Alphabeta Math
RemarkRemark: AI-adaptedProof: Not applicableverified 2026-08-06 (claude-opus-5)
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.

The Axiom of Choice is stated on this page and assumed by no proof on it; the two statements that would need it are identified and left unsettled

Remark

The Axiom of Choice is stated on this page, at The Axiom of Choice, together with the notion of a choice function it is stated in terms of, Choice function. Nothing on the page assumes it: every proof here is carried out without any choice principle, and two natural-looking statements are missing from the page for exactly that reason. This is the account of where the line falls.

What is proved without choice, and why. A construction is choice-free when the object it produces is determined by the data rather than selected from several candidates. The two-sided inverse of f:A→B is a bijection if and only if there is a function g:B→A with g∘f=ΔA and f∘g=ΔB; such a g is unique, equals the inverse relation f−1, and is itself a bijection is determined: for a bijection f and a point b of the codomain there is exactly one a with f(a)=b, so no selection is made. The left inverse of For f:A→B with A≠∅: f is injective if and only if there is g:B→A with g∘f=ΔA; for A=∅ the empty function is injective and has a left inverse if and only if B=∅ needs one arbitrary point a0 of a nonempty set, which is a single existential instantiation and not a choice principle: one point is chosen once, not one point for each index. The function produced by Let ∼ be an equivalence relation on A with quotient map π:A→A/∼, and let f:A→B. There is a function g:A/∼→B with g∘π=f if and only if a∼a′ implies f(a)=f(a′); and such a g is then unique is likewise determined, because its value on a class is forced to be the common value of f on that class, and The equivalence classes of an equivalence relation are nonempty, cover A, and are pairwise equal or disjoint; conversely every such cover arises from exactly one equivalence relation is proved from the three defining properties alone.

A choice function is an element of a product, and the two formulations of the axiom are one statement. The Axiom of Choice gives the axiom twice, once as "every family of nonempty sets has a choice function" and once as "a product of nonempty sets is nonempty", and both objects are now defined on this page, so the passage between the two readings can be written out. Let F be a set all of whose members are nonempty, and index it by itself: the identity relation ΔF of The identity relation ΔA={ (a,b)∈A×A:a=b } and the membership relation ∈A ={ (a,b)∈A×A:a∈b } is a function with domain F sending each S to itself, by clause (ii) of If f and g are functions then g∘f is a function with domain f−1[dom⁡g] and (g∘f)(x)=g(f(x)) there; ΔA is a function with ΔA(a)=a; and f∘ΔA=f=ΔB∘f for f:A→B, hence an indexed family (An indexed family (Ai)i∈I is a function with domain I; {Ai:i∈I} is its range) whose range is F, so its indexed union is ⋃F (⋃i∈IAi:=⋃{Ai:i∈I}, and ⋂i∈IAi:=⋂{Ai:i∈I} for I≠∅). Unfolding The product ∏i∈IAi:={ f:I→⋃i∈IAi ∣ f(i)∈Ai for every i∈I } for that family gives the set of functions g:F→⋃F with g(S)∈S for every S∈F, which is word for word the set of choice functions for F in Choice function. So a choice function for F is exactly an element of the product of the members of F, and the two formulations assert the same thing about the same object.

The first statement that would need choice: a right inverse for a surjection. For a surjection f:A→B (Injection, surjection, bijection) each b∈B has at least one preimage, and a right inverse is a rule picking one preimage for every b at once. That is a simultaneous selection over the whole of B, and no proof on this page makes one: the assertion that every surjection has a right inverse is equivalent to the Axiom of Choice.

The second: a nonempty product. ∏i∈∅Ai={∅}; if Aj=∅ for some j∈I then ∏i∈IAi=∅; and for I={j} the evaluation f↦f(j) is a bijection ∏i∈IAi→Aj settles ∏i∈IAi when I has no element, when some member is empty, and when I has exactly one element, and For I={∅,{∅}} with A∅=A and A{∅}=B, the map f↦(f(∅),f({∅})) is a bijection ∏i∈IAi→A×B settles the two-index case, in each case by writing an element down. For an arbitrary index set the assertion that ∏i∈IAi is nonempty whenever every Ai is nonempty is the product formulation of the axiom, which by the identification above is the choice-function formulation read at the family indexed by itself; The product ∏i∈IAi:={ f:I→⋃i∈IAi ∣ f(i)∈Ai for every i∈I } says so where the product is introduced.

Where the rest of the library's account of choice lives. The strength of the weaker choice principles relative to one another, and what survives without any of them, is recorded at The proved choice ledger: hypotheses, equivalences, and upper bounds ↗.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · two levels

37 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