Mathematics

Andrey Kolmogorov: The Father of Modern Probability

The extraordinary life and vast mathematical legacy of Andrey Kolmogorov, who laid the axiomatic foundations of probability theory and made transformative contributions to topology, turbulence, algorithmic complexity, and a dozen other fields.

Profiles Probability

Early Life

Andrey Nikolaevich Kolmogorov was born on 25 April 1903 in Tambov, Russia. His mother, Mariya Yakovlevna Kolmogorova, died in childbirth, and his father, Nikolai Matveyevich Kataev, was largely absent. Andrey was raised by his maternal aunt, Vera Yakovlevna, in the village of Tunoshna near Yaroslavl, and later in Moscow.

As a child, Kolmogorov was precocious and curious about everything. At the age of five, he noticed patterns in arithmetic — for example, that $1 = 1^2$ , $1 + 3 = 2^2$ , $1 + 3 + 5 = 3^2$ , and so on — and contributed this observation to the school magazine. This foreshadowed a lifetime of finding deep structure in patterns.

Kolmogorov entered Moscow State University in 1920, at the age of seventeen, during the turmoil following the Russian Revolution. He initially studied history but soon turned to mathematics, where his talent was recognized by Nikolai Luzin, the leader of the famous "Luzitania" school of mathematics in Moscow.

Early Mathematical Work

Kolmogorov's first major result came in 1922, at age nineteen, when he constructed a Fourier series that diverges almost everywhere — a stunning counterexample to the widely held belief that Fourier series of integrable functions should converge:

Kolmogorov's Counterexample (1922)

There exists a function $f \in L^1[0, 2\pi]$ whose Fourier series diverges at every point:

$\sum_{n=-\infty}^{\infty} \hat{f}(n) e^{inx} \quad \text{diverges for all } x$

This result astonished the mathematical community and established Kolmogorov's reputation at the age of nineteen.

In the early 1920s, he also made fundamental contributions to descriptive set theory and intuitionist logic, proving (independently of Brouwer) that classical propositional logic can be embedded into intuitionistic logic.

The Axiomatization of Probability

Kolmogorov's most influential work is his 1933 monograph Grundbegriffe der Wahrscheinlichkeitsrechnung ("Foundations of the Theory of Probability"), in which he established probability theory on a rigorous axiomatic basis using measure theory.

Before Kolmogorov, probability was a collection of techniques without a unified foundation. Different authors used different definitions, and there was no consensus on what "probability" meant mathematically. Kolmogorov resolved this by defining probability in terms of measure theory:

Kolmogorov's Axioms of Probability

A probability space is a triple $(\Omega, \mathcal{F}, P)$ where:

$\Omega$ is a sample space (the set of all outcomes)
$\mathcal{F}$ is a $\sigma$ -algebra of subsets of $\Omega$ (the events)
$P: \mathcal{F} \to [0,1]$ $P : F \to [0, 1]$ is a probability measure satisfying:
- $P(\Omega) = 1$
- For any countable collection of pairwise disjoint events $A_1, A_2, \ldots$ :

$P\left(\bigcup_{i=1}^{\infty} A_i\right) = \sum_{i=1}^{\infty} P(A_i)$

This framework unified all of probability theory within measure theory, allowed the rigorous treatment of continuous distributions, infinite-dimensional processes, and conditional expectations, and opened the door to the modern theory of stochastic processes.

Conditional Expectation

Kolmogorov rigorously defined conditional expectation $\mathbb{E}[X \mid \mathcal{G}]$ as the unique (a.s.) $\mathcal{G}$ -measurable function satisfying:

$\int_A \mathbb{E}[X \mid \mathcal{G}] \, dP = \int_A X \, dP \quad \text{for all } A \in \mathcal{G}$

This definition is the foundation for martingale theory, filtering, and modern mathematical finance.

Kolmogorov's Contributions Across Mathematics

Kolmogorov's work spans an astonishing range of fields. Few mathematicians in history have made first-rate contributions to so many different areas.

Turbulence Theory

In 1941, Kolmogorov published his famous theory of turbulence (known as K41), one of the most cited results in all of physics. For fully developed turbulent flow, he predicted that the energy spectrum follows a power law:

$E(k) \sim C \varepsilon^{2/3} k^{-5/3}$

where $k$ is the wavenumber, $\varepsilon$ is the energy dissipation rate, and $C$ is a universal constant. This $-5/3$ law has been confirmed experimentally and remains the fundamental result in turbulence theory.

The Kolmogorov–Arnold–Moser (KAM) Theorem

In classical mechanics, the KAM theorem (1954–1963) addresses the stability of nearly integrable Hamiltonian systems. Kolmogorov initiated the theorem, which was later completed by Vladimir Arnold and Jürgen Moser:

KAM Theorem (Kolmogorov 1954, Arnold 1963, Moser 1962)

If an integrable Hamiltonian system is subjected to a sufficiently small perturbation, then most of the invariant tori (on which quasi-periodic motion occurs) are preserved, provided they satisfy a Diophantine condition — their frequencies $\omega$ satisfy:

$|\langle k, \omega \rangle| \geq \frac{\gamma}{|k|^\tau} \quad \text{for all } k \in \mathbb{Z}^n \setminus \{0\}$

for some constants $\gamma > 0$ and $\tau > n - 1$ .

This result resolved fundamental questions about the long-term stability of the solar system and has deep connections to number theory through the Diophantine conditions.

Algorithmic Complexity (Kolmogorov Complexity)

In the 1960s, Kolmogorov (independently of Gregory Chaitin and Ray Solomonoff) defined the Kolmogorov complexity of a string $x$ as the length of the shortest program that produces $x$ on a universal Turing machine:

$K(x) = \min\{|p| : U(p) = x\}$

A string is random if $K(x) \approx |x|$ — that is, it cannot be compressed. This definition made precise the intuitive notion of randomness and became foundational for information theory, data compression, and the philosophy of induction.

Topology

Kolmogorov made early contributions to algebraic topology, including the definition of cohomology groups (independently of and simultaneously with J.W. Alexander). The Kolmogorov product (cup product) in cohomology:

$\smile: H^p(X) \times H^q(X) \to H^{p+q}(X)$

makes the cohomology ring a graded-commutative algebra, a structure of fundamental importance in topology.

Superposition of Functions (Hilbert's 13th Problem)

In 1957, Kolmogorov proved a remarkable result on the representation of continuous functions of several variables:

Kolmogorov Superposition Theorem. Every continuous function $f: [0,1]^n \to \mathbb{R}$ can be written as:

$f(x_1, \ldots, x_n) = \sum_{q=0}^{2n} \Phi_q\left(\sum_{p=1}^{n} \phi_{q,p}(x_p)\right)$

where $\Phi_q$ and $\phi_{q,p}$ are continuous functions of a single variable. This resolved a version of Hilbert's 13th problem.

The Teacher and the School

Kolmogorov was one of the great mathematical educators. He supervised over 80 doctoral students, many of whom became leading mathematicians, including Vladimir Arnold, Yakov Sinai, and many others. He was deeply involved in mathematical education at all levels, helping to reform the Soviet mathematics curriculum and founding specialized mathematics schools for talented youth.

"He strode through mathematics like a giant, and left footprints everywhere." — Yakov Sinai, on Kolmogorov

Personal Life

Kolmogorov lived with his lifelong companion, the topologist Pavel Alexandrov, in a dacha (country house) outside Moscow from the 1930s until Alexandrov's death in 1982. They were inseparable companions for nearly five decades, sharing an intellectual and personal bond that was central to both their lives.

Kolmogorov was an avid outdoorsman — he loved hiking, swimming, and skiing. He believed in the unity of physical and intellectual vigor.

Awards and Recognition

Kolmogorov received virtually every major honor available to a mathematician:

Stalin Prize (1941)
Lenin Prize (1965)
Wolf Prize in Mathematics (1980)
Hero of Socialist Labor (twice)
Foreign member of numerous academies, including the Royal Society and the National Academy of Sciences

Death and Legacy

Andrey Kolmogorov died on 20 October 1987 in Moscow, at the age of 84. His influence on mathematics is pervasive: the Kolmogorov axioms, Kolmogorov complexity, the KAM theorem, K41 turbulence theory, the Kolmogorov extension theorem, Kolmogorov's zero–one law, and many more results bear his name.

"Kolmogorov was, in the breadth of his interests and the depth of his contributions, the closest twentieth-century equivalent to Euler."

— Multiple attributions in the mathematical community

References

Kolmogorov, A.N., Foundations of the Theory of Probability, Chelsea, 1950 (English translation of the 1933 German original).
Shiryaev, A.N. (ed.), Kolmogorov in Perspective, AMS/LMS History of Mathematics, 2000.
Kolmogorov, A.N., "The Local Structure of Turbulence in Incompressible Viscous Fluid for Very Large Reynolds Numbers," Doklady Akademii Nauk SSSR, 30(4), 1941.
Wikipedia — Andrey Kolmogorov
MacTutor — Andrey Kolmogorov

Back to all posts