 |
|
| all images © 1994 - 2008 Joel Murphy | |
We are surrounded by a technological density: electric lights, hammers, dishwashers, venetian blinds, cook stoves, refrigerators, plumbing, stereo systems, automobiles... We use some as prosthetics to extend our control or manipulate the world around us, some we wouldn't notice until we loose them-as in a power failure, and some, like computers have brought difficult challenges to understanding humans' relationship with machines. Computer technology is so complex that our engagement with it is often erroneously referred to as an 'interaction', terminology which encourages commission of the pathetic fallacy. The naive belief that these machines are intelligent and can attend a reciprocal relationship with humans perpetuates a lie that supports an unrealistic faith in machines' ability to provide original content or meaning. 2 In all of our button pushing and lever pulling we rightly have a relationship with machines and devices, but claiming a state of interactivity with even the most sophisticated machines is like demanding fidelity from a slave. When the direction of our engagement is down and away, we have a dictatorship where the rules are arbitrarily laid down by humans for machines to follow. Our relationship with the world of machines is about hierarchy and control: it's important that the device functions as predicted and when it doesn't we say it needs to be discarded or fixed. The groping attempt to focus on interactivity in art is defined in this way, and its gesture serves to enslave technology further while it incorrectly assures the viewer that a sophisticated level of engagement is taking place. There will be no opportunity for engagement in quality discourse with machines until they have the agency to defy us.
Observing Systems steps toward a resolution of the interactivity problem by taking the radical view that machines must indeed embody an autonomous identity. Jack Burnham introduces the notion of a Systems Aesthetic in art in Beyond Modern Sculpture , where he addresses the question of interface head on:
"The logical outcome of technology's influence on art... should be a series of art forms that manifest true intelligence, but perhaps more meaningfully, with a capacity for reciprocal relationships with human beings (in this case the word viewer seems quite antiquated). The need for intelligent response and self-recognition which we have instinctively sought, and sometimes found, in art will reappear in fantastically powerful forms." 3
And elsewhere:
"...the cultural obsession with the art object is slowly disappearing and being replaced by what might be called "systems consciousness." Actually, this shifts from the direct shaping of matter to a concern for organizing quantities of energy and information. Seen another way, it is a refocusing of aesthetic awareness--based on future scientific-technological evolution--on matter-energy-information exchanges and away from the invention of solid artifacts. These new systems prompt us not to look at the "skin" of objects, but at those meaningful relationships within and beyond their visible boundaries" 4
Central to the theme of a systems aesthetic is feedback technology and the science of Cybernetics. Without dumping the entire contents of the homeostatic container out onto the table, I want to address the seminal feature of cybernetics as a metascience which transgresses terminological boundaries between the sciences as its primary function. Cybernetics is "about relation, not essence," 5 therefore mechanical function (threshold sensor triggers motor) can be discussed in terms of biological function (behavioristic terms) through the use of analogy. 6 In a real sense, cybernetics is stopping just short of what Art has done for centuries: extend analogy through to metaphor for its own purpose. The developing science of robotics has shown that simple machines, with one or two sensorial inputs giving them the ability to navigate an environment, exhibit movements akin to behavior. Robots appear to demonstrate emotions, as in Grey Walter's Tortoises of the 1940s 7 , and many constructions at MIT exhibit insect-like activity. Machines engineered to simple but strict specifications can operate with such complexity that constructors find it useful to describe the performance in behavioral terms. This analogous transformation of machine into living or conscious thing is accompanied by a collision between science and emotion which fuels the fire driving a Systems Aesthetic in art. The point of that collision is an opportunity for us to redefine our relationship with technology from an authoritarian lording over, to one of reciprocity. Observing Systems is among the first steps toward providing technology an opportunity to define its own desires through self regulation and feedback. We can present machines with that opportunity by relinquishing our control over them. Only when technology becomes a willing participant in dialogue can meaningful or interesting interactivity take place.
Materials & Methods
The primary materials I'm using in this work are found domestic objects. The things we surround ourselves with out of necessity or habit, and have a comfortable physical connection with embody a cultural and personal meaning. We are all familiar with rolling the dishrack in and out of the dishwashing machine, the steamer basket folds and unfolds petal-like in a very satisfying way. Constructing them into robots makes use of the cultural baggage they contain, and that relevance amplifies the transformation of our relationship with the objects' meaning and character. A viewer is in a strong position to evaluate the gesture of a machine, when she has a close understanding of the source of that gesture.
A useful starting point toward true interaction is to invent a system of reciprocity among machines themselves. By constructing two groups of cybernetic robots which control each other, I'm presenting a situation where concepts of emergence, machine intelligence, consciousness, cognition, interactivity, etc. can operate in a creative artistic context. Among my goals is encouraging debate about these concepts and how closely they are tied to artistic expression. The dynamic equilibrium I anticipate resulting from vital pairing of machines within a feedback loop can be read as a direct response to Burnham's call for a systems aesthetic in the evolution of sculptural forms. By utilizing the skills of a prototyper and mechanical engineer as tools for art production, I'm connecting my work to a spontaneous entrepreneurial vitality with which I have consistently found myself engaged. This work aims to understand how machines converse with each other in the given constructed situation, each provided for in terms of sensors and actuators to effectively control both the others' environment and their own as mediated through the other.
Dishracks and their environment
In a room there are six dishracks of the type commonly found in standard dishwashing machines. Custom fitted with power supplies, drive motors, and legs, the Dishracks are roboticized and able to kick themselves across the floor. On board bumper switches allow the robots to change direction when they collide with an obstacle (the wall, or one of its sisters). The Dishracks are tethered to the ceiling with 25 foot coiled telephone handset cord, and rove in a space that is 10x15 feet. The telephone cords serve to conduct signals to and from individuals in the Dishrack colony. Incoming signals activate the robots' drive motors, and exiting signals communicate directions of their trajectory (fore / aft).
Steamers and their environment
In a room there are twelve stainless steel vegetable steamer baskets of the type with metal petals that unfold around the edges. Two drive motors and wheels are wired to batteries through photo cells and power transistors in an ipsilateral fashion (right photo cell controls left wheel, and vice versa). This design gives them a sensitivity to light which generally draws them toward any illuminated beacon. These robots are untethered, and travel on a California King sized platform of electrified metal panels (six foot by seven foot). As the individual Steamers move, they drag the edges of their metal bodies on the bed, thereby acting as switches between the electrified panels. This positional switching sends signals to individual Dishracks, turning them on and off. Above the robots hangs an inverted rotating clothesline. Attached to the line are 12 beacons constructed with flashlight parts. These beacons are switched on and off according to the directional bias of individual Dishracks. The clothesline is designed to rotate back and forth 360 degrees in an effort to help keep Steamers from getting 'stuck' in shadow.
In summary, these two colonies of robots are controlling each other. Information integral to their function is exchanged through a system of switches and wires that criss cross the architecture between their respective environments. In the Dishrack colony, when a robot changes direction a signal is sent through the coiled telephone handset cord, then along a wire to a room where the Steamers are, switching light beacons on or off. Similarly, as each Steamer drags its metal body along the floor, it conducts an electrical signal between the panels it brushes against producing a signal which travels to the Dishrack colony to turn on or off individual robots. Here the robots are given a simulated self reflection, and a vital [reciprocal] relationship with one another.
Because of the tightness of the feedback loop, it can also be said that the two colonies control their own environments, mediated through the other. The physical actions of the robots provides visceral clues to the meaning of this system's structure. Coupling that with output from the data mapping tools discussed below helps to define the interactions taking place.
Command Control Center
The device used to focus and control communication between the two colonies is a bank of 12 four-pole-double-throw relays (4PDT). Six are operated by the Steamer baskets position on the bed to direct the on/off position of the dishracks, and six are operated by the Dishracks fore/aft bias to illuminate 12 beacons above the steamers. The cause-effect trajectory is direct, for example, a given position of a steamer basket will connect two or more panels on the bed to switch a relay at the command control center which activates the drive motor of a specific dishrack.
From the system of the two colonies, Command Control Center relays are accessed on only one of their four poles. The extra poles are utilized in communicating command information from the two colonies to a third system which passively maps the outflowing data stream. Real time data, reflecting the colony pair as they manoeuver around one or more equilibrium points, is mapped in three different ways by machines which I shall refer to as lenses.
1-D lense
The first lense focuses every command aspect of the robotic system (24 0/1 potentials) toward a one dimensional representation. The device utilized is a bank of 64 solenoids which directly activate 64 keys of a piano (specs) centered around middle F. In order to magnify the 24 discreet on/off signals coming from the Control Communication Center, a mediating Window is introduced. The Window is, essentially, a device which translates a binary code into decimal output. For example, when counting in binary with three bits, the corresponding numbers in the decimal system are from zero to seven: eight discreet integers. The binary-decimal Window is made up of 24 4PDT relays, divided into eight three bit units, each utilizing three switching inputs directly from the robot system to produce 64 potential outputs which operate keys on the piano. With this design, the one dimensional mapping device activates eight of the 64 keys at any given time. While the magnification potential of the Window is somewhat obscured by its resolution, any lack thereof is more than made up for by a sophisticated compositional read of the acoustic output.
2-D lense
In constructing a two dimensional mapping tool to accommodate the volume of signals which represent the robot system, an Etch-A-Sketch tm is operated by two reversible 12VDC motors. Information streaming from the Control Communication Center directs each knob of Etch-A-Sketch tm in the following ways: on/off control of each motor, right/left or up/down by switching polarity of the signal to each motor, and the potential for high or low speed is directed by switching the on/off signal through a power resistor. Six such devices are used, calling for a total of 36 switching potentials, more than is available from the system as I have it set up. Solving this problem presents an opportunity for multiple combinations of control sources. One tablet will be controlled by dishracks exclusively, a second by steamers. The remaining four will be controlled by combinations of signals from the two groups. (See illustration x). The specificity of x y axes within the structure of an Etch-A-Sketch tm tablet provide an opportunity to examine the dialogic nature of the robots' interaction. The multiple images are viewed as a single drawing of behavior in the system, providing a more comprehensive view of the trajectory of the system.
3-D lense
By operating the 3-dimensional mapping device, the robots construct an object in aerosol foam (Dap ® Tex ® Insulating Foam Sealant). Two linear actuators (12vDC) position a foam output nozzle in vertical x,y space. A circular platform, serving as a base for the foam object, rotates underneath providing the variable z. Foam ejection is triggered by a solenoid activated lever. The machine's simplicity limits connection with the data stream to six control signals: four switches control linear actuators' power (on/off) and polarity bias (extend/retract); one powers the z axis motor; and one powers the foam trigger. In order to augment the connection, two or more relays in the Control Communication Center are wired in series to control the 3-D lense. This means that coincident events will have to occur in the robot environment in order for any action to take place at the 3-D lense station.
It is important to distinguish the work being done in the data mapping room. This aspect is not to be viewed as mere output from 'drawing machines', but as lenses focused on the robots' story of themselves. Raw data is always difficult for humans to interpret in a significant way. Use of instruments like the Etch-A-Sketch tm affords the viewer a familiar interface with which to engage a direct translation of the information. Sounds produced by the piano, for example, can be measured against known compositions in the medium, and judgements can be made about meaning. Giving the viewer an essentially artistic interface makes it possible to connect on an intuitive level with the expression of the system.
Installation
An economy of materials and design holds throughout the installation. Simple, direct solutions to problems of mechanical and electrical engineering are chosen in order to expose the working logic of the system. For example, the circuitry is transparent enough that one could follow and decode its logic simply by tracing wires through the installation. The 4PDT relays are packaged in a clear 'ice-cube' style housing, revealing the inner mechanisms' workings. The choice of plywood and 2x4s as primary construction material is made for its afford ability and mutability, as are the robots' structural base. The very structure of a vegetable steamer or dishwasher rack provide 'affordances' beyond their standard use. 8 As a prototyper of cybernetic machines, I'm taking advantage of their breadboard qualities, along with their idiosyncratic forms as design shortcuts. Adding economical cost and ready availability, makes them an ideal scaffolding for robot construction.
By building machines with the products as well as the materials of our electro-mechanical world, I am in a way collaborating with technology. To the point that, the design of the Steamer/SteamerBed switching connection was an answer to the question "How to get a Steamer to signal the Dishracks, given the behavior potential in its robotic design?", not an initial paradigm from the inception of the work. Similarly for the use of a clothesline for a light canopy over the steamers, the directional switches on the Dishracks. Once presented with the initial material (Dishracks and Steamers), a process of discovering solutions to the problem of actualizing a thought with that material ensued with great care in preserving the integrity of both.
Starting with the structural base of the robots, I've tried to keep as many variables of the installation wedded to the domestic space. For example, confining the dimensions of the Steamer bed to that of a California King Size, and using bulb housings from flashlights as beacons, provide non-arbitrary solutions to design issues that uphold a larger relevance. The commonality of sources from which the objects emerge lends constructive support to the interactions that will take place. These devices are already seeded with baggage before I animate them, and that gives the work a head start on its own empowerment. The viewer has some say by bringing his or her own histories and imperatives to the found object. The dishrack moves in and out of the washing machine in an intimate and satisfying way; the Etch-A-Sketch tm tablet is a familiar and awkwardly playful drawing tool. What happens when these devices are re-contextualized is conditioned by the viewers' casual domestic experience. Additionally the use of simple materials without secret panels or hidden microchips exposes the logic of the system as eminently readable. The wonder evoked by a machine actively pursuing its own desire is amplified by this bare construction technique, and behavior is more intensely embodied by the machines. This work places two mechanical systems in a command/control feedback cycle with each other, not to pitch them in battle, but to make discoveries about how they define themselves, and what that may mean to our relationship with them. If we are to ever find ourselves in an honest conversation with machines, there must first be a minimum level of trust and familiarity on both sides. The immediacy of material and transparency of circuit design invite curiosity and encourage discovery.
Owing to the size of the project, it required construction in sections over an extended period of time. Construction of the robots took place under firm deadlines. The Dishracks were completed in time for an academic review deadline in November of 2001 under the Title: OS Test Lab 11.01 , and the Steamers met the self imposed goal of mid February 2002 [ OS Test Lab 02.02 ]. In each case, processes were streamlined as much as possible, and problems of material and design were solved according to this 'deadline' strategy with which I've been experimenting: that no matter what, the piece will be completed and shown at the appointed time. The design and look of the entire installation is a function of this deadline strategy, it is a way I've found to expose the elegance of the process, and root it firmly to material choices. As a result of this process a significant parasite on optimum signal processing is embodied by entropy. The Dishracks underwent limited tests before their first exhibition, and only minor tuning before the final performance. The sensitive design of the steamers was equally sensitive to the wear of time.
Overall, Observing Systems took place in and around three venues of the Visual Arts Facility (VAF) at UCSD. Studio 237 provided an environment for the Dishracks. The Steamers, because of their need for specific lighting, inhabited the black box of the Performance Space, and the Data Mapping machines operated in the Marcuse Graduate Gallery. A covered atrium just outside the Performance Space was the location of the Command Control Center, at or near the center of the triangle described by these three rooms (see diag. on cover leaf). Command signals travel from the Control Communication Center 40 feet to the steamers and 90 feet to the dishracks via 24g color coded copper wire commonly used for telephone/data communication in homes and businesses. Multicolored signal wire will connect machines and interact with the architecture. While the placement of buildings and accesses to rooms will define the shape of the installation, gestural lines drawn in wire will define a novel logic to the structure of the VAF. Decisions regarding specific circuit construction (i.e. which dishrack will power which relay to switch which aspect of the 2-D lens, or which way to route wires) were left to the actual act of installation, thereby preserving a spontaneity in the way the work finally came together.
Installing OS was, to its constructors, a narrative one. Moving from the broad strokes of arraying the communication lines to setting up various rooms, to the tabs and sockets of logical relay connection, each task had its necessary precursor and follower. The process mimicked the structure in that its tasks were large and small finished wholes which came together to create something larger than the sum of themselves. While this process followed a smooth trajectory in order to define its anticipated outcome, the work was to defy any expectation of its actions in performance.
Observing Systems Debut
Thursday, April 23 2002
Due to what are considered electro-mechanical issues, the dishracks did not operate in performance with the activity level projected during OS Test Lab 11.01 (conducted in November of 2001). Symptomatic of this was the inability of the dishracks to lever themselves across the floor resulting from a lack of power, also the self-unscrewing of two drive legs (pee-pees). These occurrences (unexpected but accepted on the part of the author as the proverbial crumbling of the cookies) lead to initial instances of observer interference which then occurred at a variety of levels of the system throughout the performance. On the electro-mechanical level, observers entered the Dishracks' field of operation within Studio 237 and made attempts to re-attach, or 'fix' the pee-pees of the crippled individuals. On the vital system information level, light sensors on some of the Steamers were excited by independent light sources (flashlights) carried in to the Performance Space by observers. Also, Command Control Center switches, which function to power the entire system, were flipped off and on by at least one observer. On the parasitic level, manual activation of the foam spray lever. As a result the system mapping in the gallery was fundamentally unpredictable, and unreliable as representation of robot communication.
Possible reasons for observer interference (beyond the enigma of Quantum Mechanical wave/particle duality) may be due to a number of causes:
1) The lack of verbal or written directives.
2) The gestures made by the robots, especially the Dishracks, elicits pathos in a viewer.
3) Expectations about the robots' behavior based on visual clues of their physical structure and environments were thwarted.
4) Vandalism.
Viewer participation in the context of art is sometimes necessary: Anissa Mack's new work Pies for a passerby 4 for which she baked pies in a playhouse and placed them on the window sill in Brooklyn's Grand Army Plaza is an extreme invitation to spontaneous engagement with the artwork. However, Observing Systems was designed specifically not to be interactive. Rather, a distant observer is required for the system to construct itself in equilibrium. Through some fluke of material choice or what-have-you observers eagerly became intimate parts of the system, almost unfailingly in an effort to 'help' the robots become what they, the viewers, desired them to be. This could be read as undermining my initial thesis by solving the problem of interactivity in an undeniably visceral way, however I see it only affirming the point that humans will see in machines only their own desire, and act greedily in an effort to fulfill it. While I accept the first performance, I can do so only by deciding that the true communications of the machines were obscured by large amounts of environmental noise.
Observing Systems Encore
Friday, April 24 2002
after repair of electro-mechanical issues with the dishracks, the placement of tape strips over switches on CCC, and firm verbal enforcement of a 'do not touch' policy, the noise input stayed at negligible levels. Here, effects of the lighting adjustment in the clothesline above the steamers manifested themselves as the so called 'functional operating failure' during Friday night's performance. Withstanding that, all three mapping systems appeared to generate direct representations of behavioral patterns in the robots' equilibrium without evidence of known parasites or electro-mechanical signal failure/enhancement potentials.
Portraiture and Machine Intellect
One of the main threads throughout the work has been the nature of machine consciousness. A full discussion on the nature of consciousness is beyond the scope of this paper, but I would like to bring to light some thoughts that mesh well with Systems Aesthetic. In Theories of Consciousness , William Seager continuously butts against an intellectual brick wall which he refers to as the generation problem:
In his attempt to out maneuver the generation problem, Seager makes the claim that consciousness is a fundamental feature of the world, where he means by fundamental "something elemental, not dependent upon the satisfaction of any functional description by any physical system, and not subservient to the principle of causal grounding." 11 The idea of a universal force of consciousness leads inexorably to the theory of panpsychism, "the doctrine that 'all matter, or all nature, is itself psychical, or has a psychical aspect.'" 12 Is this the pathetic fallacy all over again? Seager rescues himself from that precipice by bestowing upon conscious forces the same enigmatic and non-scaling qualities attributed to the fundamental forces of quantum mechanics. These properties provide that consciousness won't manifest itself in an amoeba or a jellyroll as some fractional equivalent of what we perceive it as. For the tetherball and the skateboard, conscious experience may only be understood by humans as an incomprehensible absurdity akin to the wave-particle duality of the quantum realm. An examination of the Cogito Paradox , as stated by Stanislaw Lem, is helpful in deepening this point.
"We humans are unable to achieve complete certainty (as proof) as to whether a machine thinks and, in thinking, experiences its states as mental ones - since conceivably one may be dealing with nothing more than an externally perfect simulation whose internal correlative is a kind of void of total "soullessness". For their part, machines are similarly unable to obtain proof of wether we, as their partners, think consciously - as they do. Neither side knows what experiential states the other subsumes under the label 'consciousness.'" 13
Universal consciousness and Panpsychism, combined with cybernetics, provide a healthy medium in which to culture Burnham's systems aesthetic.
In addressing these concerns, I've created in the Dishrack-Steamer system a machine intellect, albeit a small one, not in order that it may be discovered, but that it may discover itself. By collaborating with a material based process of cybernetic construction as an initial step in understanding what machine thought might look or sound like, I may be in a way constructing a portrait. The installation of the communication wires reads very much like neruonal ganglia, which makes the Steamers and Dishracks in their environments become sense and/or motor organs, and the Data Mapping devices a metaphor of thought. Another rout would be to apply the Conscience / Sub-conscience paradigm onto the robot system / data mapping dichotomy. In any case, reading OS as portraiture is useful in that it is not an easy fit. The nature of the wedge I'm driving at is that a truly intelligent machine may be as unrecognizable to us as it is oxymoronic, and may be able to defy us in ways we could never comprehend.
Notes
1 This title is lifted from a work of essays by Heinz von Forester:
"It was not until the Macy Conferences had run their course, however, that von Forester tried to develop more fully the epistemological implications of including the observer as part of the system. The punning title of his essay collection, Observing Systems , announces reflexivity as a central theme. 'Observing' is what human systems do; in another sense, (human) systems themselves can be observed."[Hayles, Katherine How We Became Posthuman p. 133] It becomes evident in my description of the performance that this title correctly anticipated the significant participation of observers.
2 Sophisticated machines are designed to place the human user in a position of control through the use of feedback:
"The basic principle of feedback is that the user should always be kept informed about the state of the system, where they are within any navigational structure...what actions they have taken, and whether these actions have been successful. Feedback information given should be kept constantly up to date...When the system moves into a state which has consequences for the user's behavior--particularly in forcing them to wait...then clear and immediate feedback which gets the user's attention is important" [Baumann User Interface Design p.99].
The psychology of the 'user' is examined in Human Factors and Human Engineering studies. For example, the poptimum amount of time between user action and machine feedback, or with what type of feedback (sound, animation, etc.). (This design paradigm can also ironically place the human in a subservient postition of tending to the tasks of the machine. In either case, the relationship is not reciprocal.) But the very people who place machines in a subservient role through design choices talk about interaction in the sense of a 'human-machine dialogue ' [Baumann User Interface Design p. 250]. I feel this paradoxical design structure is at the heart of the interactivity problem3 Burnham, Jack Beyond Modern Sculpture p. 16-17
4 Burnham, Jack Beyond Modern Sculpture p. 369
5 See Hayles, Katherine How We Became Posthuman (pp. 97-100) for a discussion on use of analogy in cybernetics.
Also: Bowker, Geof How to Be Universal (pp.107-127) for how cybernetics operates as a 'metascience'.
6 Hayles, Katherine How We Became Posthuman p. 97.
7 Pickering, Andy The Tortoise Against Modernity , prepublication draft of paper delivered to the Max Planck Institute for the History of Science conference on Experimental Cultures: Configurations of Life Sciences, Art, and Technology . Berlin, 6-9 December 2001.
Aslo: Burnham, Jack. Beyond Modern Sculpture p. 335.
8 Gibson, J.J. Reasons for Realism . p. 401-418
9 Bell, Annie "Steal this Pie: Anissa Mack's art project is edible, iconic and free for the taking." Time Out New York . May 16-23 Also heard a piece on All Things Considered , National Public Radio.
10 Seager, William Theories of Consciousness p. 18
11 Seager, William Theories of Consciousness p. 240
12 Seager, William Theories of Consciousness p. 240
10 Lem, Stanislaw Imaginary Magnitude
References
Baumann, Konrad. 2001. User Interface Design for Electronic Appliances . New York: Taylor and Francis.
Bell, Annie. 2002. "Steal this Pie: Anissa Mack's art project is edible, iconic and free for the taking." Time Out New York . May 16-23
Bowker, Geof. 1993. "How to Be Universal: Some Cybernetic Strategies", 1943-70. Social Studies of Science v23 n1 (February):107-127
Braitenburg, Valentino. 1984. Vehicles: Experiments in Synthetic Psychology . Cambridge, Mass.: MIT Press.
Burnham, Jack. 1967. Beyond Modern Sculpture . New York: George Braziller.
Gibson, J.J. 1982. Reasons for Realism . Hillsdale, New Jersy: Lawrence Erlbaum Associates, Publishers
Hayles, Katherine.1999. How We Became Post Human: Virtual Bodies in Cybernetics, Literature, and Informatics .Chicago: Chicago Univ. Press.
Lem, Stanislaw. 1974. Cyberiad . New York: Seabury Press.
-----------. Mortal Engines.
-----------. 1990. Pirx The Pilot . New York: First Harvest
-----------. Imaginary Magnitude.
Mussa-Ivaldi, Sandro. 2000. "Connecting Brains to Robots: The Development of a Hybrid System for the Study of Learning in Neural Tissues." Artificial Life 6(4) (Fall): 307-324.
Penrose, L.S. 1959. "Self-Reproducing Machines." Scientific American 200 (June): 105-114
Pickering, Andy. 2001. "The Tortoise Against Modernity: Cybernetics as Science and Technology, Art and Entertainment". Max Planck Institure for the History of Science, Berlin. December 6-9
Rosenblueth, A. Wiener, N. Bigelow, J. 1943. "Behavior, Purpose and Teleology." Philosophy of Science vol.10 no.1
Seager, William. 1999. Theories of Consciousness: An Introduction and Assessment . New York: Routlidge.
Serres, Michael. 1982. The Parasite . Baltimore MD: Johns Hopkins University Press.