Transcendental Number Theory

Transcendental number theory asks a deceptively simple question: which real or complex numbers can never be the root of a polynomial equation with whole-number coefficients?

9 frameworks, from 1844 to today · Number Theory · Mathematics

still a working view ended · each leads to its entry below

Overview

Transcendental number theory asks a deceptively simple question: which real or complex numbers can never be the root of a polynomial equation with whole-number coefficients1? Such numbers are called transcendental; the rest are algebraic12. The distinction matters because algebraic numbers are, in a precise sense, the numbers that arithmetic can reach—they are built from the integers by the operations of algebra. A transcendental number escapes that world entirely. Proving that any particular number is transcendental is hard, because it requires ruling out every polynomial at once, and the field's history is largely a history of inventing indirect methods that do exactly that.

The stakes reach beyond curiosity. Transcendental results settle ancient geometric questions, constrain what equations can be solved in closed form, and connect the arithmetic of numbers to the theory of approximation. The field's central open problems—above all the conjecture of Schanuel—describe a landscape that is still mostly unexplored.

Liouville and the First Transcendental Numbers

Before anyone could exhibit a transcendental number, the existence question had to be framed. Joseph Liouville supplied the key idea in the 1840s1. An algebraic number of degree \(d\) (the degree of its lowest-degree polynomial) cannot be approximated by rationals too well: there is a constant \(c\) such that \(|x - p/q| > c/q^d\) for every rational \(p/q\). This is the first instance of what became the field's signature tool, Diophantine approximation—the study of how closely irrational numbers can be approached by fractions.

Liouville turned the inequality around. If a number can be approximated by rationals faster than any power of the denominator, it cannot be algebraic1. He then constructed such numbers explicitly, for instance the sum of \(10^{-n!}\) over \(n \ge 1\), whose partial sums are rationals converging absurdly fast2. This gave the first proven transcendental numbers, in 184432. The method was powerful but narrow: it certified only numbers built to be badly approximable, and said nothing about constants like \(e\) or \(\pi\).

Hermite, Lindemann, and the Classical Constants

The next breakthrough came from a different direction: instead of approximating a number, one studies the polynomials and exponential relations it satisfies. Charles Hermite proved in 1873 that \(e\) is transcendental12. His argument introduced an auxiliary function—a function constructed for the proof, here built from exponentials and a polynomial—whose integral is estimated in two contradictory ways: it must be a nonzero integer, yet it is too small to be one. This "integer that is too small" contradiction became the field's standard engine.

Ferdinand von Lindemann extended the method in 1882 to prove that \(\pi\) is transcendental, and more generally that if \(\alpha\) is a nonzero algebraic number then \(e^\alpha\) is transcendental14. The consequence settled a question open since antiquity: squaring the circle with straightedge and compass is impossible, because a constructible length must be algebraic, and \(\pi\) is not. The Hermite–Lindemann theorem also shows that any nonzero logarithm of an algebraic number is either transcendental or, in special cases, algebraic—a template for later results.

Hilbert's Problem and the Gelfond–Schneider Theorem

David Hilbert, in his 1900 address, listed among his problems the transcendence of numbers such as \(2^{\sqrt{2}}\) and \(e^\pi\)15. These are values of the exponential function at algebraic points, but the exponent is irrational algebraic, so Hermite–Lindemann does not apply. The problem stood for three decades.

Aleksandr Gelfond (1934) and Theodor Schneider (1935), independently, proved the general theorem: if \(\alpha\) is algebraic, not 0 or 1, and \(\beta\) is irrational algebraic, then \(\alpha^\beta\) is transcendental. The proof again uses an auxiliary function, but now a polynomial in \(z\) and \(e^{\beta z}\), and a clever counting argument: one shows that sufficiently many values of the function vanish, forcing a contradiction with a linear-algebra (determinant) estimate. The theorem resolved Hilbert's problem and gave transcendence of \(2^{\sqrt{2}}\), \(e^\pi\) (since \(e^\pi = (-1)^{-i}\)), and the Gelfond constant \(e^\pi\)167.

Gelfond–Schneider and the Logarithmic Form

The Gelfond–Schneider theorem has a clean reformulation in terms of logarithms: if \(\alpha\) and \(\beta\) are nonzero algebraic numbers and \(\log \alpha / \log \beta\) is irrational, then that ratio is transcendental8. This "linear form in logarithms" viewpoint became the organizing principle of the field. It also exposed the limits of the method: it handles two logarithms but not three. The transcendence of sums like \(\log 2 \cdot \log 3\) or of \(e^{e}\) remained out of reach, and the general problem of linear independence of logarithms of algebraic numbers became the field's central target.

Baker's Theory of Linear Forms in Logarithms

Alan Baker's work in the mid-1960s broke the two-logarithm barrier1. Baker proved that any nonzero linear combination of logarithms of algebraic numbers with algebraic coefficients is transcendental—equivalently, that such logarithms are linearly independent over the algebraic numbers unless an obvious multiplicative relation forces dependence. The proof combined the auxiliary-function method with several complex variables and a sophisticated induction on the number of logarithms.

The theorem's importance was not only qualitative. Baker's method yielded effective lower bounds: explicit, computable estimates for how small a nonzero linear form in logarithms can be8. This made it a tool in Diophantine equations—for instance, in bounding integer solutions to Thue equations and in the resolution of Catalan-type problems—and it earned Baker the Fields Medal in 1970. The theory of linear forms in logarithms remains one of the most-used effective tools in number theory.

The Auxiliary-Function Method as a General Technique

Across these results, a single strategy recurs, and it deserves to be named plainly. One constructs an auxiliary function—typically a polynomial times an exponential, or a sum of such terms—that vanishes at many prescribed points. Then one estimates the same quantity in two ways: an arithmetic argument shows it must be a nonzero integer (or a nonzero algebraic integer with bounded size), while an analytic argument shows it is smaller than any nonzero integer could be. The contradiction proves the theorem. The art lies in choosing the function and the points so that both estimates hold. This method, refined by Gelfond, Schneider, Baker, and later by many others, is the field's workhorse, and its descendants—including the "zero estimate" and multiplicity arguments—drive modern results.

Diophantine Approximation and the Geometry of Numbers

Parallel to the transcendence proofs, a quantitative theory of approximation developed. The geometry of numbers, founded by Hermann Minkowski around 1890, studies lattices and convex bodies and gives sharp bounds on how well numbers can be approximated by rationals with bounded denominator. It supplied the structural language—successive minima, lattice determinants—that later approximation arguments use. The Thue–Siegel–Roth theorem (Roth, 1955) showed that algebraic irrationals cannot be approximated by rationals better than \(1/q^{2+\varepsilon}\) for any \(\varepsilon > 0\), a sharpening of Liouville's exponent that is itself a transcendence-adjacent result19. These approximation results and the transcendence theorems feed each other: better approximation bounds often translate into stronger transcendence statements, and vice versa.

Schanuel's Conjecture and the Open Frontier

The field's grand unifying conjecture was stated by Stephen Schanuel in the early 1960s. It predicts that if \(z1, \dots, zn\) are complex numbers linearly independent over the rationals, then among the \(2n\) numbers \(z1, \dots, zn, e^{z1}, \dots, e^{zn}\) at least \(n\) are algebraically independent. The conjecture would imply both Hermite–Lindemann and Gelfond–Schneider, and would settle a host of open questions—the transcendence of \(e^{e}\), of \(\pi^e\), and the algebraic independence of \(e\) and \(\pi\). It remains unproved even for \(n = 2\) in full generality. The Schanuel conjecture programme is the attempt to prove special cases and to derive consequences; progress has been incremental, and the conjecture functions as the field's guiding star rather than a settled theorem.

The Present

Today transcendental number theory proceeds along several fronts. Baker's linear-forms theory is applied constantly in computational and effective Diophantine problems. The auxiliary-function method continues to be refined, with multiplicity and several-variable versions pushing the boundaries of what can be proved. The Schanuel conjecture remains the central open problem, and the algebraic independence of \(e\) and \(\pi\) is still unknown. The field's methods have also migrated: they inform the theory of \(p\)-adic transcendence, the study of values of modular and automorphic functions, and questions about periods and motives. What began as a hunt for individual exotic numbers has become a general theory of how algebraic and transcendental quantities constrain one another—and the deepest constraints are still conjectural.

References

  1. Transcendental number theory (introduction, Transcendence, Approximation by rational numbers: Liouville to Roth, Auxiliary functions: Hermite to Baker, Major results), Wikipedia
  2. Transcendental number (introduction, History), Wikipedia
  3. Liouville number, Wikipedia
  4. Lindemann–Weierstrass theorem, Wikipedia
  5. Hilbert's seventh problem (introduction, Background and the statement), Wikipedia
  6. Gelfond's constant, Wikipedia
  7. Gelfond–Schneider constant, Wikipedia
  8. Baker's theorem (History, introduction), Wikipedia
  9. Roth's theorem, Wikipedia

The frameworks of Transcendental Number Theory

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1844-Present

    Diophantine approximation

    How closely can irrationals be approximated by fractions?

    Fly to its tower in Nousfarer

  2. 1844-1900

    Liouville's Approximation Method

    Which numbers resist rational approximation enough to be transcendental?

    Fly to its tower in Nousfarer

  3. 1873-Present

    Auxiliary function

    How can a constructed function force a contradiction?

    Fly to its tower in Nousfarer

  4. 1873-1935

    Hermite–Lindemann Exponential Method

    Which exponential values of algebraic numbers are transcendental?

    Fly to its tower in Nousfarer

  5. 1890-Present

    Geometry of Numbers

    How do lattices bound rational approximation?

    Fly to its tower in Nousfarer

  6. 1934-1970

    Gelfond–Schneider theorem

    When is an algebraic number raised to an algebraic power transcendental?

    Fly to its tower in Nousfarer

  7. 1960-Present

    Schanuel's conjecture

    Which exponential and logarithmic values are algebraically independent?

    Fly to its tower in Nousfarer

  8. 1966-Present

    Baker's theorem

    When are logarithms of algebraic numbers linearly independent?

    Fly to its tower in Nousfarer

  9. 1970-Present

    P-adic and Modern Transcendence

    How do p-adic methods extend transcendence results?

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Diophantine approximationLiouville's Approximation MethodAuxiliary functionHermite–Lindemann Exponential MethodGeometry of NumbersGelfond–Schneider theoremSchanuel's conjectureBaker's theoremP-adic and Modern Transcendence

Beyond the field

Where Transcendental Number Theory's ideas came from in other fields, and what they led to there.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Transcendental Number Theory is a planet in Nousfarer

Its city has a tower for each of these 9 frameworks. Land, walk in, and climb one concept at a time.

Land on Transcendental Number Theory

Related fields

Fields its ideas are linked to

More in Number Theory

See where Transcendental Number Theory sits on the Map of Human Knowledge