What is the act of creation? – an interview with Hou Hsiao-hsien

Owen Brademas used to say that even random things take ideal shapes and come to us in painterly forms. It’s a matter of seeing what is there. He saw patterns there, moments in the flow.

Don DeLillo The Names

For critics like us, the cinema that we loved at the end of the 90s, beginning of the 00s, had reached such a level of complexity as to be a direct challenge to the art of criticism itself. It felt as if there was little room for conventional responses to these films, the temptation was to eschew rationality altogether and deem poetry a more adequate medium of response than conventional film analysis, or alternatively, to play up the rational and the novelistic: to see in the films aspects of character and situation that standard criticism couldn’t always access. This meant not just seeing the films as fictional cinematic works (commenting on story, form etc.), but also believing they demanded some of the approaches usually utilised in writing poetry or stories and novels. Rather than thinking of the technique behind the films, you would have to muse over what was in front of our eyes. One of the filmmakers best exemplifying this, and around whom a shared interest in the work helped our friendship develop, was Hou Hsiao-hsien. In some of the films we could see that the poetic approach to criticism was necessary, in others the novelistic aspect more useful. In three successive films in particular, Goodbye South Goodbye (1996), Flowers of Shanghai (1998) and Millennium Mambo (2001) the narrative, formal and figural patterns attained such a level of abstraction as to threaten comprehension, the ‘almost known’ vied with ‘that which can never be known’ in an aesthetic of the indistinct, inventing shapes on the threshold of taking form or of disappearing. If in these films the flow of sound and image was momentarily stalled and nascent forms and figures traced in light for the briefest period, the result was no longer to isolate the time to live and the time to die but simply to capture the time to be seen and heard. It seemed to those who had absorbed all the goals and promises of Modernist cinema that access was again granted to its most subtle perceptual and pensive qualities and pleasures. Here was the poetic cinema our generation had dreamed of.

But we could also see in Goodbye South Goodbye and Flowers of Shanghai, as well as in earlier work like City of Sadness, Dust in the Wind and A Time to Live, A Time to Die (and to an extent in the later films Cafe Lumiere and The Flight of the Red Balloon), a director who needed to be understood not just through images (which indicated the poetic) but also through character and situation (indicating the novelistic). This didn’t mean that everything had to be grounded in explanation: it allowed for an interim space between the poetic which wasn’t interested in the explanatory, and the critical which can often in the process of offering it remove the ambiguity of the work. Yet in the novelistic we could see the hypothetical, the speculative possibilities available in the work. We could see in Flowers of Shanghai for example the sense in the opening scene that Tony Leung’s character was quiet and slow to humour, as though his mind was on other things and the opiate calm was less effective at stilling the chaos inside him than for the other characters sitting chatting. We noticed in Goodbye South Goodbye that, like Scorsese with, say, Mean Streets, Hou could create characters that had reason that reason didn’t know, obscure, dangerous drives that could never be corralled into the conventions of plot, yet neither were they arbitrary. A novelistic sensibility could find in these characters a certain causation without narrative motivation. Hence when the central character starts a fight early in Goodbye South, Goodbye, we are not surprised when he turns on his brother a few scenes later. Yet we call this causation rather than motivation because Hou doesn’t build into his films either a narrative thrust or a categorical progression. The scenes remain situations, moments that we puzzle over so that when one scene finishes we aren’t looking forward to the next one, we are still musing over the one just completed. By contrast even though Mean Streets is one of Scorsese’s loosest films, it is no surprise when the debt-ridden Johnny Boy gets shot at the end of the film: it is part of the film’s inexorable movement. This also means that Hou doesn’t create much anticipatory force to his scenes; they can exist almost as if for themselves. A few minutes of poetic suspension! Writing about such scenes we draw less upon situation and character than pure image, trying to find the poetic language to describe the play of light and the importance of colour.

As cinephiles and critics the ultimate question we sought to ask of the work was: what is the key to this level of creativity? Now, this is probably the last question that any great artist should ask themselves lest, to paraphrase Bergman, they seek in their subsequent films to repeat the style of their previous ones. Bergman’s jibe at Tarkovsky and the Russian’s perceived overthinking of his unique stylistic traits, seems to imply or lay claim to a realm wherein certain artists preserve an unquestionable and intuitive knowledge or innate sensitivity, as if they are the possessors of an internal guidance system that only locates the most creative zones and is immune to critical inquiry, even on the artist’s own behalf. This is where we can invoke not only the poetic and the fictional but also the clinical (as opposed to the psychoanalytic). When the British writer Dennis Potter was asked why he didn’t seek therapeutic help for his appalling psoriasis he said he wouldn’t want to be cured if it risked creative inertia as a consequence. That he would have lost his drive to write along with the loss of the disease. This is an exaggerated example, of course, but there are numerous filmmakers one feels are not inclined to involve themselves in analysis of their own work and would prefer it to be done when they aren’t in the room. How often have we heard filmmakers belligerently respond to questions that seem to be probing an aspect of the work the filmmaker is happy remains unknown? We can think perhaps of Bela Tarr, of John Cassavetes, of Werner Herzog. Of course there are also filmmakers whose work is inherently self-reflexive, the folding, unfolding and refolding back on itself proving the genetic element for each new work (JLG being exemplary in this respect). Bergman was well aware that the risk of Mannerism is an ever present one for filmmakers and we’re tempted to think that in this context Hou may not yet have left behind the delectable snares of the Ozu universe he entered for his tribute film Cafe Lumiere (and he wouldn’t be the first to whom that capture occurred, one thinks of Wenders and Kiarostami, both of whom also paid tribute to Ozu and remained somewhat ‘lost in Japan’ in their subsequent careers). Yet perhaps there is a difference between a refusal of reflection and the final acknowledgement of the self-reflexive. Bela Tarr’s films became distilled attempts at formalising elements that became an essence of world, with The Turin Horse a final film because where else was there for Tarr to go? He had created a very self-conscious style that allowed Mannerism to meet minimalism and miserabalism.

But even if HHH is well known for his reluctance to conceptualise his filmmaking process, knowing this didn’t at all quell our enthusiasm at meeting him at the recent Chinese-language Film Festival in Dublin, curated by Marie-Pierre Richard (Hou was accompanied and often assisted in the interview by his scenarist Chu Tien-wen). An hour later we left feeling that with certain great artists it’s enough to start and finish with the notion of the artisanal and the pragmatic, and in this Hou reminded us of someone like Sidney Lumet. Faced with the pragmatic approach to creativity, it’s best to let the artist’s own discourse govern rather than trying to impose on the conversation a set of terms from outside and with which they have no relationship—all an interviewer needs to do is to try to ask questions that will provoke some new thoughts, rigorous reflections on the day-to-day process of filmmaking and if the answers aren’t satisfying, ask again from a slightly different angle.

HHH has spoken a lot about atmosphere taking precedence over plot in his work, and we began by asking him how he might define atmosphere and how he achieves it.

Atmosphere comes from character, and without it, whether it’s a period piece or a contemporary drama, and every period has its unique style, there is no story. I start with the character and I look at the character in a situation, starting with the family background of the character, what this person does, in terms of occupation, personality etc. Once I have this character in mind everything comes together, I begin to construct a full picture of a being, it’s not just this person alone, it’s where they come from, the family background, the parents, the domain of work—by looking at the different facets of this person’s life I create a complete character. I really believe in character. By character I don’t mean the character in the film but the character of the actor. Take Shu Qui for example, who I first used in Millennium Mambo. I start with the actor herself, what kind of person she is, her characteristics, outside of any character she might play, so that through conversations with her I find what kind of person she is, I construct an image of her in my mind and through that image comes the personality and behaviour—things she would do, actions she would perform, whether in love, in life, anything like that. And then from there comes what I would think is the atmosphere, the basis of atmosphere in a film. For me it’s also about the emotions and when humans interact on screen with these emotions they bring something into it and that might be what viewers perceive as an atmosphere. It comes very naturally. These are just observations you make about life all around you and that translate into the film.

It’s not just the camera objectively capturing what’s put in front of it, what role does point-of-view play in the process?

In terms of framing there is of course a perspective but its not always there from the beginning. I always come back to the actor him or herself. When I work with a new screenplay and plan the filming of it I already know who I want to cast for the role and I always bear in mind the actor’s own character in trying to develop the character further. At times when the actors are right for the part they evolve in terms of the character, at times they don’t, so when they don’t evolve then that can become an problem that must be overcome. In terms of Flowers of Shanghai, we started out with an event, something happened that day at the event, the function of the long opening shot is to show what happened and from all these different peoples’ perspectives. But it’s not an exact science so when we were first doing a scene I’m not going in with the idea ‘this is the way I want it to unfold’ but rather through filming and through different shots it becomes clear, becomes more salient what I want to happen and that becomes the end result.

This pragmatic method of always focusing on the ’here, now, things in the making’ guides other factors like shot length and camera movement.

The length of shots is not a deliberate choice, it arises naturally, organically, out of necessity, the reason being I don’t do rehearsals. We are immediately thrown into a scene and the way it works is that I know well that to get even one shot done is very rare. Because of my method of working without rehearsals it becomes necessary for the actor to just go in there and develop the scene as they see fit. Sometimes it works and if it works, great, and if it doesn’t I have no issue with trying again and again and again. Actors that work with me are quite used to this immersive method of working and therefore, for example, in Flowers of Shanghai, we were working within quite a confined space, and within a confined space if you try to do frames, to cut the scene into several frames, I’m not interested in doing that. So it became more natural to have the cinematographer follow the movements of the characters and he needs time to capture what’s going on and to create the storyline. That’s what my longtime collaborator Mark Lee started doing and it made sense. Its important to depend on actors to deliver a performance on their own behalf and it doesn’t matter how many times I have to shoot a very long scene. I don’t mind doing it over and over, over several days.

Did your work change when you began to move the camera?

That method of working hasn’t changed whether I used fixed camera positions like in my earlier work or the moving camera of my more recent films.

I don’t build sets, everything is shot in real places, often there are limitations in how much room there is to move—some spaces are very confined and there is very little movement possible. For example, the character of Tony Leung’s elder brother in City of Sadness is a very dramatic actor so when we were filming, he tended to be overly dramatic for what the scene required. The way to combat this, to overcome the problem, was to pretend to him that we were just doing a test and it wasn’t real, and whenever he felt he was doing a test he would tone it down, for whatever reason, but it was actually a real shot, he just didn’t know it and then the minute I said ‘ok this is a real take’ he would revert back to his very dramatic style. That was the only way to get him to give the more realistic kind of performance that I’m looking for.

To give another example from Flowers of Shanghai, some of the actors were having a very difficult time remembering their dialogue because it’s in Shanghai-ese and trying to remember any other instructions would be too much for them to handle. Within that very confined space the way to tackle movement and without having rehearsed, the problem became ‘when will the cinematographer know when to pull focus or move the camera?’ There’s a maid, a tea lady, walking around pouring tea and we decided that the actor doing this would know to move when someone begins to speak. So it was decided that the minute the main character speaks she will move over there and the cinematographer will follow. So I provide a frame and within that frame it’s up to the actors to do what they want to do. Fixed camera position or moving camera, the problems are the same centred on the question: what is the point of the scene?

A lot is therefore left open to chance. The one thing I do care about, take City of Sadness as an example, if it happens to be a scene involving lunch or another mealtime its always a fresh meal, an actual meal, and its important that the actors are hungry and that they’re not acting, so that, say, the children come and get some food, they run around, it’s what happens in normal life. So those things are not manipulated, not controlled. What’s provided is this frame of reference. The scene I mentioned is following a previous scene where something happened that feeds into the meal scene so it’s more organic—the feelings, the emotions, the atmosphere follow from that previous scene. I don’t mind doing retakes but if it doesn’t work I’ll leave it and return some other day after other scenes are worked through, for example in this case, I’ll wait until they are hungry again.

If something doesn’t work I’ll change some element, it can be the way of shooting the scene, the camera angle or even the dialogue, to see if it works better.

My actors aren’t precise so if you ask them to be the same each time it’s not possible. I’m very sensitive to an actor’s state of being, if their state of being is a good one it will work. If it doesn’t there’s no point in trying to keep going and no point in returning to it immediately to see if I’ll get better results. A lot of times, if I’m reworking a scene I’ll put them in a new situation, it won’t be exactly the same—by changing something it might create a better result.

Often therefore this style of filmmaking gives rise to unintended meanings?

It can happen that I don’t intend a meaning. Again, it comes back to the limitations of the location, the actual space we’re in. Once I’ve selected the frame, whatever happens within the frame or outside the frame there’s no point in trying to control it. Sometimes we use tracks to follow what’s off screen but very rarely.

So there’s no identifiable philosophy guiding your poetics, your editor Liao Ching-sung has suggested there’s a touch of Taoism at play?

I think it’s more to do with what he thinks, if he thinks there’s Taoism, I don’t, at least it’s not intended. [Here Chu Tien-wen interjects]: Liao believes you need to spend a lot of time with your negatives, your raw footage, because then they’ll talk to you and tell you how they like to be cut and that might be quite Taoist.