Alphabeta Math
LemmaStatement: AI-adaptedProof: AI-adaptedSession-authored (Fable 5 assisted)precheck passjudge pass (z-ai/glm-5.2)audited 2026-07-29
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.

Transfinite recursion along the ordinals: a class rule determines exactly one operation defined at every ordinal

Statement

Let GG be a class function: a rule, given by a formula in the language of set theory, that assigns a set G(h)G(h) to every function hh whose domain is an ordinal (Ordinal (von Neumann)). Then there is a class function FF, given by a formula and defined at every ordinal, such that

F(β)=G(Fβ)for every ordinal β,F(\beta) = G(F \restriction \beta) \qquad \text{for every ordinal } \beta,

and FF is the only one: any class function FF' defined at every ordinal and satisfying F(β)=G(Fβ)F'(\beta) = G(F' \restriction \beta) for every ordinal β\beta agrees with FF at every ordinal. Here FβF \restriction \beta is the restriction of FF to the set β\beta of ordinals below β\beta, which is a set even though FF is not.

Like Transfinite recursion, this is a theorem schema of ZF: one theorem for each formula defining GG. It uses Replacement, inherited from that theorem, and it uses no form of the Axiom of Choice.

Why the published theorem does not already say this. Transfinite recursion is stated for a well-order (W,<)(W, <), that is for a set, and it delivers one function whose domain is that set. An operation such as α+β\alpha + \beta has to be defined at every ordinal β\beta, and the ordinals are not a set (Burali-Forti: there is no set of all ordinals), so no single instance of the published theorem defines it. What is proved below is exactly the bridge: the instances at the individual ordinals cohere, and the coherence is supplied by the published theorem's own uniqueness clause. No new recursion principle is introduced.

Facts & Assumptions

Given: A class function GG as in the statement, and the axioms of ZF. No choice principle is assumed. For an ordinal γ\gamma we write (γ,)(\gamma, \in) for γ\gamma carrying the membership relation.

[A1]

G(h)G(h) is a well-determined set for every function hh whose domain is an ordinal, and the rule is given by a formula.

[L1]

Transfinite recursion on a set: for a well-order (W,<)(W, <) (Well-order and well-ordered set) and a class function GG defined on functions whose domains are proper initial segments of WW, there is exactly one function FF with domain WW such that F(a)=G(FW<a)F(a) = G(F \restriction W_{<a}) for every aWa \in W (Transfinite recursion).

[L2]

An ordinal is a transitive set on which \in is a strict well-order (Ordinal (von Neumann)).

[L3]

W<a={xW:x<a}W_{<a} = \{x \in W : x < a\}, and every proper initial segment of a well-order is W<aW_{<a} for exactly one aa (Initial segment of a well-order).

[L4]

Every element of an ordinal is an ordinal, α+=α{α}\alpha^{+} = \alpha \cup \{\alpha\} is an ordinal, and αβ\alpha \subseteq \beta if and only if αβ\alpha \in \beta or α=β\alpha = \beta (claims (a), (c), (f) of Basic closure properties of ordinals).

[L5]

Every nonempty set of ordinals has an \in-least element (Trichotomy and well-ordering of the ordinals).

Proof

technique · direct
1.1

For every ordinal γ\gamma the pair (γ,)(\gamma, \in) is a well-order, by clause 2 of the definition of an ordinal.

L2
1.2

For βγ\beta \in \gamma the initial segment of (γ,)(\gamma, \in) determined by β\beta is γ<β={xγ:xβ}=β\gamma_{<\beta} = \{x \in \gamma : x \in \beta\} = \beta, because βγ\beta \subseteq \gamma by transitivity of γ\gamma; so by [L3] the proper initial segments of (γ,)(\gamma, \in) are exactly the ordinals βγ\beta \in \gamma, and a function whose domain is one of them is a function whose domain is an ordinal, to which GG applies.

L2L3A1
1.3

If ξβ\xi \in \beta then ξ+β+\xi^{+} \in \beta^{+}: transitivity of β\beta gives ξβ\xi \subseteq \beta, and ξβ\xi \in \beta gives {ξ}β\{\xi\} \subseteq \beta, so ξ+β\xi^{+} \subseteq \beta, whence ξ+β\xi^{+} \in \beta or ξ+=β\xi^{+} = \beta by [L4], and either way ξ+β{β}=β+\xi^{+} \in \beta \cup \{\beta\} = \beta^{+}.

L2L4
2.1

Applying [L1] to the well-order (γ,)(\gamma, \in) and to GG yields, for each ordinal γ\gamma, exactly one function FγF_\gamma with domain γ\gamma satisfying Fγ(β)=G(Fγβ)F_\gamma(\beta) = G(F_\gamma \restriction \beta) for every βγ\beta \in \gamma.

step 1.1step 1.2L1A1
3.1

Coherence: for δγ\delta \in \gamma put u=Fγδu = F_\gamma \restriction \delta, a function with domain δ\delta; for βδ\beta \in \delta transitivity of δ\delta gives βδ\beta \subseteq \delta, so uβ=Fγβu \restriction \beta = F_\gamma \restriction \beta and u(β)=Fγ(β)=G(Fγβ)=G(uβ)u(\beta) = F_\gamma(\beta) = G(F_\gamma \restriction \beta) = G(u \restriction \beta), so uu satisfies the recursion on δ\delta and the uniqueness half of [L1] applied to (δ,)(\delta, \in) gives Fγδ=FδF_\gamma \restriction \delta = F_\delta.

step 2.1L1L2
3.2

Define F(β):=Fβ+(β)F(\beta) := F_{\beta^{+}}(\beta), which makes sense because β+\beta^{+} is an ordinal by [L4] and ββ+\beta \in \beta^{+}; the defining condition is a formula in β\beta, so FF is a class function defined at every ordinal.

step 2.1L4construct
4.1

For every ordinal β\beta and every ξβ\xi \in \beta we have ξ+β+\xi^{+} \in \beta^{+}, hence Fβ+ξ+=Fξ+F_{\beta^{+}} \restriction \xi^{+} = F_{\xi^{+}} and therefore Fβ+(ξ)=Fξ+(ξ)=F(ξ)F_{\beta^{+}}(\xi) = F_{\xi^{+}}(\xi) = F(\xi); so Fβ=Fβ+βF \restriction \beta = F_{\beta^{+}} \restriction \beta, which is a set because Fβ+F_{\beta^{+}} is.

step 1.3step 3.1step 3.2
5.1

Consequently F(β)=Fβ+(β)=G(Fβ+β)=G(Fβ)F(\beta) = F_{\beta^{+}}(\beta) = G(F_{\beta^{+}} \restriction \beta) = G(F \restriction \beta) for every ordinal β\beta, which is the required recursion equation.

step 4.1step 3.2
6.1

For uniqueness, let FF' be a class function defined at every ordinal with F(β)=G(Fβ)F'(\beta) = G(F' \restriction \beta) for every β\beta, and suppose F(β0)F(β0)F(\beta_0) \ne F'(\beta_0) for some ordinal β0\beta_0; then D={ξβ0+:F(ξ)F(ξ)}D = \{\xi \in \beta_0^{+} : F(\xi) \ne F'(\xi)\} is a set by Separation, it is a set of ordinals by [L4], and it is nonempty because β0D\beta_0 \in D, so it has an \in-least element μ\mu by [L5].

step 5.1L4L5
7.1

Every ξμ\xi \in \mu lies in β0+\beta_0^{+}, because μβ0+\mu \in \beta_0^{+} gives μβ0+\mu \subseteq \beta_0^{+}, and ξD\xi \notin D by minimality of μ\mu, so F(ξ)=F(ξ)F(\xi) = F'(\xi); hence Fμ=FμF \restriction \mu = F' \restriction \mu and F(μ)=G(Fμ)=G(Fμ)=F(μ)F(\mu) = G(F \restriction \mu) = G(F' \restriction \mu) = F'(\mu), contradicting μD\mu \in D.

step 6.1step 5.1L4
8.1

No such β0\beta_0 exists, so FF and FF' agree at every ordinal, and FF is the unique class function on the ordinals satisfying F(β)=G(Fβ)F(\beta) = G(F \restriction \beta).

step 7.1step 6.1step 5.1

Remarks

What is spent. Replacement, through Transfinite recursion, and Separation, at step 6.1. No choice principle appears anywhere, for the same reason as in the published theorem: at every stage the object used is the unique function with a given domain, never one selected from many.

Three cases, not two. The lemma says nothing about how GG is given. In practice GG is defined by the three-way split of Successor and limit ordinals — a value at 00, a rule at a successor, a rule at a limit — and that split is exhaustive and exclusive for every ordinal. Writing "successor or limit" and forgetting 00 is the standard way to define an operation that is undefined at 00.

Why the restriction is a set. Step 4.1 is not bookkeeping. FF is a proper class, so "FβF \restriction \beta" needs an argument, and the argument is that it coincides with the restriction of the set function Fβ+F_{\beta^{+}}. Without it, G(Fβ)G(F \restriction \beta) would not even be an application of GG to a set.

The naming. Some texts state this as "transfinite recursion on the class of ordinals" and prove it directly by a least-counterexample argument. The route taken here spends nothing new: it reuses the published theorem at each ordinal and glues, and the glue is that theorem's uniqueness clause.

Depends on

Used by

Dependency tree · next 3 levels

Direct dependencies and their dependencies through the next three levels: 33 results over 15 levels. An arrow runs from a result to what uses it, and this result sits at the bottom with a heavier outline. Click the chart to enlarge it.

Sources