Home > Positioning > Persons > Zermelo

Ernst Zermelo (1871–1953)

Zermelo is known to physicists for the recurrence objection to Boltzmann’s statistical mechanics and to mathematicians for the axiom of choice and the first axiomatisation of set theory. The two contributions are unrelated in subject but share a common intellectual character: both involve pushing a formal argument to its logical limit and insisting on the consequences. The recurrence objection (1896) pointed out that Poincaré’s recurrence theorem — which guarantees that a mechanical system in a finite phase space will eventually return arbitrarily close to any initial state — is incompatible with a permanent increase in entropy. The axiom of choice (1904) asserted the existence of a mathematical object whose construction no one could exhibit, provoking a foundational crisis in mathematics that shaped the discipline for decades.

Life

Born 27 July 1871 in Berlin, Prussia. Educated at the universities of Berlin, Halle, and Freiburg. PhD in mathematics at the University of Berlin (1894), under Hermann Schwarz. Early work was in the calculus of variations and mathematical physics, including hydrodynamics.

Privatdozent at the University of Göttingen (1897–1910), where he came under the influence of David Hilbert. It was at Göttingen that Zermelo produced his most important mathematical work: the well-ordering theorem (1904), the axiom of choice, and the first axiomatic set theory (1908). Professor of mathematics at the University of Zurich (1910–16); resigned due to health problems (tuberculosis). Honorary professor at the University of Freiburg (1926). Resigned in 1935 after refusing to give the Nazi salute at the opening of lectures — an act of defiance that cost him his position at a time when most of the German academic establishment complied. He lived in relative isolation thereafter, financially strained and mathematically unproductive; his late writings on set theory and the foundations of mathematics circulated little. Reinstated at Freiburg after the war (1946). Died 21 May 1953 in Freiburg im Breisgau.

The recurrence objection

Zermelo’s objection (1896) to Boltzmann’s H-theorem:

Poincaré’s recurrence theorem (1890) proves that a mechanical system confined to a finite region of phase space — the space of all possible positions and momenta of its particles — will, given enough time, return arbitrarily close to any previous state. This includes its initial state. If a gas starts in a low-entropy configuration, evolves to equilibrium (high entropy), and then — by the recurrence theorem — eventually returns to a state arbitrarily close to its initial low-entropy configuration, then entropy cannot increase monotonically. The H-theorem, which appears to prove that entropy increases and stays high, is contradicted by the recurrence theorem, which guarantees that entropy will eventually decrease.

Boltzmann’s response. The recurrence time for a macroscopic system is staggeringly long — for a cubic centimetre of gas, it exceeds any physically meaningful timescale by many orders of magnitude (Boltzmann estimated timescales on the order of 10^(10^19) years). The recurrence is mathematically real but physically irrelevant: no observation, however patient, would ever witness a spontaneous return to a low-entropy macrostate. Boltzmann also pointed out that the H-theorem describes the overwhelmingly probable behaviour of a system, not its guaranteed behaviour — it is a statistical result, not a deterministic one.

The Zermelo-Boltzmann exchange, coming twenty years after the Loschmidt-Boltzmann exchange on reversibility, completed the clarification of the statistical nature of the second law. Between them, the two objections established that irreversibility is neither a logical consequence of the laws of motion (Loschmidt) nor compatible with the long-term behaviour of a finite mechanical system (Zermelo) — it is a consequence of the statistical properties of the initial conditions and the overwhelming probability of entropy increase on any humanly relevant timescale.

Set theory and the axiom of choice

Zermelo’s mathematical work is independent of his physics but equally consequential.

The well-ordering theorem (1904). Zermelo proved that every set can be well-ordered — arranged so that every non-empty subset has a least element. The proof required a new axiom: the axiom of choice, which asserts that for any collection of non-empty sets, there exists a function that selects exactly one element from each set. The axiom seems intuitive for finite collections but is non-constructive for infinite ones — it asserts the existence of a selection without providing any procedure for making it.

The 1904 paper provoked immediate and vigorous controversy. Émile Borel, Arthur Schoenflies, Felix Bernstein, Julius König, and G. H. Hardy all published criticisms — some questioning the proof’s validity, others accepting the proof but rejecting the axiom of choice on which it rested. Zermelo responded with a second, more detailed proof in 1908, accompanied by a systematic reply to his critics that clarified the axiom’s status and the distinction between accepting a proof and accepting its premises. The controversy was the first sustained foundational debate in modern mathematics and established the axiom of choice as a permanent point of contention.

The axiomatic set theory (1908). In response to the paradoxes that had shaken the foundations of set theory — Russell’s paradox (1901), the Burali-Forti paradox — Zermelo proposed the first axiomatic system for set theory: a list of axioms specifying which sets exist and which operations on sets are permitted. The system was later refined by Abraham Fraenkel and Thoralf Skolem into Zermelo-Fraenkel set theory (ZF), which, with the axiom of choice added (ZFC), became the standard foundation of mathematics — the axiomatic framework within which virtually all of modern mathematics can be formalised.

Where Zermelo stops

The recurrence objection is a logical point, not a physical theory. Zermelo showed that Boltzmann’s H-theorem is in tension with the Poincaré recurrence theorem; he did not offer an alternative account of irreversibility. His position — that a mechanical account of the second law is impossible — was too strong: Boltzmann’s statistical interpretation survives, qualified by the acknowledgment that the second law is probabilistic and depends on initial conditions, which is exactly the clarification Zermelo’s objection forced.

The axiom of choice and its consequences remain contested in the foundations of mathematics. The axiom is independent of the other axioms of set theory (proved by Gödel in 1938 and Paul Cohen in 1963): it can be neither proved nor disproved from the other ZF axioms. The most vivid demonstration of its unsettling consequences is the Banach-Tarski paradox (1924): using the axiom of choice, a solid ball in three-dimensional space can be decomposed into finitely many pieces and reassembled — using only rigid rotations and translations — into two solid balls, each identical in size to the original. The result is a theorem, not a conjecture; the “pieces” involved are non-measurable sets that cannot be constructed or visualised, but their existence follows from the axiom. Banach-Tarski is the standard example invoked to question whether the axiom of choice should be accepted — it gives the contestation its concrete force.

Mathematicians who reject non-constructive existence proofs — constructivists and intuitionists — reject the axiom of choice or restrict its use. The mainstream accepts it (ZFC is the default foundation), but the acceptance is pragmatic rather than principled: the axiom is useful, and no contradiction has been found. Whether the axiom of choice describes a feature of mathematical reality or is merely a convenient postulate is a live question in the philosophy of mathematics.

Key works

See also: Boltzmann · Loschmidt · Russell