Normal Extension

lean4-prooffield-theoryvisualization
K/F normal    fF[x],  f irred., has a root in K    f splits in KK/F \text{ normal} \iff \forall f \in F[x],\; f \text{ irred., has a root in } K \implies f \text{ splits in } K

Statement

An algebraic extension K/FK/F is normal if every irreducible polynomial fF[x]f \in F[x] that has at least one root in KK splits completely over KK:

K/F normal    fF[x] irred.,  f has a root in K    f splits in K.K/F \text{ normal} \iff \forall f \in F[x] \text{ irred.},\; f \text{ has a root in } K \implies f \text{ splits in } K.

Equivalently, K/FK/F is normal iff KK is the splitting field of some (possibly infinite) family of polynomials over FF.

In Mathlib: Normal F K is the predicate. Splitting fields are always normal: Normal.of_isSplittingField.

A Galois extension is one that is both normal and separable.

Visualization

Not normalQ(23)/Q\mathbb{Q}(\sqrt[3]{2})/\mathbb{Q}:

The minimal polynomial of 23\sqrt[3]{2} over Q\mathbb{Q} is x32x^3 - 2, which has three roots:

x32=(x23)(xω23)(xω223),ω=e2πi/3x^3 - 2 = (x - \sqrt[3]{2})(x - \omega\sqrt[3]{2})(x - \omega^2\sqrt[3]{2}), \quad \omega = e^{2\pi i/3}

Only 23\sqrt[3]{2} lies in Q(23)R\mathbb{Q}(\sqrt[3]{2}) \subset \mathbb{R}. The complex roots ω23\omega\sqrt[3]{2} and ω223\omega^2\sqrt[3]{2} are not real, so x32x^3 - 2 does not split in Q(23)\mathbb{Q}(\sqrt[3]{2}).

NormalQ(2)/Q\mathbb{Q}(\sqrt{2})/\mathbb{Q}:

The minimal polynomial x22x^2 - 2 has roots ±2\pm\sqrt{2}, both in Q(2)\mathbb{Q}(\sqrt{2}). Extension is the splitting field of x22x^2 - 2.

ExtensionPolynomialRoots in extensionNormal?
Q(2)/Q\mathbb{Q}(\sqrt{2})/\mathbb{Q}x22x^2 - 22,2\sqrt{2}, -\sqrt{2}Yes
Q(23)/Q\mathbb{Q}(\sqrt[3]{2})/\mathbb{Q}x32x^3 - 2only 23\sqrt[3]{2}No
Q(24,i)/Q\mathbb{Q}(\sqrt[4]{2}, i)/\mathbb{Q}x42x^4 - 2all four rootsYes
Q(i)/Q\mathbb{Q}(i)/\mathbb{Q}x2+1x^2 + 1i,ii, -iYes

Proof Sketch

  1. Splitting fields are normal. Let KK be the splitting field of pp over FF. Take any irreducible fF[x]f \in F[x] with a root αK\alpha \in K. There is an FF-isomorphism F(α)F[x]/(f)F(\alpha) \cong F[x]/(f). All roots of ff are conjugates of α\alpha; each lies in KK because KK is generated over FF by all roots of pp, and adjoining any root of ff stays within the algebraic closure available. A careful argument using the universal property shows ff splits in KK.
  2. Converse. If K/FK/F is normal and finite, then KK is the splitting field of the product of the minimal polynomials of a generating set.

Connections

Every normal separable extension is Galois, linking normality directly to the Fundamental Theorem of Galois TheoryFundamental Theorem of Galois Theory{intermediate fields}{subgroups of Gal(K/F)}\{\text{intermediate fields}\} \longleftrightarrow \{\text{subgroups of } \text{Gal}(K/F)\}Bijection between intermediate fields and subgroups of the Galois groupRead more →. The classification of which extensions are normal also underlies the Impossibility of the Quintic FormulaImpossibility of the Quintic FormulaS5 is not solvable    no radical formula for degree 5S_5 \text{ is not solvable} \implies \text{no radical formula for degree } \geq 5No general algebraic formula exists for polynomials of degree 5 or higherRead more → — the key obstruction is that certain degree-5 extensions are not normal.

Lean4 Proof

/-- The splitting field of any polynomial is normal over the base field.
    Mathlib: `Normal.of_isSplittingField` with the `IsSplittingField` instance
    provided by `SplittingField`. -/
theorem splittingField_normal (F : Type*) [Field F]
    (p : Polynomial F) : Normal F (p.SplittingField) :=
  @Normal.of_isSplittingField F _ p.SplittingField _ _ p inferInstance