William Forsythe’s methods of choreography are strikingly algorithmic and give rise to a style of movement and interaction that is distinctively his own. This conversation between Forsythe and Kaiser was recorded in 1998 and later published in Performance Research, v4#2, Summer 1999.
I first met William Forsythe in his kitchen in Frankfurt in 1994. The first thing Bill did was to try to explain how he goes about creating new movements. He started drawing imaginary shapes in the air, and then running his limbs through this complicated and invisible geometry. As a non-dancer, I was completely lost.
Later that year, I suggested that he use computer animation superimposed on videos of himself explaining them to make this geometry visible. Together with Chris Ziegler and Volker Kuchelmeister at the Center for Art and Media Technology (ZKM) in Karlsruhe, he created a multimedia work along these lines entitled Improvisation Technologies . Since then, he’s exhibited this extraordinary catalogue of dance procedures in several museums, and still uses it in training new members of his dance company.
Three aspects of Bill’s approach have always struck me:
In April 1998 Bill and I spent three days together, at the end of which we recorded the following conversation. Bill had read transcripts of my Unreal Pictures and Playground dialogues with Michael Girard and Susan Amkraut, so our conversation built on that earlier discussion.
KAISER: Why did you begin using spatial concepts such as rotation, extrusion, inscription, and refraction to create dance?
FORSYTHE: Necessity was the mother of that invention because our dance company never had a lot of time to make the work. My basic method, developed over a period of 15 years, is to find ways to use what my dancers already know. Since I work primarily with ballet dancers, I analyze what they know about space and their bodies from their intensive ballet training. I’ve realized that in essence ballet dancers are taught to match lines and forms in space.
So I began to imagine lines in space that could be bent, or tossed, or otherwise distorted. By moving from a point to a line to a plane to a volume, I was able to visualize a geometric space composed of points that were vastly interconnected. As these points were all contained within the dancer’s body, there was really no transition necessary, only a series of “foldings” and “unfoldings” that produced an infinite number of movements and positions. From these, we started making catalogues of what the body could do. And for every new piece that we choreographed, we would develop a new series of procedures.
Some of these procedures worked with what is already in ballet. If you analyze the basic ballet position where the hands are held over the head, you realize that there are two curves involved, one on the right and the other on the left. You can create innumerable transformations from that simple position, which is a given in ballet, and can act like a keyframe. You can extend it out into space, or by let it move through the body as a natural continuation of the curves. You can also make dancers perceive the relationships between any of the points on the curves and any other parts of their bodies. What it boils down to in performance is the dancer illustrating the presence of these imagined relationships by moving.
And in the process discover new ways of moving.
What it actually does is to make you forget how to move. You stop thinking about the end result, and start thinking instead about performing the movement internally. That’s what pulls your body through its “rigors,” as it were.
That approach diverges from classical ballet, where the final position is paramount, as opposed to what goes on internally and in between.
Well, I don’t know about that.
Take the ballet position of épaulement, which is the crowning accomplishment of great ballet dancers. It entails a tremendous number of counter-rotations determined by the relationships among the foot, hand, and head ‘ and even of the eyes. As in Indian classical dancing, it dictates rules of gazing past the body. For me épaulement is the key to ballet because it demands the most complex torsion. The mechanics of épaulement are what give ballet its inner transitions.
At the Frankfurt Ballet, we’ve created a new “paradigm of rigors,” in which the dancers maintain very complex torsions during physically antagonistic events. This happens in motion. For example, you can spin out of a classic position, and as this spinning undoes the position, you look at the resulting distortions to your body and correct them. You correct them within the aesthetic rules of ballet ‘ you never lose your balletism.
So it’s ballet under stress.
Well, all of ballet is about maintaining decorum under extreme physical stress.
What we do differently from traditional ballet is to focus on the beginning of a movement rather than on the end.
To see what spills out from there?
Exactly. We use the reflexes that we’ve learned in classical ballet to maintain a kind of residual coordination, which allows the body to acquire elastic surfaces that bounce off one another. This elasticity is derived from the mechanics of torsion inherent in épaulement.
In my conversation with Michael Girard, he defined “grace” as the smoothest and most efficient transition between two positions. This is certainly the case in classical ballet, where both positions and transitions are highly formalized. At the Frankfurt Ballet, however, your definition of grace seems to be based on unstable, complex movements rather than smooth, simple ones. As opposed to traditional choreography, which can be memorized and duplicated rather easily, your pieces must be much more difficult to teach — and to learn.
The simplicity of classical ballet is precisely what enables it to be reproduced with such ease. I sometimes think of it as an unconscious mimicry of the printing press in Gutenberg’s time. In fact, there is something extremely alphabetical about traditional ballet figures and positions – they resemble glyphs.
Since today’s technology is digital rather than alphabetical, why shouldn’t we go with the flow?
Your choreography does seem to take what is spatial and fixed in ballet and make it temporal and unfixed.
One of our ideas is to imitate a computer application that can wrap a different quality around an existing event, thus altering its very nature. This is another reason why I’ve stuck with ballet. It defines a very precise spatial environment, which I’ve through a series of distorting operations.
A lot of what we do in our company is based on states of fold. We teach our body how to fold and unfold again, at various rates and moving through different body parts. So we create what I call a “many-timed body” folding and unfurling towards and against itself.
One aspect of classic ballet is the constant folding and unfurling of just the leg, which the dancer always brings back to one of the prescribed positions. Our fold differs in that it not just in the knee but also in the hips, thus affecting the torso as well. This means that instead of remaining at a 90-degree angle to the floor, the torso begins to fold down and become parallel with it. An entirely new set of mechanics then takes over, since the body has achieved a new state of balance.
Since your dancers focus on the beginning rather than the end of a particular movement, how can they predetermine their final position? And if they can’t, isn’t this significantly different from classic ballet?
Well, they still have all the reflexes of the traditional ballet dancer, and they have essentially the same basic mental training, which lets them picture points in space very precisely. They orient their positions very quickly within those points. Of course, the mental images we use are not traditional.
You recently told me that your body happens to have a high proportion of “fast twitch” muscles, and that you look for similar bodies in choosing new dancers. In what ways has your physical constitution affected your choreography?
I like the physical thrill of rapid shifts, as opposed to smooth transitions, and a “fast twitch” body allows you to do this. Our company has something in common with collies, who can herd and change direction so quickly.
In classical ballet, in addition to adagio and grand allegro, there is something called petit allegro, which involves small, fast movements made primarily by the feet and legs. In applying petit allegro to the entire body, I’ve found that it’s possible to move it in counterpoint to itself. It’s more like playing the organ than playing the clarinet: you’re not just using one part of your body, you’re using it all.
As you know, Shelley and I originally toyed with using “Motion Alphabets” as the title for this book. I’ve since learned from you that you yourself have invented a procedure you call “movement alphabets,” which you use to help make new dances. How does that work?
I chose the alphabet because it’s simple, familiar, fixed, and arbitrary. We use it primarily as an index for a database of movements. Everyone knows alphabetical order, and the dancers can navigate through its sequence easily even if they only do so in chunks.
Again, this is not so different from traditional ballet where each motion and position has its own name. In a similar vein, I created a non-balletic vocabulary of 135 movements, which I then taught to my dancers until they knew it backward and forward. No matter where or when the dancers move through the zone of one of those movements, they immediately know its place in the sequence. It’s like rapidly scrolling through a list of names in a computer program.
We use our alphabet in connection with the kinesphere — the total volume of a body’s potential movement. Dancers are always conscious of their kinespheres, which exist in the air around them. For us, it becomes a huge field for jogging memory.
Let me give you an example of how exactly how this works. When I cup the back of my neck with my hand, it’s as if I were swatting a mosquito — and so, using this arbitrary association, we say that I’m spelling the letter “I” for “insect.” Now suppose that while I’m dancing, I suddenly find my hand cupped around my knee, which reminds me of the insect element. Bearing in mind that my focus is always on the beginning of a movement rather than on its end, I will have to fold my neck down to that point in space rather than performing it standing up, as in the original alphabet. Now, keeping to the sequence of the movement alphabet, I can perform the movement either directly before or after I — that is, the movements associated with either H or J.
In this kind of dancing, I can lose my equilibrium within a dance phrase, then remember everything from the point of that dislocation, so to speak. My body exists in the sphere of its own memory.
In Alien Action, were you performing operations based on the elemental motions represented by each letter?
Yes. Each letter of the alphabet, which covered as many human configurations as possible, was then modified by the various operations that we’ve developed over the past 15 years. These modifications reminded us of more letters, which in turn recalled more operations. It was a total immersion system.
Your description makes me think of recursive algorithms, where procedures call themselves, modify the results, call themselves again, and so on.
In fact, Alien Action was the first time that I actually began to produce movement based on recursive algorithms. However, they were fixed variations that we created through a long, painstaking process, not unlike that of computer programming, where every step has to be repeated ad infinitum.
The dancers in Alien Action face a challenge similar to that of the characters in the film Alien. Both are trying to find their way in an unknown architecture, and both are using a diagram. The dance diagram, however, does not depict any concrete or existing space, but rather a potential space — as the piece forms, an architecture emerges. The goal of the piece, actually, is to form another and smaller stage within the real stage. This is a drastic scale shift — the whole thing suddenly has to happen on a stage one eighth the size of the original stage, a dramatic condensation.
Dancing Alien Action is like navigating levels on the computer. You can’t just move directly sideways to the desired destination, you have to go down to a different floor, so to speak, and then walk a ways and cross over and move back up.
Are your methods now more advanced — in Eidos, for example?
I believe so. In Eidos I was searching for a counterpoint algorithm.
There were also certain emotions associated with Alien Action that led to extremely idiosyncratic, almost narrative events built into the movement. Eidos, on the other hand, is completely abstract, even though the scale of the entire structure provokes a powerful emotional reaction.
All this supports my notion that your dancers have to think even faster than they move – and they move very fast.
Yes, but don’t forget that visceral thinking is something that’s acquired over a long period of time. Even so, the first act of Eidos requires an encyclopedic command of a huge kinetic field. The dancers must be able to recover any part of the piece instantaneously, since there is always a physical “accident.” When the force of gravity throws them into another configuration, for example, they have to analyze themselves and their current state in relation to the entire piece. In this sense, they are always in a “possessed state,” whether it be Apollonian or Dionysian.
That’s quite different from a more traditional dancer, who’s simply moving through a sequence that’s well known in advance.
On the contrary, I believe that truly great dancers, such as Gelsey Kirkland, are equally “possessed” by the act of defining what they’re experiencing. When she performed, she was entirely in the moment.
Still, that seems very different to me: her moment is not nearly so uncertain!
However, let’s move on to another topic. I know that you’ve spent a good deal of time creating and manipulating drawings, processes that you’ve likened to that of choreographing for dance.
Let’s take this scene from Slingerland, for instance. What I see here is not exactly a tracing, but an extrapolation of lines from a still photograph.
This kind of drawing is an attempt to mask origin. It does work by extrapolation. Where it diverges from the original photograph is in its repetition of elements that get in the way of one another, which creates an unusual kind of architectural space that emerges entirely from itself. It’s a proliferating space, and also a space of loss: you’ve lost any sense of the concrete, leaving you with nothing but indications of its origins.
In one of our earlier conversations, you spoke about colonizing a photograph as if you were an alien element.
That’s exactly what we’re attempting to do when we take one cultural event and use it as a host in order to create something entirely different. I’m sure that the makers of the film Alien could never have imagined how we would use it as a model for our own mechanics.
In this drawing, you’re colonizing yourself. You’re taking off from a still of one of your own productions. Whereas the photograph shows a stage space that’s fixed in the usual way, your drawing creates an architectural space that’s precisely unfixed.
Well, I think the space was potentially there, and it was just a matter of choices. The drawing suggests that from that space there could have been these vectors generated.
Recently I’ve also appropriated (or “colonized”) some Tiepolo drawings, which I found in a Dover book. The figures in the ink and charcoal drawings are like knots of figures hovering in the air, suspended and tangled in the sky. From their limbs, heads, shoulders, arms, wrists, knees, and butts, I drew rather complex vectors.
I used this as the basis for a dance, which took the form of the following task. Given these complex, knotted, puzzle-like configurations, the dancers were asked to “solve” these configurations by unknotting them via the vector paths I’d drawn. Each separate page became a key frame. Using the vectors, the dancers had to invent a transition to the next frame they entered.
Let’s talk about the future. In your thirty years of dance-making, you’ve discouraged people from writing books or making films about your ballets. Yet here we are exploring the possibilities for making virtual dances. Such works are made not of flesh, not of paper, not of celluloid, but of numbers. In principle, at least, they could last forever.
Well, you raise a number of points. Let me start by saying that up until recently I’ve created works specifically for the stage, and not for the page or the screen. The quality of light and of sound, not to mention the physical presence of the dancers, cannot be reproduced, so I’ve wanted my productions to stay intact as live performances. Recently, however, I have made two short films, Solo and Duo, with the specific aim of bringing viewers closer to the dancing.
Now, people often ask, where is the book of photographs of the Frankfurt Ballet? Ballet has been blessed and cursed by the profusion of coffee table books, each with ever more beautiful pictures of graceful bodies frozen in air.
But our work is about moving between positions and passing through positions, not maintaining positions. This is actually a fact of ballet in general, new and old: one moves through a position with greater or lesser accuracy.
No one has ever done arabesque, they’ve passed through an approximation of it. Arabesque will always remain primarily a prescription, an ideal. I mean, there is a good arabesque and a bad arabesque and a phenomenal arabesque, but arabesque is about passing through. It’s more about time than it is about position.
Now to answer your question about the future, I’d say that the virtual dance is certainly not for posterity, it’s for now. As Balanchine once said, the dance of today will not be the dance of tomorrow.
We’ll see about that! That certainly isn’t the case for other temporal arts like music or literature, where reproduction is not a problem.
In any case, what interests me about your virtual dance ideas is that my thinking has mysteriously or surprisingly coincided with developments in computer programming. In reading your dialogue with Michael and Susan, I’ve noticed that the questions are virtually the same. In fact, it reads like my diary — as if I’ve come across messages I wrote to myself. When you talk about phases of movement shifting through parts of the body, and about their visible duration and rates of decay ‘ that’s dance. That’s exactly what we talk about at the Frankfurt Ballet. All of us seem to be posing the same kinds of questions about how to organize kinetic events.
Some choreographers create dance from emotional impulses, while others, like Balanchine, work from a strictly musical standpoint. My own dances reflect the body’s experiences in space, which I try to connect through algorithms. So there’s this fascinating overlap with computer programming.
For Eidos, I gave my dancers ‘ and myself ‘ the following general instruction: “Take an equation, solve it; take the result and fold it back into the equation and then solve it again. Keep doing this a million times.”
Recursion again! Where is all this heading?
If you look back over the last couple of centuries, the dominant paradigm for what I call the temple arts — music and dance — has been counterpoint.
Now once you begin to analyze the nature of an event carefully, as we did with ballet, you begin to see completely new possibilities for counterpoint. We looked at ballet and asked, what makes this function? We looked at something classical, Symphony in C by Balanchine, for example, and the logic of its functions began to emerge. This logic is simply about creating ways to connect.
Now we find that these ways to connect can be algorithmically redefined — infinitely. Since we’re no longer restricted to the prescribed classical methods of connection, we’re open to an extraordinary leap in connection, which is just a matter of defining connective space.
That’s where your focus on spatial procedures and the architecture of movement maps so well onto computer algorithms and virtual spaces. As you said before, it’s as if we’re all on the same quest.
How do you define that quest?
Shelley and I have spoken about it as the search for a new art form, which seems about to emerge from this odd confluence of the dance, visual art, and computer worlds. I imagine that in this new form, performance and recording and notation ‘ three strands of the performing arts that have always been separate ‘ will be fused. So that you can have the notation shaping the performance, the performance shaping the recording, the recording shaping the notation, and so on. Perhaps this new process, which builds on itself, can bootstrap a new way of making art.
Where I’d start is with the score. What’s been missing so far is an intelligent kind of notation, one that would let us generate dances from a vast number of varied inputs. Not traditional notation, but a new kind mediated by the computer.