Kiju Yoshida is often considered, along with several other loosely connected filmmakers, part of the so-called Japanese New Wave of the 1960s. A book on that unofficial movement (by David Desser) took its title from the director's most famous work, Eros + Massacre (1969). Recently, a series of worldwide retrospectives, including the current one at the Harvard Film Archive, has allowed his films to be freed from their historical context so that they can address a new audience as "pure cinema"—something Yoshida greatly values.

The psychological atmosphere and the formal strategies of Yoshida's films are very different from those of the work of Nagisa Oshima, another New Wave director to whom he is frequently linked. Like Oshima, Yoshida, who is now 76, started directing at Shochiku at a young age, benefiting from the studio's perceived need to target the youth market, but soon left the company when it became clear that there were limits to Shochiku's willingness to accommodate its new directors' political and aesthetic radicalism. Yoshida then went on to make, from 1965 to 1973, a series of independently produced films, almost all of them starring his wife, Mariko Okada, an established star who had worked with Naruse, Mizoguchi, and Ozu. This series culminated in three films examining 20th-century Japanese history: Eros + Massacre, on the Taisho-era anarchist Sakae Osugi; Heroic Purgatory (Rengoku eroika, 1970), on the radical student movements of the 1950s; and Coup d'état (Kaigenrei, 1973), about Kita Ikki, the ultranationalist intellectual whose ideas inspired an abortive military coup in 1936. Since Coup d'état, Yoshida has interspersed narrative feature filmmaking with other projects, including television documentaries and a major book on Yasujiro Ozu, whom Yoshida came to know in the last years of the older director's life.

The Ozu book is called Ozu's Anti-Cinema, a term that may give rise to misunderstandings. It should be understood that "anti-cinema," for Yoshida, means roughly "counter to the rules and conventions of commercial cinema." In this sense, Yoshida's own work must also be considered anti-cinema. In each of his films he starts with a rigorously limited subject, a small number of locations, and a small group of characters, and proceeds to free the patterns formed by these elements from their moorings in time and space. He multiplies angles on scenes, thwarting and denying the narrative function of cinema in favor of a dynamic interplay of different points of view. A Yoshida film doesn't narrate a story, but describes gaps: the intensely charged spaces across which his characters look at one another. Yoshida depicts and organizes these spaces with increasing freedom and intricacy through films such as Lake of Women (Onna no mizuumi, 1966), The Affair (Joen, 1967), and Flame and Woman (Honoo to onna, 1967).

Thanks to the kindly facilitation of film critic Shigehiko Hasumi, I interviewed Yoshida in Tokyo a few days before his departure for Harvard (where he and Okada will appear for several post-screening Q&As, starting April 3). Film director Toshi Fujiwara served as interpreter.

Last year and this year, your work has been the subject of retrospectives in Japan and around the world. What is your perception of the way audiences have responded to your films at these events?

For a filmmaker it's not necessarily a joyous thing to have one's retrospective organized. But as a human being, at a certain point in your life, you have to accept that you are now in the position of being seen in retrospective terms. It also may be because I made Women in the Mirror (Kagami no onnatachi, 2002), which took me 13 or 14 years to make. Since it takes me more than 10 years to make a film, there may not be enough time left for me to add another one to my filmography, and the logical assumption is that Women in the Mirror may be my last film. And as such, it is acceptable to me. It has no wasted moments, and it's as close as possible to what I consider the ideal form of cinema. So I guess I've accepted that my work is now complete and an appropriate subject for a retrospective.

Do foreign audiences and Japanese audiences respond differently to your work?

In Japan there are two distinct categories of audience. One is the generation who saw my films at the time they came out. The other is the young people who are discovering them now. For the audience of my own generation, a certain sense of "actuality" and a certain nostalgia are mixed together in their reactions. If I may allow myself to become a little controversial, when the people of my own generation, or the generation right after my own, watch a film, they try to understand, and if they don't understand, they ask why, and if things happen differently from the way they understand, they reject the film. The younger generation is free from these kinds of prejudices and preconditionings. They don't have the historical context or the context of the characters in these films, so they see and question them purely as beings projected on the screen. This may also mean that my work has now become part of history and that it just has to be accepted. But I also optimistically feel that my films are now seen as pure cinema.

When foreign audiences watch my films, they see them as Japanese films. But perhaps my films don't seem to them to belong to the framework of Japanese cinema. I think that's good for me and for my films, and I feel lucky about that.

Over the 1960s and 1970s, did your conception of your relationship with your audience change?

I don't think that I was really conscious of the audience back then. As a child born before the war, I saw people being pushed in a mass-collective way to go to war. Therefore I felt a kind of inborn rejection of the idea of controlling the masses and manipulating the audience. Perhaps that is why I never thought about the audience when I was making my films. Also, my generation had to face the cruelty of the time before and after the war. In just five or 10 years, the society changed completely. I never experienced an era as being continuous, so I couldn't really have any attachment to a certain period as my era. I never had the capacity to believe in an era or to predict what would come next on the basis of the present.

The program of the Harvard Film Archive refers to the six films you made after leaving Shochiku and before making Eros + Massacre as "anti-melodramas." Do you accept this term?

Naturally there is always a rejection on my part whenever someone else categorizes my works. But when people explain to me why they make such categorizations, then I often accept that such a point of view is possible. I often say that at the same time I am both myself, and also "the other" to myself, a stranger to myself. So when somebody tells me, "You are such a person," then I have to accept it, because I believe that I myself don't know myself. So I find the categorization of those films as anti-melodrama acceptable, to some extent. But you can also call some of them anti-youth films. Whatever you put after "anti-," as long as "anti-" is there, it fits my films.

I was never really conscious of doing something to be anti-something, but whenever I wrote a screenplay, it naturally came out like that. For example, Shochiku once asked me to make a youth film, so I put all my energy and effort into writing a good youth film. Then I made it and showed it to the company, and they said, "Oh, this is not a youth film." With Akitsu Springs (Akitsu onsen, 1962), the company was expecting a romance, a melodrama. And I too was thinking of trying to make a romance film. But romantic love is always betrayed by time, because it's only for a very short period of time that a man and a woman can think in unison. So for me a true romance film is a film that shows the gap between the woman and the man and shows that they go past each other without really finding a common meeting point.

In your films, the male characters tend to be driven by very specific, limited wants and needs, which define their actions in advance, whereas the women are making choices as they go along and are therefore more interesting than the men. Would you agree with this statement?

Yes, exactly. I wasn't so conscious about it, but because I'm a man, I understand men very well. But what a filmmaker—or maybe I can say an artist—really wants to do is express something that one doesn't understand. Since I'm a man, very naturally the most important subject for me to describe is women, because they are an enigma for me. So most of the time my protagonists are women. I don't really want my films to be categorized as women's films, because for me my films are about the most important enigma for me in my life, which is women. And I can stress that point not only because I'm a man and a male filmmaker but because Japan still today is predominantly a male society. Men are still the main characters. We've never had a woman prime minister. Women are oppressed, discriminated against, and rejected; they are outside the thinking of the society or the state. So the only way really to look objectively at that male-oriented society is to take the side of women, who are rejected outside that society.

I don't think it's possible to have an abstract idea of a man and an abstract idea of a woman and put those two characters in a story. Maybe the society, the mainstream, the capitalists, the studio want those abstract ideas of man and woman, who either become happy or unhappy. But when you go down to the level of the individual reality of human beings, there is no abstract idea of man or of woman. There are just concrete man and concrete woman, each looking at the other from a point of view with certain individual distortions, and I've always emphasized this idea in making my films.

Not that I'm criticizing Mr. Mizoguchi, but the women in his films are women seen from a male point of view. They are the women that Mr. Mizoguchi sees. He had a certain idea of the essence of women, or the sexuality of women, and of course it had a certain reality. But still it is just women seen by Mr. Mizoguchi. There is no looking back from the side of women, looking back at men; no exchange of gazes happens between male and female. Of course, even if a woman filmmaker makes a film about women, it doesn't necessarily mean that that film portrays what women are.

How have your ideas about film acting evolved over time?

Maybe for me it's not an evolution, but more that each time I make a film, after making the film, I think perhaps the performances should be more this way, or we should work more on this side. So my ideas about performances in films changed constantly. Often both the filmmaker and the audience are forced to see the character according to a certain set of rules depending on the image of the actor: one actor is good for comedy; another actor is good for tragedy or melodrama. When I make a film, I first take away all these restrictive rules. Which also means I don't have a common language with the studio and with the audience. But that's the front line of my filmmaking. Having taken away all the common rules, either I go forward, or I go backward and make compromises. There's nothing to hang on to, as far as a common understanding with the audience or with the studio in the commercial system of cinema is concerned.

Looking at it more broadly, I also went beyond the normal set of rules for setting up the camera and framing a scene. The common rule is that when you make a close-up, the focus of the shot should be at the center of the frame, so that for most people it's easy to look at, it's comfortable. Which also means that as part of the set of rules of cinema, the person at the center is often unconsciously defined as the protagonist. So I very often frame only half of the face of the actor. It's a kind of resistance, telling the audience, "Don't trust so blindly what you see on the screen. Please try to find by yourselves what is really important to you as the audience, in what you see within this frame." That kind of feeling became stronger and stronger for me.

Was there a large role for improvisation in making your films?

In a way. As far as dialogue is concerned, I write very strictly without any wasted lines, and I don't write explanations: I never have the characters explain why they feel this way, why they cry here, why they're laughing here. So the improvisation doesn't happen so much in the dialogue as in the description of the action. I always want to incorporate in my films what happens by accident during shooting. One example I can give you is from Affair in the Snow (Juhyou no yoromeki, 1968). There is a scene that we shot very early in the morning in Muroran station in Hokkaido. It's a scene between Mariko Okada and two men, her former lover (Isao Kimura) and her present lover (Yukio Ninagawa). They're starting a very dangerous trip. I arrived with my crew before my actors. It was dead winter, very cold. A cargo train was already in the station, filled with coal, and it was so cold that the coal was frozen in the coal compartment and they couldn't take it out. Seventy or 80 workers were hitting the wagon with huge hammers so that the coal would loosen up and they could take it out. So we started frantically making shots of them doing this, and when Okada and the actors arrived I asked them to wait while we shot it. Then the train that we borrowed came in, and we started shooting the scene with the three actors. We could still hear the sound of the workers hammering at the wagon. After we finished the scene, the actors told me that hearing these sounds all the time affected their performances. That was a case of incorporating something accidental into a scene.

I hope you understand this as a joke, but when you finish filming and you go to the editing room, when you cut, you have to have a reason to cut. The clumsiest way is when somebody gets angry and hits a table with his hand, and you cut at this moment and go to another shot. It's such a clichéd technique that people who are used to seeing films just accept it automatically without any resistance. My task is how to replace that kind of ordinary cinematic experience that people have in the theater with something unusual. For that, you need accidents. Because when I calculate certain things or prepare certain things, the audience can always tell.

Did accident also play a role in the complex interweaving of past and present, reality and fantasy, in films such as Flame and Woman and Eros + Massacre? Or were the structures of these films completely planned in advance?

Usually I don't make many changes to the scripts that I write. But even if I write a realistic, coherent, chronological scene, when I shoot it, at a certain point in the middle of the scene, it may turn into fantasy or a flashback. Because the images I created become independent from myself and take on a life of their own, the audience is free to interpret a scene as a fantasy or a flashback even if I never intended it that way. I feel that if I say in words that this scene is 10 years ago, a sophisticated audience might say to me, "You shouldn't have said that. It's better if you don't tell us this directly." A film is ultimately not about what I tell the audience to see but about what the audience sees and discovers for themselves. I myself was the kind of viewer who believes that cinema can exist as a means of communication without following the commonly accepted rules of time and space.

It seems to me that in Eros + Massacre, there is an implicit criticism of Sakae Osugi in the rather macho, self-possessed manner of the actor's performance, which is so different in style from the acting of the women. Was this your intention?

Sakae Osugi was an anarchist who put free love into his logic of anarchism. He insisted that the institution of monogamy is only a set of rules imposed by the state, or by morality. So to free themselves from those boundaries, human beings must be free to make love. But from a female point of view, this could be just a male's egoistic idea. If Noe Ito, his mistress, played by Mariko Okada in the film, had said to Osugi, "I'm in love with someone else. Why can't we live together?" Osugi would probably have said no. So I consider the idea of free love proclaimed by Osugi to have actually been male-chauvinistic egoism. When you say that he looks macho, what I wanted to portray was—not a male comic, but a male clown. His performance is more theatrical than the others; it's as if he were trying to call attention to himself. In Japanese tradition, kabuki theater has that kind of performance that attracts attention to oneself.

I think it's possible for an uninformed Western viewer to understand the issues involved in the historical story of Eros + Massacre, and their relevance to the time the film was made, just from watching the film, but this is probably not true of Coup d'état. If you were introducing the film to an audience today, what would you say?

Even in Japan Coup d'état is very difficult to understand. I had never wanted to make a film about Kita Ikki, but I thought it was necessary to make a film about him, to show that such a thinker existed in Japan, and how he thought about the imperial system. My original idea was to make the protagonist of the film an anonymous soldier from whose point of view we would see this historical figure. For the Japanese audience there is a certain resistance to seeing the film like that. They can't understand how this anonymous soldier can have a conversation with this famous figure, why he is invited into the house of this man. Maybe for a foreign audience it's easier to forget about the political context and to see the film as a drama between the anonymous soldier and this mysterious thinker, Kita Ikki. Also they quickly become aware of the strangeness of the style of the film. It doesn't fit into the mode of the jidai-geki [period film]; it also doesn't fit into the mode of political filmmaking. So the foreign audience quickly understands and accepts the alienating style of the film.

In the years since Coup d'état, how have your filmmaking concerns and style changed?

When I made Coup d'état, I felt that there was a limitation to what I could pursue in the medium of cinema. In Coup d'état there are no characters according to the preconceived rules of cinema. It's totally free. So even if I were to treat different themes, it would be only a modification of what I did in Coup d'état. For me, it's the most perfect film, in the sense that there is no waste. Just the structure is there. The rest of my career would have become just a repetition of what I had already completed. A different way of saying that is that even if I took a situation that was happening in the contemporary time and made it into a film, the structure would still be the same as that of Coup d'état. So I quit fiction filmmaking for a while. I started to make a series of television documentaries about art, The Beauty of Beauty (Bi no bi, 1974-77). Now that more than 30 years have passed, I can confess that at that time mentally I was not very stable. It was a dangerous time for me. I was 40, and I wanted to quit cinema and start something new to do for the rest of my life. But I didn't know what to do, so I was in a very dangerous mental confusion. For five years I made The Beauty of Beauty, then for five years I stayed in Mexico trying to make a film, so basically for 10 years I dropped out of my own life and escaped from myself.

After that you returned to cinema.

After 13 years I made The Human Promise (Ningen no yakusoku, 1986). It's a story of a son's mercy killing of his own mother. It was a taboo; that's why I wanted to do it. I was very interested in this. My challenge was how to treat it as a cinematic expression. It was purely a question of cinema. So the 13-year blank in my career served me well in that I could return as a simple filmmaker, thinking purely about cinematic form. Before that 13-year blank, I was also thinking about the past of cinema, trying to free myself from the pre-existing conventional set of rules of cinema. Now I was able to think about nothing but finding a good cinematic expression of a taboo issue, setting myself a cinematic challenge, instead of struggling with all the other burdens I had imposed on myself.

Then I made Wuthering Heights (Arashi ga oka, 1988), dealing with necrophilia, another taboo issue. This film too I made as pure cinema, just as a cinematic challenge.

You have written of Ozu's resistance to the conventional functioning of cinema, and your own work has embodied such a resistance. Have you come to see Naruse's work as showing such a resistance too?

Such a resistance to cinematic rules is not apparent to me in Mr. Naruse's work. I feel that Mr. Ozu's films are much more youthful than Mr. Naruse's films. Mr. Naruse's films seem very adult, as if the filmmaker were someone who can't easily be moved. Sometimes, as in Untamed (Arakure, 1957), he makes a very emotional, agitated film. But I think the prime example of his work is Floating Clouds (Ukigumo, 1955). Of course theoretically I can understand that the repetition of two lovers meeting and separating again and again over a period of time, if done properly, can become very strong drama. But you can't do it just because you know it theoretically. Mr. Naruse did it in a very powerful way. That's really an amazing film.

Akitsu Springs and Floating Clouds have the same structure, and I was very aware of that. But I think it was Mr. Naruse's personality that made it possible for him to create his film. In my case, I had to go not just through myself, but through the filters of my generation, my society, and growing up during the war. In Floating Clouds, both the man and the woman are half responsible for what happens to them, half not, but they're equally responsible. For me, with the burdens of my generation and my society, the historical responsibility of that period is on the men's side, not the women's, so the male side must be responsible for everything bad that happens [laughs]. In Floating Clouds there are flashbacks to the time they spent in French Indochina, and both the man and the woman remember it as "We were happy back then." They're speaking their honest feeling. They share the same memory in the same way. But in Akitsu Springs, that mutual sharing of a memory is impossible. Already, the meaning of the end of the war has become very different for a man and for a woman.