Frieder Nake on “Machinic” Miracles
Frieder Nake on “Machinic” Miracles
Prior to the Herbert W. Franke Foundation’s Generative Art Summit Berlin, Frieder Nake, one of digital art’s first-ever practitioners, spoke with Peter Bauman (Monk Antony). They cover the very origins of algorithmic art, Nake’s role and its evolution to today.
Frieder Nake: I'm not a religious believer in any sense. Since the age of 14, I have given up all those thoughts. However, what’s happening here is a wonder, a miracle. You are somewhere in the world. I am somewhere else in the world. We have computers in front of us. We click here and there. We have no idea—at least I can say so of myself—how it is happening that we talk and hear each other, even see the other one. All I know is that this usually works fine. When the “system” realizes that there is a problem—like a camera not working—it suggests measures to solve it and continues working. This is how you become a religious believer [laughs].
Peter Bauman: It is a miracle, especially for me. I'm so grateful for the opportunity to speak to you. Thank you so much for your time. We are chatting before the Herbert W. Franke Foundation Generative Art Summit Berlin, organized by Susanne Paech, Franke’s widow. Can you talk about meeting Franke for the first time?
Frieder Nake: I got to meet Herbert for the first time in 1966. I think it could have been at a meeting in Frankfurt, Germany, where somebody brought together a small group of people to discuss computer art.
Computers in art—people always called the results of this “computer art,” which I hated because there is no graphic artist who would do “pencil art” or “ink art.” The means have never defined the style of art.
Lithography may be a bit of a counter example: lithos means "stone.” But, usually, when talking about some kind of art, some style or epoch of art, the means are secondary only, never primary.
The word “computer art” puts the means into the lead. I always thought this to be wrong. My term is “algorithmic art.”
Peter Bauman: You place the algorithm at the forefront. How do you see the relationship between algorithms and systems?
Frieder Nake: Those who today are nearly permanently using computers are almost never concerned about algorithms. They have essentially no clue what this is—an "algorithm.” The world of computers is, of course, full of them. But we don’t need to know anything about algorithms when we use a computer. Deplorably, many of my students nowadays have no idea anymore about programming. Which is horrible. Why? Because they are, of course, extremely clever at searching the Internet and finding something here and there. They then try to bring these things together to generate something to their surprise.
Whereas, each algorithm is a total artifact. More precisely speaking, it's a “computable function.” In older times, this used to be the essence of your work. It was the major act of creation.
In current gallery shows, we may see fantastically complex appearances that may be based on simple image-processing software, so that the breath-taking appearance hides the triviality of the software application. But I should be cautious: I am not familiar with the current exciting technology. This may have changed a bit; I don’t really know. When I was much younger, it dawned upon me what it meant to write a program. You should know that in 1963, this meant to develop machine language code. Not assembler, but one level down. The computer was naked. It did not even have an operating system.
We were caught between the algorithm and the machine. We were in a distance from the work we wanted to create. Once we had formulated the algorithm in a form suitable for the machine, we started it to run on the computer and stood there observing the drawing machine as it was doing its work.
Most likely, some error would show up. So, you would have to jump in again to solve the problem. Always, your question was: What detail did I forget? What did I do wrong again?
We experienced a fantastic, distant relationship between this kind of art, the machine, and ourselves. We became aware of us as the creator forcing a machine to do something we wanted to see and show to people.
Peter Bauman: What about the code itself? Because for Matrix Multiplications in the Tate permanent collection, you have the actual code sheet hanging with the outputs that it produced. The code even appears first so you are quite explicitly placing the code at least on an equal level with the results. Is the code the system? Or is the system the idea and the code itself perfunctory?
Frieder Nake: I'm thinking now of a person who is writing the entire code from scratch. If you like: Into the naked machine. Of course, now I depend to a large extent on the operating system. The operating system has become absolutely necessary, and without it, I cannot do anything.
In 1963, when I, for the first time, did anything in this direction, the computer available to me did not have an operating system yet. It had a very, very basic function that in the morning, when the machine was started, had to be put onto the machine before it could do anything. You may still know the term “assembler language?” It's a low-level programming language. There is just one level below the assembler—the machine itself. At that time, I was programming in machine code, which later became assembler language.
Therefore, I had a sheet of paper on which I organized the computer storage by hand: Storage cell #100 to contain this command, #101 something else and so forth. I put what I wanted the machine to do directly into those cells, addressing the storage cells themselves. Nobody is doing anything like this nowadays. It would be horrible.
But it was fantastic. The Computing Center allowed me to do this. I was working on the machine from eight in the evening to eight in the morning. I needed such long hours to get done anything of aesthetic relevance because the computer was so incredibly slow in comparison to today.
Peter Bauman: When did you sleep—during the days or did you just not?
Frieder Nake: No, I just didn't need any sleep as a young guy. These are all anecdotes. Which is to say, they were true and not quite true. My major belief is this: life is a series of anecdotes—your personal anecdotes that you go through. There is no sense in life. The sense of life is life itself.
Peter Bauman: That speaks to the randomness of life, which of course pervades your work. You're one of the few people in history who have created something that will live on for a very long time. How do you think about your legacy: algorithmic art today?
Frieder Nake: I’ll tell you something I'm very convinced of. Let us think of all the works that, for a number of decades, have been exhibited at great occasions and great institutions. Think of the many exhibitions of generative art during the second half of the 20th century and continuing into present times—almost a quarter century, again. Much of what was put up on the walls of galleries and museums of fine art during that time is precious, and the artists who made them contributed to the ever-growing world of fine art.
There is now no doubt of this as a revolution in the history of fine art. However, I now consider that to be the pre-history of algorithmic art only. That history has really started only recently. Just about twenty or thirty years ago. Everything before—including the second half of the 20th century—was just the pre-history of algorithmic art.
This is so because the products, the procedures and the processes that you create through algorithmic activity in a kind of pseudo-collaboration with the computer must be dynamic; its works must be moving.
The still image is still okay; it's nice as an exercise, as an assignment, as an expression of something you want to say or to show. But people and artists have done this for centuries—all the visual artists—with great results that we admire in awe. And we ourselves are still putting things up on walls for others to come and take a look at and like or dislike them. I myself owe a lot to all those who were friendly enough to present and exhibit the stuff I had done.
Now, art must be dynamic; it must be changing—changing permanently—nothing fixed anymore. The computer must be present and running as the work is happening. I'm not talking of something recorded on videotape, perhaps even generated by an algorithm. There must be more. The result must be displayed live on a large screen. That's what I would nowadays call “algorithmic art.”
Peter Bauman: There are some contemporary artists that have that same spirit today, creating what they refer to as real-time generative art, preferring to display it running live on large screens—artists such as Kim Asendorf, Andreas Gysin and Leander Herzog and their group AGH.
Frieder Nake: You know what is interesting? You several times, without any hesitation, use the term generative. I love this very much. Do you know the origin of it? Apparently, everybody is now using it. It was first used by Max Bense at the opening of the first exhibition of computer art at his institute in Stuttgart. The 5th of February, 1965. There and then, he coined that term. In the form of Generative Aesthetics.
Peter Bauman: You, of course, studied with Bense in Stuttgart at that time. How do you see your studies of cybernetic theory from then and its relationship to contemporary AI art? Much of today’s AI art is based on neural networks, designed to learn from themselves in a feedback-loop mechanism. These concepts of machine self-learning through feedback are directly rooted in cybernetic art and theory, which you studied with the great Max Bense.
Frieder Nake: What is interesting here is that history doesn't repeat itself. However, certain concepts may occasionally repeat. Concepts like cybernetics may have been used in some small, perhaps tiny little circles but didn't spread beyond those circles until the time was ripe, so to speak, to accept the new. In the case of generative art, it seems that now we need it. However, we should be careful. Generative grammar was defined by Noam Chomsky around 1955. He was more or less alone, claiming that there are generative mechanisms in grammar. He didn't use a computer to produce text, but some people very soon did so (Theo Lutz in Stuttgart, 1959). Chomsky showed how the concept of grammar could create rules that allow you to automatically generate correct sentences. He must have had a wonderful, looking-into-the-future brain by reducing linguistics to its very basics.
Linguists soon came up with more and more complex grammars that for many, were difficult to understand and, therefore, were even discriminated against as being trivial. When we use computers, we are always trivializing something that people have done before without really being able to explain what they are doing and why. They just did it.
Peter Bauman: A lot of early generative art was text-based—or music-based—because it relied on sets of rules. How did Chomsky’s ideas influence your own thinking?
Frieder Nake: When I came to Vancouver, they had hired me to teach computer graphics. Two days after arriving, I had a meeting with the chairman of the department. We talked about computer graphics, of course. He asked me, “What are you going to do?” And our conversation was going on. But when we were about to leave, he said, “Oh, by the way, we want you to teach a course in computational linguistics.” “What?” I had no real idea of what this could get me into. I realized that I had just four weeks to teach myself computational linguistics. It turned out that this was enough for me to learn enough to, at least, start a seminar. I felt honored by his trust in this young German being capable of doing such a seminar. In a way, it was a fantastic challenge. You need challenges in order to learn.
Peter Bauman: How long were you teaching in Vancouver? This must have been before you joined Bremen.
Frieder Nake: In a way, unfortunately, I was in Vancouver for only two years. I had actually emigrated from Germany. I did not want to return to this country. But then they started this university in Bremen. In Canada, a slightly reduced edition of a very good German weekly newspaper was published, which I often read. There I learned that a new university was about to be started in Bremen, based on, in some way, revolutionary pedagogical principles. So I wrote to them something like, “I hear there will be a university in Bremen. What should I do to apply?” This ended in me getting that position (where I got stuck, but love it). In my early Bremen years, I occasionally regretted having left Canada. Otherwise, I would now be a Canadian, I suppose.
Peter Bauman: Life gives us these unexpected, random paths.
Frieder Nake: This is another of those three terms that you have used several times: randomness.
Randomness is one of them, generative is the second, and algorithmic is the third. Picture them as the three vertices of a triangle. Within that triangle—between its three poles—is what we call algorithmic, digital, media art or whatever other term you may use. If you do something—resulting in visuals—that is influenced by randomness, algorithms and generating, we have set up the conditions for what is now algorithmic art.
We need randomness, which I take as the machinic correspondence to human intuition. I mean this as a little provocation. In the arts, you must provoke to some extent. You shouldn't do it with too heavy a hand. This would not be good, because we still want others to understand. But the question comes up now: what is it that we call “the artist's intuition?” The artist used to be standing in front of an easel with paints ready. They hold a brush, ready to dip it into the paint, thinking, “But heaven’s sake, what should I do now?” Think of Jackson Pollock. You have seen photographs of the situation: the large canvas on the floor. He is bending over it, not quite touching it. He is throwing the paint onto the canvas. Great. Revolutionary. Shouldn’t there be a bit of yellow somewhere, as some great surprise in the dark? For me, Pollock is a precursor to algorithmic art.
Peter Bauman: I think certainly he used randomness and he had some kind of system that he gave a certain degree of autonomy. How can these features of computation—like randomness, algorithms and generativity—teach us about what it means to be human?
Frieder Nake: As I was studying mathematics and, in particular, concentrating a bit on probability theory, I had the great luck to know a bit more than others and, in particular, about randomness. As part of this, I had my own random number generators. Everybody else in the community of algorithmic artists was using standard random number generators. Therefore, mine possessed different qualities. A good random number generator would run continually for, perhaps, a month before it would start to repeat itself. This would not really make a great difference; nobody would take notice of it. It would be an unrecognizable quantitative event. But, knowing this, I took some arrogant pride in being aware of the difference between quality and quantity in applying randomness in the decision processes of generative art.
Computers have no qualities whatsoever, only quantity. As humans, we are the great event in the world. We are those beings in the world who look for and are capable of experiencing the quality that we usually call feeling.
Peter Bauman: I wonder if machines will ever get to that point. In your present practice, what languages are you using or exploring?
Frieder Nake: For the last twenty years or, most likely, for much more than this, I have used nothing but Processing. I assume you have heard the name, of course.
It is so fantastic that Casey Reas, one of the two founders of Processing, is an artist himself. And co-founder Ben Fry is at least close enough to a kind of aesthetic or artistic thinking. They were still students when they constructed Processing explicitly as a programming language for artists and designers. Still following the traditional ways of thinking algorithmically, the two were close enough to the artists’ approach that it became an almost trivial act to write programs.
The fact that here, an artist who was an artist before he was using computers, together with his friend, defined that programming language, Processing, became an historic fact of great impact in the world of art. I should add: in that part of the world of art that is close enough to being aware of the cultural revolution that we call the Algorithmic Revolution.
Casey’s and Ben’s teacher at MIT’s Media Lab, John Maeda, introduced them to the world of algorithms and fine art. This situation is the background of Processing becoming the powerful tool or system that it is.
Processing is really simple. Ever since I myself became familiar with it, I've never taught anything else to my students. It is so powerful because it's based on Java. But you don't have to know Java, which would be much too complicated and cumbersome to learn. It would, most likely, discourage people from trying to write their own programs.
---
Frieder Nake is an artist and professor considered a founder of digital art. He produced his first digital work in 1963 and was among the first three to exhibit in 1965. His work has been shown globally and is part of the permanent collections at institutions such as Tate Britain, the Victoria & Albert Museum and the Tama Art University Museum.
Peter Bauman (Monk Antony) is Le Random's Editor-in-Chief.