Autopoiesis
David Halsell
In the early 1970s, Chilean biologists Humberto Maturana and Francisco Varela sought an accurate definition of life when they first formulated a theory to distinguish living systems. This led to the idea of autopoiesis, (literally "self-producing"), based in the "systems theory" approach to theoretical biology. The theory of autopoiesis describes a closed system that is autonomous from, yet interactive with, its environment. When stimulated, this living system will reorganize itself internally; yet always retain an order that makes it unique, reflexive, and self-perpetuating.
The cell, a simple biological machine, is a useful model in autopoietic theory. As a closed system with permeable boundaries, it is both an autonomous entity but also coupled with its environment. Its functions are the result of mechanical interoperations within the cell. These mechanisms give the entity its character and behavior, based on the interactions and physical limitations of its component structures. As a cell grows and lives, the development of all the functionality of life comes from itself and continues to do so until it dies–that is, the autopoietic cycle ceases. The cell takes in chemicals for growth, but to the cell’s components there is no real difference between food and a toxin–they are both perturbations–affecting the efficacy of self-propagation, favorably or not. In response to perturbations by the environment or medium in which it exists, an entity will adapt or evolve to maintain the structural integrity of its components–which is the cyclical nature of autopoiesis (Winograd & Flores 1986, 45). When multiple individual entities interact, these same structural processes continue symbiotically and in self-similarity to more and more complex, higher-order systems (such as neurological systems), which is perhaps one of the more difficult, but epistemologically interesting aspects of the theory. Autopoietic theory suggests a structurally-based explanation of cognition and the development of language. It is here that the rather esoteric but elegant idea of autopoiesis meets art.
As noted, autopoietic theory grew out of systems theory, which itself has had precedents in the arts. One of the earliest artists to explore systems theory was Hans Haacke, who directly engaged these ideas in works such as 1964’s Condensation Cube. Art historian Jack Burnham explored these manifestations in curating the seminal 1970 exhibit Software, and with books such as Beyond Modern Sculpture. Other permutations of systems theory have had a great impact on the technological arts and the development of theoretical computing, such as cybernetics.
Artists, especially those in the plastic arts, have always wanted to capture the essence of life. Today, this legacy is embodied by explorations of Artificial Life and Artificial Intelligence. One of the prime concerns in these ideas is the search for a cohesive theory to understand how life was created and how it sustains. Artists such as Simon Penny, Yves Amu Klein, Joel Slayton, Louis Bec, and Sommerer & Mignonneau are working with Artificial Life artworks, exploring innovative heuristic models such as genetic algorithms, neural networks and complex-adaptive systems. In Artificial Life research and artwork production, artist Ken Rinaldo has noted the existence of two schools of thought: the "hard" variety, in which the artist attempts to actually produce life, and the "soft" variety, in which a simulation of life is sought (Rinaldo 2001).
The artwork Autopoiesis by Ken Rinaldo is a large-scale interactive robotic installation piece designed to emulate the theory of its namesake. Consisting of audio and video projections and fifteen mechanized structures that hang in an exhibition space like humanoid arms, it is at once both organic and mechanical. When the viewer/participant moves amongst these structures–which are bigger than the average adult but dangle several feet off the ground–they respond as if "alive," smoothly swinging and gesticulating. A tiny video camera is mounted in the tips of two of the arms, projecting signals onto the walls, as if to "see" from its viewpoint. To give a further sense of sentience, these devices use telephone tones as a "language" to "communicate" among themselves. On each arm, a series of light-emitting diodes signals the status of information input and exchange among the group. Computer-controlled feedback loops, smart sensor configurations, and randomization algorithms produce and control its movement. Through the interaction of the viewer/participant, the piece seemingly evolves, producing unexpected, emergent, behavioral gestures and emotive sounds. Depending on the conditions of interaction, at times it seems to act as though it were a single entity, and at other times, as independent, individual parts, leaving the viewer/participant with the feeling that he is interacting with a sentient machine. The installation succeeds in provocatively addressing our concepts and perceptions on the nature of existence.
The functional workings of Autopoiesis are a technical accomplishment in and of themselves. The quick, lifelike responsiveness and the low predictability of repetitive behaviors are a particularly illustrative example of successful interactivity in an artwork. A network of infrared sensors responds to people moving amongst the arms, coordinating their movements depending on the conditions. At the top of each arm are four passive infrared sensors that are configured to overlap their range and provide eight zones of feedback, doubling their efficiency. This economy of design is supplemented by an active infrared sensor in the tip of the arm that provides the specific proximity data to the viewer/participant. By the resultant push/pull of the sensory data, a multitude of subtle and complex behavior responses are allowed.
The piece is also quite metaphorically autopoietic, in that its structure determines a limited, albeit very complex, range of possible behavioral states. This sophisticated level of interaction regulates the "experience" of the system, similar to the coupling between an autopoietic system and its environment. The interactions of the robots are defined within a set group of rules for their behavior individually and collectively, which could be considered their structural identities. With a provocation, actions elicited are generated not by the source of stimulus, but only by the possibilities inherent in the robot’s programming repertoire and material constraints. This structural identity in the physical sense is what defines the structural identity of actions, and a central notion to autopoietic theory.
Fundamentally, autopoiesis is not only a theory of the nature of living systems; it is also a theory of the development of cognition and perception. The linguistic and communicative aspect in Rinaldo’s Autopoiesis is not actually its audio component, but in the networked data packets distributed among the fifteen robot processors and the global controller that exchange sensory information and drives evolutionary behavioral change. This is a rudimentary step towards what Maturana calls linguistic behavior, a situation where organisms engage in a coordinated activity of communication that is not yet a language.
One of Rinaldo’s goals in the piece was the disruption of the typical user interface of computers and robots. As in all art, the interface between the user and the artwork is manifest in the physical space, but actually takes place on the cognitive level. In Autopoiesis, the user is confronting a robotic entity that is indeed responding in real time and in a seemingly sentient manner. This, according to autopoietic, the theory, is one of the elements of cognition. Two neurons in the brain, structurally dependent and coupled by their mutual environment, will rearrange themselves when perturbed, but can only do so within the limits of the range of possibilities defined by their structural makeup. The physical and mental domains are unified and one makes the other possible. In a cyclical process of perturbation and reorganization, the user and the robot are in a similar interaction but on a grander scale. The user is a closed system that senses and reorganizes itself to the environmental conditions (in this case a machine that is emulating a similar autopoietic process). While the machine is rudimentary in comparison, both parties are sensing and responding–inducing an exchange of stimulation that is both highly interactive, artistic, and demonstrative of the idea of autopoiesis.
The process of transliteration of the theory into the piece gives insights into the artistic process of exploring the boundaries of Artificial Life/Intelligence research. Just as learning is an autopoietic process, so too is the creative one. From an autopoietic perspective, Rinaldo’s piece is not communicating information, it is adapting to perturbations in its environment and at the same time perturbing the internal organizational systems of the viewers. The interaction of the viewer/participant brings information to the piece, and in so doing the viewer/participant creates the aesthetic and form by their presence.
As a conceptual inquiry, Rinaldo has demonstrated some of the key issues of the theory. We can conclude that the artwork Autopoiesis, (which might be classified as "soft" Artificial Life with a physical presence) is not autopoietic in the true sense, but an allegory of a living organism using the principles of Artificial Life, Due to its rigorous conceptual underpinnings and technical complexity, this work is a substantive contribution to this area of cultural discourse.
In autopoietic theory specifically, and systems theory in general, the idea of evolution is central. Burnham recognized the relevancy of evolution when he recognized art’s ability to prepare society for future social and biological evolutionary changes. (Burnham 1968, 314). Because of our self-awareness, this evolution takes place in our psychic and social domains in a relatively short time. The similarities between the autopoietic process of perturbation-reorganization-perpetuation and Burnham’s views on cultural evolution are striking, at least on a metaphorical level.
Artists are seeking a cohesive theory to understand how life was created and how it sustains itself. Society has now reached a level of conceptual sophistication and mastery of technological tools to support greater artistic exploration of these ideas. The recursive nature of the arts, the self-referential, cyclical nature of ideas and processes, is reflected in both the conceptual and practical considerations of autopoiesis. The artist who attempts this type of cultural dialogue must conceptualize as an artist, design as an engineer, and maintain the procedural rigor of the scientist. This cross-disciplinary practice is what makes technically and conceptually complicated artworks such as Autopoiesis both interesting and possible. Artists continue to cross boundaries, pushing the limits of our understanding, with some of the more enticing intellectual models of how we, and the universe around us, work. Art such as Rinaldo’s, and the work of others exploring similar ideas, present fresh perspectives on the questions of perception and consciousness–what it means to be alive.
