Home > Positioning > Persons > Reichenbach

Hans Reichenbach (1891–1953)

Reichenbach was a philosopher of science whose posthumous work The Direction of Time (1956) developed an account of the mechanism by which traces form — the concept of branch systems. A branch system is a subsystem that interacts with a larger system, acquires a low-entropy imprint through that interaction, and then evolves in quasi-isolation toward equilibrium. A footprint in sand, a fossil in rock, a crater on the moon — each is a branch system whose current ordered state is only explicable by reference to a past interaction. The concept connects Boltzmann’s statistical mechanics to the everyday experience of records and traces: we find records of the past, not of the future, because branch systems inherit their low-entropy initial states from interactions that occurred when the larger system was at lower entropy. Beyond the direction of time, Reichenbach was a leading logical empiricist whose work on probability, induction, the philosophy of space and time, and the foundations of relativity was influential across the philosophy of science.

Life

Born 26 September 1891 in Hamburg, Germany. Studied civil engineering, then physics and mathematics at the universities of Berlin, Munich, and Göttingen. PhD in philosophy from the University of Erlangen (1915), with a dissertation on probability. Served in the German army during the First World War. Attended Einstein’s first lectures on general relativity in Berlin in 1917 — an encounter that shaped the direction of his philosophical work.

Privatdozent in physics at the Technische Hochschule in Stuttgart (1920–26). Professor of philosophy at the University of Berlin (1926–33), where he joined the Berlin Circle — the logical empiricist group allied with the Vienna Circle. Founded and edited the journal Erkenntnis with Rudolf Carnap.

Dismissed from Berlin in 1933 under the Nazi racial laws (Reichenbach was of Jewish descent). Emigrated to Turkey, where he held a professorship in philosophy at the University of Istanbul (1933–38) — part of Atatürk’s programme of importing European academics to modernise Turkish universities. Moved to the University of California, Los Angeles (1938), where he spent the remainder of his career, building a significant philosophy department and training a generation of American philosophers of science.

Died 9 April 1953 in Los Angeles, aged sixty-one. The Direction of Time was published posthumously in 1956, edited by Maria Reichenbach.

The direction of time

The Direction of Time (University of California Press, 1956). The book asks: what gives time its direction? The laws of physics are time-symmetric; the thermodynamic arrow is statistical. Reichenbach’s contribution is to connect the statistical arrow to the formation of records and traces — the asymmetry that makes the past knowable and the future not.

Branch systems. The central concept. Consider a closed, isolated system at or near equilibrium. Now a subsystem separates from the larger system through an interaction — a footprint is pressed into sand, a signal is recorded, a fossil is deposited. This subsystem — the branch — inherits a low-entropy initial state from the interaction. Once separated, it evolves toward its own equilibrium, its ordered state degrading over time. The footprint erodes; the recording degrades; the fossil weathers.

The asymmetry: branch systems acquire their low-entropy states from interactions in the past, because the larger system was at lower entropy in the past. We find records of past events and not future events because record-formation requires the entropy gradient that the cosmological boundary condition provides. The direction of records is the direction of the thermodynamic arrow.

The common cause principle. Reichenbach formalised a related idea: if two events are correlated but neither causes the other, there exists a common cause that explains the correlation. This “screening off” principle connects probability theory to causal structure and has been influential in the philosophy of causation, in Bayesian network theory, and in the foundations of causal inference.

Probability and induction

Reichenbach’s earlier work addressed the problem of induction — how empirical generalisation can be justified — through a frequentist theory of probability. The Theory of Probability (1935; English translation 1949) developed probability as the limit of relative frequency in a sequence of trials, and argued that the “pragmatic justification” of induction is the best available: if regularities exist in nature, the inductive method will find them; if they do not, no method will succeed. The argument does not prove that induction works; it proves that induction is the best available bet.

Philosophy of space and time

The Philosophy of Space and Time (1928; English translation 1958). Reichenbach’s analysis of the conceptual foundations of relativity theory. The book argues that the geometry of physical space is not known a priori (contra Kant) but is an empirical question, to be settled by measurement. The conventionalist element — that certain aspects of spacetime description (the simultaneity convention in special relativity, the choice of geometry in general relativity) involve conventional choices rather than empirical discoveries — became a central topic in the philosophy of physics.

Where Reichenbach stops

Reichenbach’s branch-system concept explains how records form and why they point toward the past. It does not address what happens when a system actively maintains its records — when the trace is not passively degrading toward equilibrium but is sustained at thermodynamic cost. A fossil is a branch system; an organism carrying a memory is not — it is actively importing free energy to keep its records intact. The step from passive trace to active retention marks a threshold that Reichenbach’s framework identifies by implication but does not theorise. Schrödinger’s What is Life? and Prigogine’s dissipative structures later addressed what happens on the other side of that threshold.

Key works

See also: Boltzmann · Prigogine · Rovelli · Kant