Cybernetics

views updated

CYBERNETICS

Cybernetics is defined classically as the study of "control and communication in the animal and the machine" (Wiener 1948). After the decline of classical cybernetics, the field underwent a rebirth as "second-order cybernetics" in the early 1970s. Second-order cybernetics is more closely and more obviously involved with ethics than classical cybernetics (and certainly promotes a radically different worldview), but both have important contributions to make to reflections on science, technology, and ethics. Cyberculture, an increasingly important phenomenon that includes elements as diverse as email and chat rooms, electronic commerce and gaming, virtual reality and digital politics, has its origins not just in computers but also in the lesser known field of cybernetics (from which it takes its name).

Cybernetics

Cybernetics was originally promoted by the mathematician Norbert Wiener (1894–1964) in his 1948 book of that name (although W. Ross Ashby's 1956 book, An Introduction to Cybernetics, is considered the classic introductory text). The terms of cybernetics (including goals and purposiveness, feedback, and mechanism as metaphor) had been previously used, as was the concept of control as attaining and maintaining desired states, rather than restricting the actions of others—but not as concepts forged into a coherent field. In the development of cybernetics, two groups were particularly important: the informal association of Wiener, Arturo Rosenblueth (1900–1970), and Julian Bigelow (1913–2003) at the Massachusetts Institute of Technology (MIT); and the Josiah Macy Jr. Foundation meetings on "Circular, Causal, and Feedback Mechanisms" (which assumed the supertitle "Cybernetics" after the publication of Wiener's book), which included Warren McCulloch (1898–1969), Walter Pitts (1923–1969), Margaret Mead (1901–1978), Gregory Bateson (1904–1980), Heinz von Foerster (1911–2002), and Wiener and Rosenblueth.

The term cybernetics was derived from the Greek kybernetes, meaning "helmsman," and the field initially examined the behavior of (often complex) systems to develop models for improving system performance. The models were based on a notion of universally applicable mechanism: No essential differentiation was made between animate and inanimate systems. Examination of behaviors meant that systems which seemed impossibly complex or obscure no longer needed to remain so. If cyberneticians could not see what constituted a system, they could treat the system as a black box, which, through careful study of the inputs and consequent outputs, could be notionally "whitened" to the point that a viable mechanism relating input and output could be imagined, even if the actual mechanism remained unknown.

The intention was that systems would become controllable or better able to achieve the aims for which they were intended. The systems that cyberneticians studied were assumed to have observer-defined goals. Potential for error was understood to be omnipresent. To correct an aberration in the behavior of a system, differences between the (hypothesized) goal and behavior were examined, and the system adjusted to compensate for any difference (error). The process of error determination and correction continued until the system began to attain (and continue to attain) its goal.

Although the physical systems initially considered by cyberneticians were military and mechanical (starting with antiaircraft guns and developed through W. Grey Walter's electronic "tortoise" and Ashby's "homeostat," as much as through the computer and the robot), the animate quickly grew to be of equal significance. Application to social, anthropological, and psychological issues was pursued by Mead and Bateson (Bateson 1972a), especially in regard to mental health issues—a concern that Bateson shared with Ashby, also a psychologist. Management cybernetics was born of Stafford Beer (1926–2002) in the 1960s, and Gordon Pask (1928–1996) began cybernetic studies of teaching and learning in the 1950s.

There are many similarities between classical cybernetics and the slightly later mathematical theory of communication, or information theory, of Claude Shannon and Warren Weaver (1949); and general systems theory and its siblings, such as systems science, as developed by Ludwig von Bertalanffy (1950), making differentiation between these approaches difficult. Which term is used is frequently no more than a personal preference or historical accident. All of these approaches made notable contributions to such scientific and technological understandings and developments as the relationship between wholes and parts, automated control systems, approaches to complexity, developments in computing and communications hardware and software, and homeostasis in biological systems—to list but a few.

Early on, Wiener recognized ethical dangers in the cybernetic approach. The conjunction of animal and machine, even used metaphorically, has ethical implications—especially when the metaphor is predominantly of the animal as machine rather than the machine as animal. Another typical (and well-known) danger is that associated with the power of the machine, as exemplified, for example, in Isaac Asimov's "Three Laws of Robotics," from his science-fiction writings, which read:

First Law: A robot may not injure a human being, or, through inaction, allow a human being to come to harm.
Second Law: A robot must obey orders given it by human beings, except where such orders would conflict with the First Law.
Third Law: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. (Asimov 1942)

Wiener's Human Use of Human Beings (1950) is his attempt to come to terms with the most important of these dangers. He was not alone in this awareness. These ethical considerations, however, are not peculiar to cybernetics.

Second-Order Cybernetics

The initial promise of cybernetics was more than could be delivered, and the subject fell out of favor. By 1970 its funding base had eroded (with assistance from the Mansfield Agreement, a U.S. law introduced to prevent the military from funding any speculative research, or research that might not lead to an immediate military outcome). For some cyberneticians this indicated retrenchment, for others reconsideration leading to a new beginning: second-order cybernetics. The critical insight differentiating second-order cybernetics from classical (first-order) cybernetics is that second-order cybernetics takes cybernetic circularity more seriously.

Classical cybernetics exists within a worldview in which energy considerations reign paramount. The feedback loop is understood as requiring insignificant amounts of energy, thus creating a hierarchy. The controller, using relatively (and ignorably) little energy, controls the controlled, which is the big energy using part of the system. In second-order cybernetics, form and information are considered in preference to energy. In a second-order cybernetic control loop, the information passed between controller and controlled is understood to be of equal status. First-order hierarchy disappears. Each component in the loop contributes to the control of the whole. In effect, each component controls the other and the controller/controlled distinction is seen as a matter of role. The circular form of the cybernetic system is no longer disguised.

The difference was not initially presented this way. The originator of second-order cybernetics, von Foerster, made the following distinction on the frontispiece of his compilation "The Cybernetics of Cybernetics" (1975):

First order cybernetics—the cybernetics of observed systems / Second order cybernetics—the cybernetics of observing systems.

These two characterizations, however, appear similar if one treats observe and control as interchangeable verbs, and remembers that the observing/controlling system is observing/controlling the observed/controlled system in order to develop understanding, which requires feedback. Furthermore, these concerns are similar to those expressed in the involved observer of Ernst von Glasersfeld's Radical Constructivism (1987).

The circular systems of Second Order Cybernetics are essentially autonomous. Their stability derives from their (internal) maintenance of their circular processes. To an external observer they may appear to veer wildly. An example is the Autopoietic system of Humberto Maturana, Francisco Varela and Ricardo Uribe. This system constructs and then maintains itself, providing a model of "life"—or, rather, "living." Such systems are said to be organisationally closed but informationally open: the form of the system maintains (distinguishes) itself, is in this manner autonomous (Maturana and Varela 1992). Information enters, passes through (is processed by) and exits it. The system distinguishes itself as itself. Because these systems are autonomous, any meaning the information passing through them may have is unique, private to each system. Communication between these systems cannot be by transmission of meaning because each system builds its own meaning: Meanings are not communicated. Uncoded communication may, however, occur through conversation. Pask's conversation theory (a formalized version of everyday conversation developed, initially, to support communication in learning environments) provides a structure to sustain communication that is formally equivalent to the other circular systems of second-order cybernetics (Glanville 1996).

Admitting autonomy and conversation requires a system that accepts that, individually, one sees differently and understand uniquely, while acting as though one believes the objects one observes are the same. Otherwise, one's relativism would lead to isolation because one has nothing communicable and there is no one to communicate with. Ranulph Glanville's theory of "Objects" (1975) provides the framework that allows individuals to believe they each make different observations of the world, yet can act as if observing the same "Object"—the essential conceptual basis making second-order cybernetics and its ethical implications viable.

Second-order cybernetics has made notable contributions in such areas of human understanding as learning, conversational communication, and the emergence of the unanticipated (often through conversational processes). In particular, through the concepts and mechanisms of autopoiesis, it has aided in the understanding of how social systems acquire stability. Nevertheless, second-order cybernetics is probably better thought of more as a way of understanding than as a technology.

Ethics

There are those who would argue that, perhaps more than any other scientific or technological field, second-order cybernetics constitutes an effort to develop a scientific basis for ethics. As such it constitutes an important contribution to any discussion concerned with science, technology, and ethics. This section sketches the basis of this contribution.

Second-order cybernetics' circular systems are autonomous—the starting point for the ethical implications of second-order cybernetics. Von Foerster was among the first to register the ethical dimension in his essay, originally published in 1973, titled "On Constructing a Reality" (von Foerster 2003a); even more relevant was his 1992 essay, "Ethics and Second-Order Cybernetics." (Von Foerster's 1993 German book KybernEthik originated the term CybernEthics.)

Von Foerster proposed two imperatives:

Ethical imperative:/Act always so as to increase the number of choices. / Aesthetical imperative: / If you desire to see, learn how to act.

The ethical imperative insists that cybernetics has a dimension in ethics. Cybernetics implies generosity, increasing options. Von Foerster contrasted the essential meanness of morality (restrictions applied to others) to the generosity of ethics (which comes from within.)

The origin of this ethical concern can be seen to lie in the age-old question of what reality, if any, we can know independent of our knowing (i.e., is there a mind-independent reality [MIR]?). Although making a strong assumption of MIR is now commonplace, the question is in principle undecidable. Von Foerster remarked, "only we can decide the undecidable," leaving responsibility for answering this question (and, hence, for determining how we act) with each individual: one pursues whichever option one chooses. One's approach to one's world starts from this choice, which can be made once, or remade at will.

In second-order cybernetics, one's understanding of the world may be said to derive from a position of essential ignorance. The black box provides a mechanism for this. The understanding an observer builds through interacting with experience is (in the black box model) tentative: A reliable description of behavior emanating from the box may suggest it has been whitened, but nothing about the black box and our relationship to it has changed. It remains unopened (and unopenable)—provisional, as black as ever. Knowledge gained from using this model is based in profound ignorance. One cannot, therefore, insist on rightness and should tread warily, respecting the different views of others. The ethical implication of ignorance is respect for the views of others since one can never be certain, oneself. The views of others are considered as equal in stature to one's own—which does not mean theirs—or one's own—are either correct or viable.

Furthermore, the relationship between the behaviors (or signals), that is, the input and the output that black boxes are taken to act on—causing input to become output—results from interaction between observers and their own black boxes. Causality and its legal counterpart, blame, are seen to arise not from mechanism but from patterns observed by observers. The value of this understanding in how one acts cannot be over-emphasized, and is confirmed in many psychotherapies that depend for their effectiveness on persuading people that the blaming causality they see is their construction and responsibility. It is not what happens to one that matters, but how one responds to it.

The black box model requires that one distinguishes: If there is no distinction between behaviors there is nothing to experience. In essence, why distinguish myself if I am alone? Distinguishing myself, I distinguish myself also from another. This act of distinguishing brings into being and implies mutualism: whatever qualities may be attributed to one side of the distinction may (but need not) be attributed to the other. What I take for myself I may give you—this is von Foerster's ethical imperative again.

Distinctions, made in observing, can be considered a basis upon which observers construct experience, including experience of themselves. In order to assume experience is not solipsistic we assume that the other constructs (its experience of) itself (and us) in a reciprocal manner—another form of mutuality. Self-construction and maintenance indicate organizational closure: There is a boundary (it distinguishes its self) and the system is autonomous. An autonomous system is responsible. It has built itself, maintains itself (is organizationally closed), while it remains informationally open (communicates with its environment, thus substantiating the claim that, in distinguishing, one both distinguishes and distinguishes from). Bateson brings these ideas together when he uses the notion of difference (distinction) to define information: the difference that makes a difference (Bateson 1972b). The acceptance of responsibility grows out of autonomy (von Foerster 2003b): Autonomous systems are responsible for their actions. Here is the source of the aesthetical imperative.

There remains communication—that is, conversation. When communication is understood as individual construction of—and responsibility for—meaning and understanding by each participant (rather than the transmission of meanings and understandings), one can see that to understand the other one trusts the other's goodwill, acting with generosity, trust, honesty, and openness to build the understandings one will map onto each other's. This is an interaction. Teaching and learning (and much else beside) are interactive—the reason Pask developed conversation theory.

In turn, this understanding reveals that all one knows requires an observer's (knower's) presence, an understanding crucial in how one treats learning. Maturana said, "Everything said is said by an observer." Von Foerster retorted, "Everything said is said to an observer" (Von Foerster in Krippendorf 1979, p. 5). Respecting the observer is an ethical behavior.

Conclusion

Second-order cybernetics implies individuals are willing to treat each other, and (other, second-order) cybernetic systems, with a goodwill and generosity that can and should be understood as ethical implications. These go against some of the meaner understandings people currently and fashionably hold about their position in the world. Second-order cybernetics provides, in the ethical arena, hope and delight: those behaviors that are often considered higher, more civilized, and better are assumed and sustained in this way of understanding—a better-than-good reason for taking its lessons seriously.

RANULPH GLANVILLE

BIBLIOGRAPHY

Ashby, W. Ross. (1956). An Introduction to Cybernetics. London: Chapman and Hall. The most complete textbook introduction to cybernetics presented with simple profundity.

Asimov, Isaac. (1942). "Runaround." Reprinted in his Robot Dreams. London: Victor Gollancz, London, 1989.

Bateson, Gregory. (1972a). Steps to an Ecology of Mind: Collected Essays in Anthropology, Psychiatry, Evolution, and Epistemology. San Francisco: Chandler. Collected essays and commentaries from a multidisciplinary scholar who was one of the founding fathers of cybernetics, and whose writings can be seen to anticipate second-order cybernetics in several respects.

Bateson, Gregory. (1972b). "Pathologies of Epistemology." In Steps to an Ecology of Mind. An exploration of why we need the sort of shift that second order cybernetics embodies.

Glanville, Ranulph. (1975). A Cybernetic Development of Theories of Epistemology and Observation, with reference to Space and Time, as seen in Architecture. Ph.D. Thesis, Brunel University, 1975. Also known as The Object of Objects, the Point of Points,—or Something about Things.

Glanville, Ranulph. (1996). "Communication without Coding: Cybernetics, Meaning and Language (How Language, Becoming a System, Betrays Itself)." Invited paper in Modern Language Notes 111 (3): 441–462. Explores communication, language and meaning when the mechanism is conversation rather than code.

Glanville, Ranulph. (2002). "Second-Order Cybernetics." In Systems Science and Cybernetics: The Long Road to World Sociosystemicity, ed. F. Para-Luna. Oxford: Encyclopaedia of Life Support Systems (EOLSS) Publishers. One of the clearest explanations of a second-order cybernetic position and its epistemology.

Maturana, Humberto R., and Francisco J. Varela. (1992). The Tree of Knowledge: The Biological Roots of Human Understanding, rev. edition, trans. Robert Paolucci. Boston: Shambala.

Shannon, Claude, and Warren Weaver. (1998 [1949]). The Mathematical Theory of Communication. Urbana: University of Illinois Press. The classic text introducing what came to be called Information Theory.

Von Bertalanffy, Ludwig. (1950). "An Outline of General Systems Theory." British Journal for the Philosophy of Science 1: 134–165. One of von Bertalanffy's earlier papers discussing his proposals for a General Systems Theory.

Von Foerster, Heinz. (1979). "Cybernetics of Cybernetics." In Communication and Control in Society, ed. Klaus Krippendorf. New York: Gordon and Breach. Reprinted in Understanding Understanding: Essays on Cybernetics and Cognition, ed. Heinz Von Foerster. New York: Springer (2003).

Von Foerster, Heinz. (1991). "Ethics and Second-Order Cybernetics." In French in Systemes, Ethique, Perspectives en therapie familiale, ed. Yveline Rey and Bernard Prieur. Paris, ESF Editeur. Also in Understanding Understanding: Essays on Cybernetics and Cognition, ed. Heinz Von Foerster. New York: Springer (2003).

Von Foerster, Heinz. (1993). Kybernethik. Berlin: Merve Verlag.

Von Foerster, Heinz. (1995 [1975]). The Cybernetics of Cybernetics. Minneapolis: Future Systems Inc. A collection of cybernetic papers treated to cybernetic analysis providing a starting point for second order cybernetics. Von Foerster compiled and edited this collection, which was assembled by students at the University of Illinois and published by the publisher of the Whole Earth Catalogue, Stewart Brand.

Von Foerster, Heinz. (2003a). "On Constructing a Reality." In his Understanding Understanding: Essays on Cybernetics and Cognition. New York: Springer. Originally published in Environmental Design Research, Vol. 2, ed. Wolfgang F. E. Preiser (Stroudsburg, PA: Dowden, Hutchinson and Ross, 1973).

Von Foerster, Heinz. (2003b). "Responsibilities of Competence." In his Understanding Understanding: Essays on Cybernetics and Cognition. New York: Springer. Originally published in Journal of Cybernetics 2, no. 2 (1972): 1–6. The first venture into ethics by the father of second-order cybernetics.

Wiener, Norbert. (1948). Cybernetics; or, Communication and Control in the Animal and the Machine. New York: Wiley. 2nd edition, Cambridge, MA: MIT Press, 1961.

Wiener, Norbert. (1950). The Human Use of Human Beings: Cybernetics and Society. Boston: Houghton Mifflin.