Cybernetics is an interdisciplinary field concerned with how systems communicate, regulate their behavior, and respond to changing conditions. It examines relationships among actions, information, and their consequences, particularly when those consequences influence subsequent actions through feedback. Its subjects include machines, organisms, and social organizations. Rather than defining systems primarily by their material composition, cybernetics investigates patterns of behavior that different systems may share. Norbert Wiener established its influential modern formulation in his 1948 book Cybernetics: Or Control and Communication in the Animal and the Machine. (asc-cybernetics.org)
Origins and historical development
The word derives from the Greek kybernētēs, meaning a steersman or helmsman. Steering illustrates the central idea: a person observes a vessel’s course, adjusts the rudder, and uses the resulting change to guide further corrections. Wiener adopted this terminology for a science connecting communication with control in both biological and mechanical systems. (asc-cybernetics.org)
Cybernetics emerged from several previously developing research traditions, including automatic regulation, mathematical descriptions of nervous activity, and the study of homeostasis. During World War II, problems such as predicting aircraft movement and controlling antiaircraft weapons encouraged research into feedback and prediction. These engineering problems intersected with questions about purposeful behavior in organisms. The field’s formation therefore involved a convergence of ideas rather than a single invention. (asc-cybernetics.org)
The Macy conferences, held between 1946 and 1953, provided an important setting for this convergence. Ten meetings brought together researchers from mathematics, engineering, physiology, and the social sciences. Participants included Warren McCulloch, Walter Pitts, Wiener, John von Neumann, Margaret Mead, and Gregory Bateson. Their discussions explored feedback, circular causal relationships, and parallels between biological and artificial systems. The conferences helped establish a shared vocabulary, although participants did not develop one uniformly accepted theory. (asc-cybernetics.org)
Feedback, regulation, and stability
A feedback loop exists when a system’s output affects its subsequent operation. In a simple temperature-control system, a sensor measures temperature, a controller compares that measurement with a desired value, and a heater changes the temperature. The resulting measurement returns to the controller, completing the loop. This differs from an open-loop arrangement, which acts without using the resulting output to correct its behavior. (cds.caltech.edu)
Negative feedback counteracts deviations: corrective action tends to reduce the difference between an observed condition and a reference condition. Positive feedback reinforces changes and can amplify deviations. These terms describe the direction of interaction, not whether its consequences are desirable. Feedback can improve regulation, but its effectiveness depends on the dynamics of the entire interconnected system. Delays or excessive corrective action can produce oscillation or instability rather than reliable control. (cds.caltech.edu)
Cybernetic explanation often replaces a simple linear chain of causation with a circular relationship. An action changes the environment, and that changed environment affects the system’s next action. Goal-directed behavior can consequently be analyzed through observable mechanisms without assuming that a machine possesses intentions or consciousness. This approach made feedback a bridge between engineering and the investigation of behavior. (asc-cybernetics.org)
Information and limits of regulation
Cybernetics is closely related to information theory, but their emphases differ. Information theory supplies mathematical tools for analyzing uncertainty and communication; cybernetics asks how communicated information participates in regulation and behavior. Measurements must distinguish relevant conditions, and communication channels must preserve enough information for appropriate responses. Noise can undermine this process by obscuring the differences on which regulation depends. (asc-cybernetics.org)
W. Ross Ashby developed a particularly systematic account in An Introduction to Cybernetics (1956). His law of requisite variety relates the range of disturbances a regulator faces to the range of responses available to it. Effective regulation requires enough distinguishable responses to counter disturbances that would otherwise drive important variables outside acceptable limits. The principle is not simply that a controller must be as complicated as its environment: the relevant distinctions depend on which outcomes must be prevented or maintained. (ashby.info)
Ashby also analyzed systems through states, transformations, and black-box descriptions. A black-box investigation studies relationships between inputs and outputs without requiring complete knowledge of internal construction. Such methods allow comparisons among physically different systems, while making the choice of variables and observational distinctions explicit. (ashby.info)
First-order and second-order cybernetics
The distinction between first-order and second-order cybernetics concerns the observer’s place in an investigation. First-order descriptions typically study a system as an object whose behavior can be observed and regulated. Second-order approaches include the observer’s activities within the account, asking how observation, description, and intervention participate in the relationships under study. Heinz von Foerster characterized the distinction as one between observed systems and observing systems. (asc-cybernetics.org)
An important episode was Mead’s 1967 presentation “Cybernetics of Cybernetics,” published in 1968. She proposed that the American Society for Cybernetics apply cybernetic ideas to its own organization and operation. The broader second-order tradition extended such reflexive questions into cognition and epistemology: how observers construct descriptions and how their participation affects what they describe. (asc-cybernetics.org)
Applications and disciplinary relationships
Cybernetics overlaps with control theory, which develops mathematical methods for analyzing and designing regulated systems. Control theory commonly addresses stability, performance, and robustness using models such as differential equations. Cybernetics has a wider interdisciplinary scope, including conceptual questions about communication, organization, and observation. (cds.caltech.edu)
Its history also intersects with artificial intelligence and robotics, although neither is synonymous with cybernetics. Cybernetic research emphasizes ongoing interaction between a system and its surroundings. In another branch, Gordon Pask’s conversation theory investigated learning and understanding through interactions among people and machines, contributing a framework for studying dialogue and collaborative design. Cybernetic approaches thus encompass both automatic regulation and processes in which participants negotiate meanings or goals. (pangaro.com)