Alphabeta Math
TheoremStatement: AI-adaptedProof: AI-adaptedPipeline-generatedjudge pass (gpt-5.6-terra)audited 2026-09-22
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Banach-Stone

Statement

Assume the Axiom of Choice (The Axiom of Choice). Let K be R or C, let K and L be nonempty compact Hausdorff spaces, and let T:C(K,K)→C(L,K) be a surjective linear isometry, where both spaces carry the supremum norm. Then there are a homeomorphism h:L→K and a continuous function u:L→K with ∣u(y)∣=1 for all y, such that

Tf(y)  =  u(y) f(h(y))(f∈C(K,K), y∈L).

Conversely, for every homeomorphism h:L→K and every continuous u:L→K with ∣u∣=1, the formula Tf:=u⋅(f∘h) defines a surjective linear isometry C(K)→C(L). The representation is by a pair (h,u) that is unique: h is determined by T and u=T(1). The conclusion does not say that a general linear isometry is multiplicative or unital.

Facts & Assumptions

Given: An assumed Axiom of Choice, nonempty compact Hausdorff spaces K,L, and a surjective linear isometry T:C(K)→C(L) over K∈{R,C}.

[L1]

The extreme points of the dual unit ball of C(K) are exactly the normalized evaluations: ext⁡BC(K)∗={c δx:x∈K, ∣c∣=1} (Extreme points of the dual ball of C(K), The Axiom of Choice).

[L2]

The transpose T∗:C(L)∗→C(K)∗ is bounded linear with (T∗g)(f)=g(Tf), ∥T∗∥=∥T∥, and (ST)∗=T∗S∗; consequently for a bijective isometry T one has (T∗)−1=(T−1)∗ (The transpose of a bounded operator, The transpose is bounded with the same norm, Transposition reverses composition).

[L3]

A surjective linear isometry maps the unit ball onto the unit ball and preserves extreme points: if e is extreme and Ve=(1−t)y+tz with y,z in the target unit ball, applying V−1 writes e as the corresponding convex combination of V−1y and V−1z.

[L4]

For a nonempty compact Hausdorff space M, the evaluation map M→Δ(C(M)) is a homeomorphism, and the family C(M) separates points from closed sets, so the evaluation map into the product over C(M,[0,1]) is an embedding (Characters of continuous functions are evaluations, The evaluation map of a point–closed-set separating family is a topological embedding).

[L5]

Under Dependent Choice — which follows from the Axiom of Choice — the Urysohn lemma holds in normal spaces, so in a compact Hausdorff space two distinct points are separated by a continuous function into [0,1] (Urysohn's lemma, under the axiom of dependent choice: in a normal space two disjoint closed sets are separated by a continuous function into [0,1], and conversely such a space is normal, The Axiom of Choice).

Proof

technique · direct
1.1L2algebra

T−1 is linear and isometric, because ∥T−1g∥=∥T(T−1g)∥=∥g∥ for all g; hence by [L2] the transpose T∗ is a bounded linear bijection with (T∗)−1=(T−1)∗, and ∥T∗∥=∥(T∗)−1∥=1.

1.21.1L1L3

For every y∈L the point evaluation δy is an extreme point of BC(L)∗ of the form 1⋅δy, so by [L1] and [L3] its image T∗δy is an extreme point of BC(K)∗; by [L1] there are unique h(y)∈K and u(y)∈K with ∣u(y)∣=1 and T∗δy=u(y)δh(y). This defines functions h:L→K and u:L→K.

1.3algebra

Conversely, let h:L→K be a homeomorphism and u:L→K continuous with ∣u∣=1, and set Tf:=u⋅(f∘h). Then T is K-linear, ∥Tf∥∞=sup⁡y∣u(y)f(h(y))∣=sup⁡x∣f(x)∣=∥f∥∞ because h is surjective, and T is surjective with inverse Sg:=(g∘h−1)/u(h−1).

2.1step 1.2L2algebra

Evaluating T∗δy=u(y)δh(y) at the constant function 1 gives u(y)=(T∗δy)(1)=δy(T1)=(T1)(y), so u=T1 is continuous, and ∣u(y)∣=1 for all y by [step 1.2].

2.2step 1.2L2algebra

For f∈C(K) and y∈L: Tf(y)=δy(Tf)=(T∗δy)(f)=u(y)f(h(y)), using the definition of the transpose in [L2] and [step 1.2].

3.1step 2.1step 2.2L4

The map h is continuous: for every f∈C(K) the function y↦f(h(y))=Tf(y)/u(y) is continuous, since Tf is continuous and u is continuous with ∣u∣=1 so 1/u=u‾ is continuous; the family {f∘h:f∈C(K)} therefore consists of continuous functions and the evaluation embedding eK of K into the product over C(K,[0,1]) has continuous composition eK∘h, whence h is continuous because eK is an embedding by [L4].

3.2step 1.1step 2.2

Applying [step 1.2] and [step 2.2] to the surjective isometry T−1:C(L)→C(K) (which is a surjective linear isometry by [step 1.1]) produces h′:K→L continuous and v:K→K with ∣v∣=1 such that T−1g(x)=v(x)g(h′(x)) for all g,x.

4.1step 2.2step 3.2L5algebra

From TT−1=idC(L): for g∈C(L) and y∈L one computes g(y)=T(T−1g)(y)=u(y)(T−1g)(h(y))=u(y)v(h(y)) g(h′(h(y))) by [step 2.2] and [step 3.2], with ∣u(y)v(h(y))∣=1; if h′(h(y))≠y then [L5] gives g with g(y)≠g(h′(h(y))), contradicting the displayed identity; so h′∘h=idL, and the same argument with the roles reversed gives h∘h′=idK. Hence h is a bijection with continuous inverse h′=h−1, that is, a homeomorphism.

5.1step 1.3step 2.1step 2.2step 4.1∎

By [step 2.2] and [step 4.1] every surjective linear isometry has the asserted form with h a homeomorphism and ∣u∣=1; by [step 1.3] every pair (h,u) of that form defines a surjective linear isometry; and the pair is unique since u=T1 by [step 2.1] and then h is recovered from T by the formula.

Remarks

  • Nonemptiness is a hypothesis. For K=∅ or L=∅ the space C(K) is the zero algebra and the conclusion is vacuous; the argument above uses nonemptiness to have a point evaluation to transpose.
  • The weight u is forced. Step 2.1 identifies u with T1, so the isometry is unital precisely when u≡1; nothing in the theorem requires this.

Depends on

Used by

Dependency tree · two levels

41 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources