7 Circular Causality

Circular causality is the causal structure in which the effects of a variable feed back to influence the causes that produced it, creating a closed loop of mutual determination in which no single element can be identified as an unqualified first cause or final effect. In linear causality, A causes B, which causes C, and the chain of influence is directional and non-recursive. In circular causality, A influences B, which in turn influences A, so that A and B are simultaneously cause and effect of each other, and their states can only be understood as the joint solution to a system of coupled equations rather than as independent cause-and-effect pairs. Circular causality is the foundational structure of all feedback systems and the central conceptual innovation of cybernetics.

The philosophical significance of circular causality was recognized by Norbert Wiener, Arturo Rosenblueth, and their collaborators in the early 1940s as the principle that unified the behavior of purposive machines and living organisms. Prior to this recognition, the prevailing mechanistic view of causation was linear: the organism or machine receives inputs and produces outputs, and the task of analysis was to trace the unidirectional chain from causes to effects. Circular causality revealed that goal-directed behavior requires a circular causal structure in which the system's outputs are compared to goals, the comparison generates error information, and the error information flows back to modify the inputs—closing the causal chain into a loop. This loop structure makes the system sensitive to the discrepancy between its current state and its goal, rather than merely responding passively to its inputs.

A simple representation of circular causality in a feedback control system shows the mutual dependence: the error e is caused by the reference r and the output y; the control action u is caused by e through the controller; the output y is caused by u through the plant; and the output y feeds back to influence e, completing the circle:

Substituting these equations into each other shows that y depends on e, which depends on y: the system is defined by a set of simultaneous equations rather than a sequence of independent ones. The closed-loop transfer function that results from solving this circular system—Y(s)/R(s) = C(s)P(s)/(1 + C(s)P(s))—represents the emergent behavior of the circular causal system, a behavior that neither the plant nor the controller exhibits alone.

The consequences of circular causality for understanding systems are profound. In a linearly causal chain, removing a cause eliminates its downstream effects, and effects cannot influence their causes. In a circularly causal system, removing any element in the loop changes the behavior of all others; interventions at any point propagate both forward and backward around the loop; and the steady-state behavior of the system cannot be attributed to any single cause but emerges from the circular interaction of all elements. This makes analysis more complex—requiring simultaneous solution of coupled equations rather than sequential evaluation of a causal chain—but it also makes systems more adaptive: because each element responds to the state of the others, the system can achieve coordinated behavior that compensates for perturbations affecting any component.

Ecological systems exhibit circular causality in their predator-prey dynamics. Predator populations are causally influenced by prey availability—more prey supports larger predator populations. But predator populations are also a cause of prey availability—more predators reduce prey numbers. The resulting circular causal structure produces the oscillatory dynamics characteristic of predator-prey systems: as prey increase, predators increase, which reduces prey, which reduces predators, which allows prey to recover, completing the cycle. The Lotka-Volterra equations model this circular causality:

where x is prey population and y is predator population. Each equation contains the product xy, representing the circular coupling: x affects dy/dt (predators grow when prey are available) and y affects dx/dt (prey decline when predators are present). Neither population can be analyzed independently; they are jointly determined by the circular causal system.

In social and economic systems, circular causality generates the complex dynamics of interconnected human behaviors. Confidence in a bank is maintained as long as deposits remain stable; stable deposits sustain confidence; but if confidence falters, withdrawals begin, which reduce deposits, further undermining confidence—a positive circular causality that amplifies small perturbations into bank runs. Inflation expectations are themselves a cause of inflation: workers who expect higher prices demand higher wages, which raise production costs, which raise prices, fulfilling the expectation. Central bank credibility in maintaining price stability interrupts this circular causality by anchoring inflation expectations at the target rate, so that the self-fulfilling inflationary circle cannot be initiated.

In interpersonal communication, circular causality means that what one person communicates shapes what the other person communicates in return, which shapes the first person's subsequent communication, and so on through the interaction. Communication is not a series of independent messages but a circularly causal process in which each message is simultaneously a response to what came before and a cause of what comes next. This circular structure means that the meaning and effect of any communicative act cannot be fully understood from the act alone but only in the context of the circular causal loop within which it is embedded—a principle that systems therapists, negotiation analysts, and organizational communication theorists have applied extensively in understanding how communication patterns emerge from and reinforce relational structures.