The sociobiologist Edward O. Wilson once said: “The love of complexity, without reductionism, creates art; with reductionism, it creates science.” Science and art, as two of humanity’s most creative cultures, although once expected to eliminate misunderstandings and interpretive hegemony, merging into a “third culture”, still maintain a clear dialogue between them, each following its own path. Therefore, complex science may serve as a way of thinking that expands our horizons and builds bridges for communication.
Complexity and Complex Systems
Complexity itself is a “complex” concept. In complexity science, there is a saying: “To understand the complex, one must first understand the complex.” However, we can still understand it from the perspective of order: complexity is a state between complete order and complete disorder.
The development of complex systems science can be described as having multiple origins and branches. Below are some major developmental stages:
1. Pre-conceptual Stage (～1930):
Evolutionary theory, statistical mechanics, and other abstract system laws.
2. Systems Science Stage (1930-1980):
Old Three Theories + New Three Theories
Information Theory (1948), Cybernetics (1948), Systems Theory
Chaos Theory (1963)
Complex Adaptive Systems (1968) / Autogenesis (Varela, 1972)
Dissipative Structures Theory (1977, Prigogine), Synergetics (1974), Catastrophe Theory (1972)
3. Complexity Science Stage (1980-1990):
Santa Fe Institute, computer modeling and simulation, addressing real-world problems that are difficult to solve through experimentation.
Emergence
Group behavior / levels of complex systems
4. Complex Networks Stage (2000-2010):
Skeletons and relational data acquisition.
Barabási (2012)
5. Big Data Era and the Physics of Complex Systems (2010～): In the big data era, data-driven rather than merely computational simulation. Mathematical equations describe unified laws. Biology’s Kleiber’s Law, Scale Theory (Geoffrey West, 2017)
Moreover, symmetry breaking, self-organization, and emergence are three very core concepts of complex systems, summarizing three ways complexity and complex systems emerge from space, time, and scale (real systems often possess all three, but may select a more appropriate representation). Complex science can transcend and integrate reductionism and systems theory, reestablishing connections with art, precisely because of these three methods.
Symmetry and Symmetry Breaking
If you ask what beauty is, throughout history, the most common answer would undoubtedly be order. Clearly, symmetry is a natural beauty, prevalent in the animal and plant kingdoms. Therefore, although human art particularly favors symmetry, we cannot say this is solely due to humans having a sense of balance in their inner ear or a pair of symmetrical eyes.
Philosophically, symmetry is a unity of individual and overall diversity, allowing the whole to be composed of the same elements in a regular repetition. So why do nature and ancient artisans favor symmetry? From a physics and mathematics standpoint, the reason things are as they are is that the simplicity of symmetry comes with the least “cost”.
However, symmetry can only be said to represent the most basic harmony. Beyond the complete unity of the whole and parts, and the greatest symmetry, there exist many non-isomorphic proportional divisions concerning forms involving time and growth. The most famous of these is the golden ratio. Thus, the idea that “beauty is harmony” can be further deepened into “beauty is proportion”: beauty is the harmony of the proportions of various parts of an object.
What is the relationship between the golden ratio and symmetry? From the perspective of condensed matter physics and complex systems, they are essentially a form of symmetry breaking. Symmetry breaking creates significant richness and complexity in things; in a certain sense, this intricate and complex world is a result of symmetry breaking. Therefore, if symmetry represents a classical eternal beauty, then symmetry breaking itself also represents a beauty that breaks conventions, diversity, and heterogeneity.
Self-Organized Criticality and the Edge of Chaos
When diversity arises from symmetry breaking, these elements or subjects may engage in local interactions, leading the system to produce some form of overall order, which is called self-organization (Spontaneous order), also known as spontaneous order in social sciences.
When sufficient energy is available, this process can be spontaneous, requiring no external control. It is typically triggered by seemingly random fluctuations and amplified by positive feedback. The resulting self-organization is completely decentralized, distributed across all components of the system. Thus, self-organization is often robust, capable of surviving or self-repairing severe disruptions.
Self-organization generates complexity or complex systems, mainly through dynamical system behavior over time, often based on simple rules, reaching a critical point as an attractor after sufficient evolution, thus presenting a certain order distinct from the past. This process is also referred to as self-organized criticality.
Chaos refers to systems where extremely minor inaccuracies in measuring initial positions and momenta can lead to huge errors in long-term predictions. This is commonly known as “sensitivity to initial conditions.” It directly denies Laplace’s determinism, akin to the quantum uncertainty principle.
Chaotic phenomena have been observed in many systems, including cardiac arrhythmias, turbulence, circuits, water droplets, and many other seemingly unrelated phenomena, with the most famous being the so-called “butterfly effect”: the flapping of a butterfly’s wings in the Amazon may trigger a hurricane in Texas.
Multiple Emergence and Strong Emergence
From the diverse elements or subjects arising from symmetry breaking to self-organization among them, producing self-organized criticality and the edge of chaos, if the entire system is a multi-level system, then it is possible to produce new system properties or system levels beyond the original observation scale, a phenomenon known as emergence.
From the perspective of complexity, just as with symmetry, each new type of complex system corresponds to a new complexity, meaning that the term “complexity” is always in an unfinished state. For discovered natural or artificial systems, we can seek new complexities to characterize them; conversely, we can utilize known symmetries and complexities to create and generate new systems, including various types of art.
Cyborgs and Cross-Media Art
The renowned art historian and theorist E.H. Gombrich stated in “The Sense of Order”: Regardless of whether it is poetry, music, dance, architecture, calligraphy, or any craft, it demonstrates that humans enjoy rhythm, order, and complexity.
In natural language and formal language, if a sentence directly or indirectly refers to itself, it is called self-reference. If it directly or indirectly points to itself or its class, it is called recursion.
We can see that the reason human natural language has such a powerful expressive capability, to the extent that “the limits of my language mean the limits of my world”, is due to the existence of recursive and self-referential mechanisms. This also enables language itself to accommodate various modal expressions; for example, poetry can express one sensory experience through another.
In cybernetics, whether human, animal, or machine, they are all merely components of information theory. Therefore, when I give commands to a machine, it is not fundamentally different from giving commands to a person. — Wiener, “Cybernetics in History”
Through information theory and cybernetics, we can discover that information itself can be a unified abstract language that transcends many systems; whether it is system state storage or causal action transmission, it can be accomplished through information processing. This provides a foundational basis for cross-media art.
Turing Machines and Generative Art:
AI, Evolutionary Art, and Artificial Life
Information can serve as a unified language for the operation and description of all systems, and abstract processing of information is no different from the information processes of natural systems. The Turing machine invented by Alan Mathison Turing constitutes a universal simulator for any information and computational process (with information loss).
Using the Turing machine as a conceptual prototype, John von Neumann designed and built electronic computers that can process various types of information from the natural world, including information based on different sensory and physical modalities, such as visual, audio, semantic, etc.; at the same time, according to computability theory, the computational capacity of the human brain is equivalent to that of a Turing machine. This means that humans can obtain an intelligent system capable of unifying all media information into pure digital media for processing.
Turing machines: the embodiment of mechanical judgment processes, can serve as information universal processing devices for cross-media art, based solely on information transmission commands, enabling various media and subject systems to collaborate. The next step as a complex system is for artists to utilize or create a system, using the system itself for artistic creation. Among these, digital media is central, and in a sense, digital serves as a so-called meta-medium (atoms → bits). However, not all digital art is cross-media or cross-system; for example, video or photography, which primarily records physical world information, still uses digital media as a mapping tool rather than a creative tool based on its own.
The most typical form of digital media-based art is generative art (Generative art). The most typical definition of generative art comes from Professor Philip Galanter of New York University:
Generative art refers to any art practice where the artist uses a system, such as a set of natural language rules, a computer program, a machine, or other procedural invention, which is set into motion with some degree of autonomy contributing to or resulting in a completed work of art. — Philip Galanter
It can be seen that, unlike creating art using physical media or natural language, generative art uses code language and automatic algorithms for artwork creation, thus the code and algorithm themselves cross the two media or systems with the final produced process and result. Since generative art often starts from a series of algorithms, computer programs, natural language rules, or even mathematical equations, typical generative art can be summarized as the process of formal systems generating non-formal systems, such as chaos/fractals/L-systems/generative grammar/rule systems/cellular automata/Game of Life/reaction-diffusion systems, etc. Of course, if we remove the restriction of automatically running code and algorithms, then art produced based on rules can be regarded as generative art; for example, classical art’s use of visual or other symmetries, patterns, textures, and repetitions, or the use of mathematics and geometry for composition and arrangement. In the broad sense of generative art, there exists a large category of art that simulates the human brain using neural network algorithms, which can be termed artificial intelligence art. For example, art generated and created based on neural networks such as RNN/LSTM/GAN/Diffusion, or large models with billions of parameters (Text-to-image/multimodal, etc.). Additionally, when generative art and artificial intelligence art draw inspiration from mathematical algorithms and brain neural networks respectively, and shine as creative methods, another large category influenced by natural-inspired algorithms is evolutionary art (Evolutionary Art), which is quietly developing. In the context of complexism, which blends science and the humanities, evolutionary art becomes a new dynamic portraiture, creating various artworks with high effective entropy complexity comparable to natural systems through adjusting parameters, selecting fitness functions, and studying genotype-phenotype mapping, widely utilizing cluster systems, ant colony algorithms, genetic algorithms, genetic programming, self-organization, and emergence.
Therefore, it can be said that narrow generative art, or AI art based on neural networks, represents two types of broadly generative art specialized in algorithmic systems at the rule and non-rule ends, while evolutionary art lies between rules and non-rules. As mentioned earlier, this is a system of the highest complexity, thus it may give rise to highly intelligent autonomous subjects or collective intelligences, creating art through these subjects beyond the artist. This is art of artificial life or robotic art.
A typical evolutionary art process is as follows:
Evolutionary Art: Shaping the final phenotypic effects by adjusting initial parameters and fitness functions
Evolutionary Art: Miguel Chevalier: EXTRA-NATURAL 2021
Artificial Life: Karolina Sobecka and Jim George “Sniff”, 2009
Digital Ecosystem: Joan Soler-Adillon “Digital Babylon”, 2005
Here, we can summarize several types of art from the perspective of complex systems:
Generative Art:
Creation of works using code language or automatic algorithms.
AI Art:
Creators create an autonomous intelligent subject using various media and algorithms.
Evolutionary Art & AI & Robotic Art:
Creators create an intelligent subject capable of creating art.
Complex Art: Biology, Ecology, Cryptography, and Others
It can be seen that based on Turing machines and artificial intelligence systems, generative art has profoundly changed the concepts and methods of artistic creation:
1. Creative Subject:
No longer limited to humans, but may include algorithms, machines (the symmetry breaking of creators).
2. Process Controllability:
The creative process is no longer controllable by humans, introducing real-time, interactivity, autonomy, evolution, and other features;
3. Result Predictability:
Due to the introduction of randomness and the aforementioned interactivity, the results are almost completely unpredictable and emergent.
Furthermore, from the perspective of evolutionary art, autonomous intelligent agents and human-machine collaborative creation will lead people to re-examine the art and creation process itself.
Contemporary art itself has become a process of multi-body collaborative interaction and symbiotic evolution. For instance, in terms of creative subjects, besides humans and autonomous machines, it also includes existing biologies, nature, and other ecological subjects of the Earth (such as slime molds); regarding collaborative spaces, scales, and methods, it also includes larger interactive collaborative processes among human creators, such as encrypted art based on contract systems, game art connecting virtual and physical spaces, and various design co-creation performance arts.
Biological Art:
Creation using other naturally existing organisms as media and materials.
Ecological Art:
Creation using natural ecological systems as media.
Cryptographic Art:
Creation involving numerous user subjects and digital objects based on blockchain contract systems.
Game Art:
Creation of new orders through collaboration among various subjects, including humans, based on rules within a certain spatiotemporal range.
Below are some typical examples of this kind of art:
Robert Smithson, Broken Circle / Spiral Hill
(1971, located in Emmen, Netherlands)
Environmental Art / Ecological Art / Earth Art, Multiple Systems and Emergence
Laurent Minio and Krista Zomerel,
“Insect Man” 2019, Interactive Installation,
Creating larger mutual systems using the life system itself as material.
PaK: The Merge, Cryptographic Art, using smart contracts to mobilize various subjects for purchasing, resulting in an unpredictable emergent outcome
Conclusion: The Future of Complex Art
We can see that from the perspective of complexity, viewing contemporary art reveals that art is not only an emergence, but also that the aesthetic experience of art is a co-emergence of consciousness and art. Art has evolved from classical static observation and imitation to modern expression and interaction, often accompanied by processes such as symmetry breaking, self-organization, and emergence in complex systems. The integration of complex science and art allows art to largely break free from the influence of reductionism, embracing more chaos and uncertainty. In the book “Art in the Age of Emergence” published in 2017, it is argued that the complex system characteristics described by emergence, and the relationship between emergence and consciousness, based both on scientific research and providing space for spirituality, can be particularly suitable for understanding the process of artistic creation, thus filling the void in understanding art under postmodernism. In the 2008 book “The Art of Artificial Evolution: A Handbook on Evolutionary Art and Music”, it was also suggested that complexism could serve as a new practical program for evolutionary art. Although we are unclear whether the attempts to combine complex science and art in concepts and practices can promote a new understanding and development of art, at least we have seen this possibility, as well as the new directions that are currently happening and being explored. The transformations and explorations occurring in various art fields are all worthy of our attention and exploration.
Written by | Shisanwei
Source: Excerpt from the original article of WeChat public account “Jizhi Club”