Jonathan McCabe
CHAPTER I — CELL
Artist 2 of 2
Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here. Intro by Kate goes here.
KV: Conway's Game of Life demonstrated that extraordinary complexity can arise from a handful of simple rules. How has that idea influenced your own thinking, and where do you see your work extending Conway's original proposition?
JMC: I first came across Conway’s Game of Life in the famous article in October 1970 Scientific American by Martin Gardner (Conway's Game of Life: Scientific American, October 1970) sometime in the late 1970s. I was struck by the examples of “gliders” that transform into themselves in a different position on the grid, maintaining an identity and moving by transformation. The Game of Life demonstrates that the potential complexity of a system is not necessarily due to a complicated process, but rather a complicated state undergoing transformations by what can be the simplest of rules. Alan Turing showed that a simple computing device, a “Turing Machine” can be constructed that can do whatever any computer of any complexity can do, given enough time and data storage. Stephen Wolfram has done a lot of work in this area, and I was very influenced by his work classifying one-dimensional cellular automata, where he divided them by behaviour into four classes, with his “Class 4” giving complex behaviour.
There is an idea called the “Edge of Chaos” originating from Norman Packard in the 1980s that many systems can have an adjustable tendency towards order or chaos and that the most interesting behaviour occurs on the ordered side near the transition to chaos. Brains seem to continually adjust to be near chaotic levels of feedback where the greatest sensitivity and potential for different behaviours occurs. Some people speculate that living systems are also in this region of behaviour. Conway’s rules give a system that has this kind of behaviour, and it has been shown that it is possible to build computers in his game that can do anything an ordinary computer can do, including running the Game of Life. The idea of a balance between Order and Chaos occurs in many religions (for example, Vishnu (Order) and Shiva (Chaos) in Hinduism) and in popular culture, for example in stories by Michael Moorcock and the novella Great Work of Time by John Crowley.
Conway described his system as a solitary game, where a person carries out the time steps according to the rules. This was at a time when computers were not widely available. It’s a strange kind of game because there is only choice about the initial conditions and no choices are made during game play. Conway’s system is an example of a “cellular automata”, which can be traced back at least as far as Stanislaw Ulam and John von Neumann in the 1940s. The representation of space as a grid and time in discrete steps were already embodied in games like Go and Chess, which are very old.
My work in the Cellularity series is descended from Conway’s Game of Life. Space is represented by a discrete grid, mapped to pixels in the resulting image. Time is in discrete steps. The state of each cell depends on its state in the previous time step as well as the states of its nearest neighbours in the previous time step, the same as in Conway’s game. In the Game of Life each “cell” (position in the grid) can be alive or dead, a single bit of information. In Cellularity each cell has a state consisting of four numbers (in 32-bit floating-point format). A time step involves averaging a cell’s values with its nearest neighbours (equivalent to diffusion) and then jumping to a new position in the four-dimensional space represented by the four numbers (equivalent to a reaction). The jump is based on polynomials of the four numbers. The polynomials change depending on the previous history of the four numbers, so the simple rules are contingent on the history, giving regions with different behaviour. So, you can see that Cellularity is an elaboration on Conway’s Game of Life, which is a variation of the game of Go, which is at least two and a half thousand years old.
Kate Vass: Your work often produces forms that feel unmistakably alive, yet they emerge entirely from mathematical processes. At what point does a pattern stop feeling computational and begin feeling biological?
Jonathan McCabe: There’s an essay from 1960 titled “The Unreasonable Effectiveness of Mathematics in the Natural Sciences” where Eugene Wigner asks why our theories of the natural world are often describable in simple mathematical forms. It’s not obvious on the face of it why this should be so. Many people, for example Max Tegmark, John Wheeler and Konrad Zuse, argue that information or mathematics is the actual basis of reality, what the world is made of. Atoms themselves can be described as mathematical processes and be simulated on digital computers in the field of computational chemistry, and all living things that we know of are sets of atoms. From this point of view living things can be described as a type of mathematical process.
Early computer art with its simple lines and polygons is clearly non-biological, and any raster image if you look too closely becomes a collection of coloured squares. The computer art that looks more biological tends to have structure at different scales (a “fractal” look). My work with reaction-diffusion systems produces rounded forms for a different reason (the smoothing nature of diffusion) to why cells and their substructures are rounded (surface tension).
KV: Your reaction-diffusion systems often resemble plants, membranes, or microscopic organisms, despite never being programmed to imitate nature. Why do you think mathematics repeatedly converges on biological aesthetics?
JMC: The reaction-diffusion systems have regions of different immiscible stable phases or cycles of the discrete dynamical system and the edges between them form a kind of one-dimensional membrane. Cells divide the inside and outside worlds by a membrane which is a two-dimensional liquid, a lipid bilayer. Both kinds of membrane have a tension that minimises curvature. I’d say it is a coincidence that these systems look similar.
The Evotrees have a branching structure that is like some ferns or sea creatures or feathers. They are related to L-Systems by biologist Aristid Lindenmayer and context free grammars as described by Noam Chomsky. So, they are actually based on theories of language and the structure of biological organisms. Each Evotree is a remarkably long sentence in a language of movements and rotations in three dimensions.
KV: Many of your works exist as computational processes rather than fixed objects. What do you consider to be the artwork: the algorithm, the execution, the rendered image, or the experience of watching the system evolve?
JMC: I see myself as the author of the program which embodies the algorithm that when run generates the image. Of course, the algorithms have a very long history in our shared culture, traceable back to at least Chinese, Indian and Arabic cultures. Generative art has similarities to photography, where the photographer sets up the shot, choosing the place and time and technical details like exposure and lens aperture, but then lets a process generate the image. I consider the artwork to be the whole thing, the images can’t exist without the writing of the text which is the program.
KV: As digital and software-based works become increasingly important to art history, how should we think about their long-term preservation? Is preserving the code enough, or must we also preserve the environment in which it runs?
JMC: Now this is a very interesting question. I’m reluctant to release my programs because most of them are at least thirty years old and I haven’t had many new ideas since then. I have from time to time described my algorithms precisely in ordinary language and they have been turned to programs and released by other people. I’ve then found it disconcerting to come across images on the internet that I would recognise as my work but ascribed to other people. For example “Multi-Scale Turing Patterns” now has a life independent of me and exists as a propagating idea, which bodes well for its long-term survival.
I would say we must preserve the environments that the programs run in. The programs are not self-contained and don’t fully describe the process but rely on many unstated properties of the environment. For example, in the C programming language the precision of the number formats is not stated in the program but is a property of the “headers” and “libraries” used by the compiler. An environment with different precisions in any of the floating-point number formats may well give a different output for the same program. Even which way is up for an image and the order of the primary colours (RGB or GBR) are conventions which are not part of the program.
At the moment we can make a good guess about what kind of computer, operating system, libraries etc. a program was run on, but it will become harder with time. Bundling everything as a virtual machine file might help any future creatures with an interest in 21st century generative art.
KV: The exhibition ultimately asks a single question: Can emergence itself become conscious? After spending years building systems that organize themselves without central control, where do you stand? Do you think emergence alone could ever cross that threshold?
JMC: The definition of consciousness has kept philosophers busy for ages. My understanding is that it can come in various amounts and be added together or divided. For example, some people were operated upon to sever the major connection between the two hemispheres of the brain so that an attack of epilepsy would not propagate to the other hemisphere. It seems that these “split brain patients” support two consciousnesses, which can have disagreements where, for example, the two arms and hands act independently to differing ends (like one hand dressing the body while the other hand is undressing). The brain can be further subdivided (by a mad scientist trying to prove a point) into disconnected pieces of consciousness. This supports the idea of panpsychism, where consciousness is a property of the smallest parts, and is added together and arranged to make the larger consciousnesses we enjoy. Adding together any number of zeros results in zero, so if consciousness can be added together it must exist in the parts in some amount. In this view there is no threshold to cross. Any system would be conscious in some way by some amount. A consciousness comparable to ours would probably have many parts which are connected and influence each other in a certain way. The strength of the influence is crucial, if it’s too weak the system is nearly a pile of independent parts (we wouldn’t expect a set of non-interacting objects to have a unified consciousness) , if it’s too strong the system will be like a single part because the states of the parts will be highly correlated or anti-correlated. This viewpoint is compatible with Integrated Information Theory as described by Giulio Tononi in 2004.
On the other hand, consciousness is often seen as an emergent property of a complex system, in which case there can be a threshold due to the size of the system. Such a threshold might appeal to people who want to think that humans are conscious and animals are not, but don’t mention the sizes of elephant or whale brains to them.
In either case consciousness could arise in a suitable system. It would probably be very alien to us and we wouldn’t recognise it.
NEXT: STATE II - PATTERN
11th August