Normal number

Normal number. In mathematics, a real number is called normal if, in its infinite decimal expansion, every finite string of digits occurs with the same limiting frequency as every other string of the same length. This means the digits are distributed uniformly, not just single digits but all possible blocks. The concept extends to expansions in any integer base greater than 1. While it is widely believed that fundamental constants like π and e are normal, this remains an unproven conjecture. The study of normal numbers sits at the intersection of number theory, probability theory, and ergodic theory.

Definition and properties

Formally, a real number *x* is normal in base *b* if, for every positive integer *k*, each possible string of *k* digits in the base-*b* alphabet appears in the base-*b* expansion of *x* with asymptotic frequency 1/*b*^k. This definition implies simple normality, where each single digit appears with frequency 1/*b*. A number that is normal to every integer base *b* ≥ 2 is called absolutely normal. A key property, established using tools from measure theory, is that almost all real numbers are absolutely normal, with the Lebesgue measure of the non-normal set being zero. However, constructing explicit examples and proving specific constants are normal is notoriously difficult. Normality is a property of the number's expansion, not its arithmetic value; a number normal in one base may not be normal in another, though absolutely normal numbers exist.

Examples and constructions

Explicitly constructed numbers provide the primary proven examples. The Champernowne constant, formed by concatenating all positive integers in order (0.123456789101112... in base 10), was shown by David Gawen Champernowne to be normal in base 10. The Copeland–Erdős constant, created by concatenating the prime numbers, is also normal in base 10, a result of Arthur Herbert Copeland and Paul Erdős. Gregory J. Chaitin discovered that the halting probability Chaitin's constant Ω is normal for all bases. Other constructions include those by Henri Lebesgue and Wacław Sierpiński, who gave early examples of absolutely normal numbers. Despite these artificial constructions, no fundamental constant from analysis or number theory, such as π, e, or the square root of 2, has been proven normal in any base.

Normality in different bases

A number can be normal in one base but not in another. For instance, a rational number has a terminating or eventually periodic expansion in every base and is thus not normal in any base. The concept of absolute normality requires normality across all bases simultaneously. A theorem by Émile Borel established that the set of numbers normal to a given base has full Lebesgue measure. The Stone–Čech compactification and notions from topological dynamics provide frameworks for studying base-dependent normality. A significant open problem is whether there exist numbers normal in one base but not normal in another; it is conjectured such numbers exist, but no explicit example is known. The Bailey–Borwein–Plouffe formula for π relates to digit extraction in specific bases but does not resolve normality questions.

History and development

The concept was introduced by the French mathematician Émile Borel in 1909. Borel used the nascent tools of measure theory to prove his normal number theorem, showing almost all numbers are normal. This connected number theory to probability theory. In 1933, David Gawen Champernowne constructed his famous constant and proved its normality in base 10, providing a concrete example. The mid-20th century saw further constructions by Arthur Herbert Copeland, Paul Erdős, and Wacław Sierpiński. The study was later enriched by perspectives from ergodic theory, particularly through the work of mathematicians like Andrey Kolmogorov and the application of Birkhoff's ergodic theorem. The question of the normality of fundamental constants like π was highlighted in the works of Ferdinand von Lindemann and remains a major unsolved problem, with computational checks by projects like the PiHex algorithm providing empirical but inconclusive evidence.

Related concepts

Several mathematical ideas are closely linked to normality. An absolutely normal number is normal in every base. A disjunctive sequence (or rich word) contains every finite string at least once, a weaker condition than normality. The Champernowne constant and Copeland–Erdős constant are canonical examples of normal numbers. In computer science, Chaitin's constant is a normal random number. The study of normal numbers intersects with Borel normality, uniform distribution modulo 1, and Benford's law. Concepts from dynamical systems, such as normal numbers for dynamical systems, generalize the idea. The unresolved conjectures about π and e connect to deep questions in analytic number theory and transcendental number theory, areas advanced by figures like Charles Hermite and Alexander Gelfond. Category:Number theory Category:Real numbers Category:Mathematical constants