Alphabeta Math
TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck 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.

For α>0 every ordinal β is α⋅ξ+ρ with ρ<α, in exactly one way

Statement

Let α and β be ordinals (Ordinal (von Neumann)) with α>0. Then there are unique ordinals ξ and ρ with

β=α⋅ξ+ρandρ<α.

ξ is the quotient and ρ the remainder of β on division by α; concretely, ξ is the largest ordinal with α⋅ξ≤β, and ρ is what For α≤β there is exactly one ordinal γ with α+γ=β returns from α⋅ξ≤β.

No choice principle is used.

Facts & Assumptions

Given: Ordinals α>0 and β, with + and ⋅ as in Ordinal addition α+β and Ordinal multiplication α⋅β. For a set A of ordinals, sup⁡A=⋃A is its least upper bound.

[L1]

α⋅0=0, α⋅δ+=α⋅δ+α, and α⋅λ=sup⁡{α⋅ζ:ζ∈λ} for limit λ (Ordinal multiplication α⋅β).

[L2]

From Monotonicity of ordinal + and ⋅: strictly increasing and continuous in the right argument, weakly increasing in the left, with left cancellation, and the identities 0+β=β and 1⋅β=β: 1⋅μ=μ and μ+0=μ (claim (a)); μ<ν implies α+μ<α+ν, left cancellation for +, and α≤α+μ (claim (b)); for α>0, μ<ν implies αμ<αν, hence μ≤ν implies αμ≤αν (claim (d)); and μ≤ν implies μγ≤νγ (claim (e)).

[L3]

If μ≤ν there is exactly one γ with μ+γ=ν (For α≤β there is exactly one ordinal γ with α+γ=β).

[L4]

Every nonempty set of ordinals has an ∈-least element, and exactly one of μ∈ν, μ=ν, ν∈μ holds (Trichotomy and well-ordering of the ordinals).

[L5]

μ+ is an ordinal, μ⊆ν if and only if μ∈ν or μ=ν, and μ∉μ (claims (b), (c), (f) of Basic closure properties of ordinals); consequently μ<ν if and only if μ+≤ν.

[L6]

Every ordinal is exactly one of 0, a successor, or a limit (Successor and limit ordinals).

Proof

technique · direct
1.1L1L2L5

β<α⋅β+: since α>0 gives 1≤α, claim (e) of [L2] gives β+=1⋅β+≤α⋅β+, and β∈β+.

1.2L1L2L4L5

Uniqueness: suppose αξ1+ρ1=αξ2+ρ2=β with ρ1,ρ2<α; if ξ1<ξ2 then ξ1+≤ξ2 by [L5], so αξ1+α=αξ1+≤αξ2≤αξ2+ρ2=β=αξ1+ρ1<αξ1+α by [L1] and [L2], which [L5] forbids; by symmetry ξ2<ξ1 is impossible too, so ξ1=ξ2 by [L4] and then ρ1=ρ2 by left cancellation.

2.1step 1.1L5

The collection C={η∈(β+)+:β∈α⋅η} is a set of ordinals by Separation, and it is nonempty, because β+∈(β+)+ and β∈α⋅β+ by step 1.1.

3.1step 2.1L4

Let η0 be the ∈-least element of C, which exists by [L4].

4.1step 3.1L1L4L5L6

η0 is a successor: it is not 0, since α⋅0=0 and β∉0; and it is not a limit λ, for then every ζ∈λ would lie in (β+)+ by transitivity and outside C by minimality, so αζ≤β by [L4], making β an upper bound of {αζ:ζ∈λ} and hence α⋅λ≤β by [L1], contradicting β∈α⋅λ; so η0=ξ+ for a unique ordinal ξ by [L6].

5.1step 4.1step 3.1L1L4L5

With that ξ: ξ∈η0⊆(β+)+ and ξ∉C by minimality of η0, so αξ≤β by [L4]; and β∈α⋅ξ+=αξ+α because η0=ξ+∈C.

6.1step 5.1L2L3L4

By [L3] applied to αξ≤β there is exactly one ρ with αξ+ρ=β, and αξ+ρ=β<αξ+α forces ρ<α, since α≤ρ would give αξ+α≤αξ+ρ by [L2].

7.1step 6.1step 1.2∎

Existence is step 6.1 and uniqueness is step 1.2, so β=α⋅ξ+ρ with ρ<α in exactly one way.

Remarks

Why the least η with β<α⋅η has to be a successor. Because η↦α⋅η is continuous at limits: at a limit stage its value is the supremum of the earlier values, so it cannot overtake β for the first time there. That is claim (f) of Monotonicity of ordinal + and ⋅: strictly increasing and continuous in the right argument, weakly increasing in the left, with left cancellation, and the identities 0+β=β and 1⋅β=β in the form used at step 4.1, and it is the only place the limit clause of Ordinal multiplication α⋅β is used.

The bound (β+)+ is a Separation device. "The least η with β<αη" quantifies over all ordinals, which is not a set; step 1.1 supplies a specific witness inside (β+)+, so the collection can be cut out of a set. Nothing depends on the particular bound.

Uniqueness is proved before existence, and independently of it. Step 1.2 uses only the monotonicity laws, so it applies to any two representations whatever their origin. This is the order used again in Cantor normal form: every nonzero ordinal is ωβ0⋅c0+⋯+ωβk−1⋅ck−1 with β0>⋯>βk−1 and each ci a nonzero natural number, in exactly one way, where uniqueness is what licenses the definite article in "the Cantor normal form".

The remainder can be 0 and the quotient can be 0. If β<α then ξ=0 and ρ=β; if α divides β exactly then ρ=0. Neither case is excluded, and neither needs separate treatment.

Depends on

Dependency tree · 2 levels

Sources