could see paintings in books, they would have to go to museums and spend many days copying them. It’s a way to learn. We want to be better, so we copy, like the masters copied.

In Episode 1 I was copying Hitchcock—in many aspects I was copying his style to learn his style. In the second part as well I was copying Hitchcock, but also things other than cinema, like Tintin. In the third, it was sad Cold War espionage stories like those of Fritz Lang, le Carré too. So for Episode 5, it was making Partie de campagne again, making a silent picture again, trying to generate the intensity to tell a love story through acrobatic airplanes. All the great painters learned their skills by going back. They’re very modern when they begin, but as they proceed they get more into the past. And it’s the same for us. We’re getting older so we have to go back to the masters. It’s the only thing I know.

Scope: Painting seems like it was the primary inspiration behind Episode 6, at least visually.

Llinás: Yes, but not just the last sequence. For the last sequence we’re doing da Vinci, it’s quite clear. But in some sense I think that Manet is actually the biggest influence on the film—he’s at the very heart of the picture. I’m not quite sure where exactly the halfway point of the film is, but if I had to guess I’d say it’s the Manet sequence, in the spy episode. You see a plaque in that scene that says, “Here lived Édouard Manet,” followed by an interlude of Manet paintings superimposed over one another. About five or six years ago I went with Hugo Santiago to a Manet exhibit in Berlin. At most museums you might see five Manets, but here it took two-and-a-half hours to see all the paintings. You could really study the relationships between the paintings.

Even with Hitchcock, we thought of him as an artist, as a painter. When I talk to my students about Hitchcock I tell them to go see Vertigo (1958), and not on their shitty phones, but projected. Why? Because Vertigo is one of the most important visual art objects of the 20th century. For me, as a painting, Vertigo is much more modern and beautiful than the whole of Picasso.

And in the Casanova sequence in Episode 4 we of course tried to imitate the light of the 18th-century paintings. So it’s always there. It was all part of the game. It hopefully has nothing in common with those directors like Peter Greenaway who try and make these impressive-looking pictures with elaborate lights and design. If it looks like that, it would be terrible to me. We did ours with one light, and with a humour and a lightness. I don’t think the great masters would consider their work as Greenaway considers his work. Or at least I hope not.

Scope: Can you tell me a little bit about the sequence that ends Episode 4, of the portrait-like images of the four actresses shot on set? It’s my favourite moment in the film.

Llinás: I’ll tell you what I feel now, but let me say first that if that’s your favourite sequence, then it worked. It should be like that. If not, then it’s ridiculous. But if you feel that everything has brought us here, to this moment, and you’re satisfied, then the trick worked. I’m sorry to reveal that it’s a trick, but it was!

If I had my way, if the film was just for me, the whole picture would be like that sequence. But no one would care for 14 hours of nice pictures of those girls. You have to arrive at that point, so you care about them. You know why it works? I’m going to say something terrible, but you know how they do the In Memoriam videos at the Academy Awards? It’s that kind of thing. Of course those moments are horrible because everybody cries, and they have to be judicious and intersperse people everyone loves with people nobody knows, but the same idea applies: you need the whole picture for it to work emotionally.

Another influence, especially in this instance, are the films Rossellini made with Ingrid Bergman. And the end of Stromboli (1950), when Bergman—who we all know as Joan of Arc, and as Ilsa in Casablanca (1942), and the spy in Notorious (1946)—walks up the volcano, it’s like she’s carrying the entire story of cinema. And Rossellini knew that. In that moment, the marriage between traditions is quite touching. And of course it’s a fiction, but this is a moment for me when cinema shows its own history. Rossellini is shooting the story of Bergman’s character and the story of Bergman herself. It’s thrilling. So I felt at this point in our film we needed to do the same. These actresses are not Bergman—most of them had not even shot one picture. So we had to make their career. This picture would be their career. You’ve seen their lives, and through these images you now understand their process. It’s the moment when we, the filmmakers, have the right to make that sequence, and it’s the moment when you, the audience, realize that the whole film is a portrait.