Alphabeta Math
Session-authored (Fable 5 assisted)
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.

33 results · all verified · 17 also independently AI-judged
Every result on this page is machine-checked by a proof checker and read in full and owner-audited; the judge is an additional, independent cross-model AI review of the proofs. The 16 not AI-judged were verified by owner audit (typically over a confirmed judge false positive), not failures.

Construction of the Natural Numbers

1 · Prerequisites

2 · Summary

Objective. This page builds the natural numbers N\mathbb{N} from the ground and proves the facts that every later construction silently assumes. It is the prerequisite that sits beneath the two constructions of R\mathbb{R}, the Cauchy-sequence construction and the Dedekind-cut construction: both begin "the natural numbers, with their arithmetic and order, are the assumed starting point", and this page discharges exactly that assumption. The headline is the principle of mathematical induction (The principle of mathematical induction), the proof technique used everywhere in mathematics, obtained here from its set-theoretic source.

Two starting points, reconciled. There are two standard ways to say what the natural numbers are, and we use both. The Peano axioms (Peano system) describe N\mathbb{N} abstractly as a triple (N,0,σ)(N, 0, \sigma) with a zero, a successor, and an induction axiom. ZFC set theory instead builds a concrete model: the Axiom of Infinity yields an inductive set (Inductive set), and the smallest one (The natural numbers exist: a smallest inductive set) is the set ω\omega of von Neumann naturals (The natural numbers N\mathbb{N} (von Neumann)), where 0=0 = \varnothing and σ(n)=n{n}\sigma(n) = n \cup \{n\}, so each number is the set of its predecessors. The bridge between the two views is The von Neumann naturals form a Peano system: the von Neumann model satisfies the Peano axioms, so a Peano system provably exists. Everything after the bridge is proved from the Peano axioms alone, so it holds in any model.

Induction, and definition by recursion. Minimality of ω\omega is exactly the induction principle: a set of naturals that contains 00 and is closed under the successor is everything (The principle of mathematical induction). Its indispensable companion is the recursion theorem (The recursion theorem), which licenses defining a function on N\mathbb{N} by specifying its value at 00 and at each successor. Without it the recursive definitions of addition and multiplication would be unjustified; with it they are genuine total functions (Addition is a well-defined total operation, Multiplication is a well-defined total operation).

Arithmetic, proved. Addition (Addition of natural numbers) and multiplication (Multiplication of natural numbers) are defined by recursion, and their laws are then theorems, each proved by induction: addition is associative, commutative, and cancellative (Addition is associative, Addition is commutative, Addition is cancellative); multiplication distributes over addition and is associative and commutative (Distributivity and the successor law for multiplication, Multiplication is associative, Multiplication is commutative); and there are no zero divisors (The natural numbers have no zero divisors). The archetype of the method is the one-line unfolding 1+1=21 + 1 = 2 (1+1=21+1=2 from the recursion), with 2+2=42 + 2 = 4 (2+2=42+2=4 from the recursion) right behind it.

Order and its consequences. The order mn    k (m+k=n)m \le n \iff \exists k\ (m + k = n) (Order on the natural numbers) is a linear order (\le is a linear order on N\mathbb{N}) satisfying trichotomy (Trichotomy of the order on N\mathbb{N}), compatible with addition and multiplication (Order is compatible with addition, Order is compatible with multiplication) and discrete in that σ(n)\sigma(n) is the immediate successor of nn (Discreteness: σ(n)\sigma(n) is the immediate successor). Its deepest consequence is the well-ordering principle (The well-ordering principle): every nonempty set of naturals has a least element. From it comes strong induction (Strong (complete) induction), so ordinary induction, well-ordering and strong induction are available interchangeably in what follows.

Uniqueness. Finally, the natural numbers are pinned down: any two Peano systems are isomorphic by a unique isomorphism (Categoricity: the natural numbers are unique up to unique isomorphism), proved from the recursion theorem exactly as the uniqueness of R\mathbb{R} is proved one level up. So the set-theoretic encoding chosen at the start is immaterial, and it is legitimate to speak of the natural numbers.

Why the axioms are sharp. The optional items show that the induction axiom and the injectivity of the successor cannot be dropped. Dropping induction leaves room for extra elements a proof by induction would never reach (The induction axiom is independent of the other Peano axioms); dropping injectivity of the successor allows a finite model in which distinct numbers collapse (Injectivity of the successor is independent of the other Peano axioms); and induction genuinely needs its base case, since successor-closure alone does not force a set to be all of N\mathbb{N} (FALSE: successor-closure alone forces a set to be all of N\mathbb{N}).

3 · Logical flowchart

4 · Definitions, theorems and proofs

DefinitionDefinition: Literature-sourcedProof: Not applicableaudited 2026-07-25Open item page →

Peano system

Definition

A Peano system is a triple (N,0,σ)(N, 0, \sigma) consisting of a set NN, a distinguished element 0N0 \in N, and a successor function σ:NN\sigma : N \to N (A function is a relation ff with (a,b)f(a,b) \in f and (a,c)f(a,c) \in f implying b=cb = c; f:ABf : A \to B, the value f(a)f(a), domain and codomain), satisfying the three Peano axioms:

The elements of NN are the natural numbers of the system. We abbreviate 1:=σ(0)1 := \sigma(0), 2:=σ(1)2 := \sigma(1), 3:=σ(2)3 := \sigma(2), and so on.

Remarks

Axiom (P3), the induction axiom, is the load-bearing one: it says NN has no proper subset that contains 00 and is closed under σ\sigma, so every element is reached from 00 by finitely many successors. It is what makes definition by recursion and proof by induction legitimate (The principle of mathematical induction, The recursion theorem).

This is a structural definition: it does not say what the natural numbers are, only which laws they obey. Two questions are therefore separate and both require proof: that a Peano system exists (a model built inside set theory, The von Neumann naturals form a Peano system, from the von Neumann naturals The natural numbers N\mathbb{N} (von Neumann)), and that it is unique up to a unique isomorphism (Categoricity: the natural numbers are unique up to unique isomorphism). Together these justify speaking of the natural numbers N\mathbb{N}.

We use the modern 00-based convention (0N0 \in N). Peano's original postulates started from 11; the choice is immaterial, and starting at 00 matches the set-theoretic construction, where 0=0 = \varnothing.

DefinitionDefinition: Literature-sourcedProof: Not applicableaudited 2026-07-25Open item page →

Inductive set

Definition

Working in ZFC, for a set xx write x+:=x{x}x^{+} := x \cup \{x\} for its successor (a set, by the axioms of Pairing and Union, The Axiom of Pairing: xyzt(tz(t=xt=y))\forall x\,\forall y\,\exists z\,\forall t\,(t \in z \leftrightarrow (t = x \vee t = y)) and The Axiom of Union: xyz(zys(sxzs))\forall x\,\exists y\,\forall z\,(z \in y \leftrightarrow \exists s\,(s \in x \wedge z \in s))). A set II is inductive when

Iand(xI    x+I) for every x.\varnothing \in I \quad\text{and}\quad \bigl(x \in I \implies x^{+} \in I\bigr)\ \text{for every } x.

That is, II contains the empty set and is closed under the successor operation.

Remarks

The Axiom of Infinity (The Axiom of Infinity: there is a set containing a set with no elements and closed under yy{y}y \mapsto y \cup \{y\}) is precisely the assertion that an inductive set exists. Inductive sets can be large and are far from unique (if II is inductive so is I{anything}I \cup \{\text{anything}\} once closed off), so an inductive set is not yet a good definition of N\mathbb{N}. The natural numbers are carved out as the smallest inductive set, the intersection of all of them (The natural numbers exist: a smallest inductive set, The natural numbers N\mathbb{N} (von Neumann)); minimality is what delivers the induction principle.

Here 0:=0 := \varnothing and the successor of nn is σ(n):=n+=n{n}\sigma(n) := n^{+} = n \cup \{n\}, so 1={0}1 = \{0\}, 2={0,1}2 = \{0,1\}, 3={0,1,2}3 = \{0,1,2\}: each natural number is the set of all smaller natural numbers.

LemmaStatement: AI-adaptedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

The natural numbers exist: a smallest inductive set

Statement

There is a set ω\omega that is inductive (Inductive set) and is a subset of every inductive set; it is unique. This ω\omega is the set of natural numbers.

Facts & Assumptions

Proof

technique · direct
1.1

By the Axiom of Infinity fix an inductive set I0I_0.

given
1.2

By Separation the collection ω:={xI0:xJ for every inductive set J}\omega := \{x \in I_0 : x \in J \text{ for every inductive set } J\} is a set.

givenconstruct
2.1

ω\omega is inductive: J\varnothing \in J for every inductive JJ (so I0\varnothing \in I_0 and ω\varnothing \in \omega), and if xωx \in \omega then xJx \in J for every inductive JJ, hence x+Jx^{+} \in J for every inductive JJ, and x+I0x^{+} \in I_0 since xI0x \in I_0 and I0I_0 is inductive, so x+ωx^{+} \in \omega.

step 1.2
2.2

ωJ\omega \subseteq J for every inductive JJ: any xωx \in \omega satisfies xJx \in J by definition.

step 1.2
3.1

Uniqueness: if ω\omega' is also inductive and contained in every inductive set, then ωω\omega \subseteq \omega' (as ω\omega' is inductive) and ωω\omega' \subseteq \omega (as ω\omega is inductive), so ω=ω\omega = \omega' by Extensionality.

step 2.1step 2.2
DefinitionDefinition: Literature-sourcedProof: Not applicableaudited 2026-07-25Open item page →

The natural numbers N\mathbb{N} (von Neumann)

Definition

The set of natural numbers is the smallest inductive set (Inductive set),

N=ω:={I:I is inductive},\mathbb{N} = \omega := \bigcap\,\{\, I : I \text{ is inductive} \,\},

which exists and is itself inductive by The natural numbers exist: a smallest inductive set (the Axiom of Infinity, The Axiom of Infinity: there is a set containing a set with no elements and closed under yy{y}y \mapsto y \cup \{y\}, supplies one inductive set to intersect within, and Separation, The Axiom Schema of Separation: for each formula φ\varphi, pˉxyz(zy(zxφ(z,pˉ)))\forall \bar p\,\forall x\,\exists y\,\forall z\,(z \in y \leftrightarrow (z \in x \wedge \varphi(z,\bar p))), makes the intersection a set). On N\mathbb{N} we take

0:=,σ(n):=n{n},0 := \varnothing, \qquad \sigma(n) := n \cup \{n\},

the distinguished element and the successor function. Thus 0=0 = \varnothing, 1={0}1 = \{0\}, 2={0,1}2 = \{0, 1\}, 3={0,1,2}3 = \{0, 1, 2\}, and in general n={0,1,,n1}n = \{0, 1, \dots, n-1\} is the set of its predecessors.

Remarks

"Smallest" means ωI\omega \subseteq I for every inductive set II. This minimality is exactly the induction principle (The principle of mathematical induction): a subset of N\mathbb{N} that contains 00 and is closed under σ\sigma is itself inductive, hence contains ω=N\omega = \mathbb{N}, hence equals N\mathbb{N}.

With 00 and σ\sigma so defined, (N,0,σ)(\mathbb{N}, 0, \sigma) satisfies the Peano axioms (Peano system, proved in The von Neumann naturals form a Peano system), so it is a model of the abstract natural numbers. By categoricity (Categoricity: the natural numbers are unique up to unique isomorphism) any other model is uniquely isomorphic to it, so the particular set-theoretic encoding chosen here is immaterial to every arithmetic and order property that follows: those are developed from the Peano axioms, not from the sets themselves.

TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

The principle of mathematical induction

Statement

Let SNS \subseteq \mathbb{N}. If 0S0 \in S and σ(n)S\sigma(n) \in S whenever nSn \in S, then S=NS = \mathbb{N}. Equivalently (property form): if a property PP satisfies P(0)P(0) and (P(n)P(σ(n))P(n) \Rightarrow P(\sigma(n))) for all nn, then P(n)P(n) holds for all nNn \in \mathbb{N}. This is the induction principle, the basis of proof by induction.

Facts & Assumptions

Given: N=ω\mathbb{N} = \omega is the smallest inductive set (Inductive set, The natural numbers N\mathbb{N} (von Neumann), The natural numbers exist: a smallest inductive set); 0=0 = \varnothing, σ(n)=n{n}\sigma(n) = n \cup \{n\}.

[L1]

ω\omega is contained in every inductive set (The natural numbers exist: a smallest inductive set).

Proof

technique · direct
1.1

Suppose SNS \subseteq \mathbb{N} with 0=S0 = \varnothing \in S and (xSσ(x)=x{x}Sx \in S \Rightarrow \sigma(x) = x \cup \{x\} \in S).

given
2.1

Then SS contains \varnothing and is closed under xx{x}x \mapsto x \cup \{x\}, so SS is an inductive set.

step 1.1
3.1

N=ωS\mathbb{N} = \omega \subseteq S because ω\omega is a subset of every inductive set.

L1step 2.1
4.1

With SNS \subseteq \mathbb{N} by hypothesis, S=NS = \mathbb{N}.

step 3.1given
5.1

The property form is the set case with S={nN:P(n)}S = \{n \in \mathbb{N} : P(n)\}: P(0)P(0) and P(n)P(σ(n))P(n) \Rightarrow P(\sigma(n)) give 0S0 \in S and σ\sigma-closure, so S=NS = \mathbb{N}, i.e. PP holds for all nn; conversely the set form is the property form applied to P(n):nSP(n) :\equiv n \in S, so the two are equivalent.

step 4.1
LemmaStatement: AI-adaptedProof: AI-generatedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Every natural number is a transitive set and is not a member of itself

Statement

For every nNn \in \mathbb{N}: (a) nn is transitive, i.e. xnxnx \in n \Rightarrow x \subseteq n; and (b) nnn \notin n.

Facts & Assumptions

Given: N\mathbb{N} (The natural numbers N\mathbb{N} (von Neumann)) with 0=0 = \varnothing, σ(n)=n{n}\sigma(n) = n \cup \{n\}; the induction principle (The principle of mathematical induction). Let P(n)P(n) be the conjunction "nn is transitive and nnn \notin n".

[L1]

Induction: if P(0)P(0) and P(n)P(σ(n))P(n) \Rightarrow P(\sigma(n)), then PP holds for all nn (The principle of mathematical induction).

Proof

technique · induction
1.1

Base P(0)P(0): 0=0 = \varnothing has no members, so it is vacuously transitive and \varnothing \notin \varnothing.

base
1.2

Inductive hypothesis: P(n)P(n) holds, i.e. nn is transitive and nnn \notin n.

ih
2.1

σ(n)=n{n}\sigma(n) = n \cup \{n\} is transitive: take xσ(n)x \in \sigma(n); if xnx \in n then xnσ(n)x \subseteq n \subseteq \sigma(n) by the inductive hypothesis transitivity; if x=nx = n then x=nσ(n)x = n \subseteq \sigma(n) since nn{n}n \subseteq n \cup \{n\}; either way xσ(n)x \subseteq \sigma(n).

step 1.2
2.2

σ(n)σ(n)\sigma(n) \notin \sigma(n): suppose σ(n)σ(n)=n{n}\sigma(n) \in \sigma(n) = n \cup \{n\}, so σ(n)n\sigma(n) \in n or σ(n)=n\sigma(n) = n; if σ(n)=n\sigma(n) = n then nnn \in n (as n{n}σ(n)=nn \in \{n\} \subseteq \sigma(n) = n), contradicting the inductive hypothesis nnn \notin n; if σ(n)n\sigma(n) \in n then σ(n)n\sigma(n) \subseteq n by transitivity of nn, and since nn{n}=σ(n)nn \in n \cup \{n\} = \sigma(n) \subseteq n we get nnn \in n, again contradicting nnn \notin n; hence σ(n)σ(n)\sigma(n) \notin \sigma(n).

step 1.2
3.1

Hence P(σ(n))P(\sigma(n)) holds; by induction [L1], P(n)P(n) for all nNn \in \mathbb{N}: every natural number is transitive and satisfies nnn \notin n.

step 2.1step 2.2discharge-induction
TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

The von Neumann naturals form a Peano system

Statement

The triple (N,0,σ)=(ω,,+)(\mathbb{N}, 0, \sigma) = (\omega, \varnothing, {\cdot}^{+}) satisfies the three Peano axioms (Peano system): (P1) σ(n)0\sigma(n) \neq 0 for all nn; (P2) σ\sigma is injective; (P3) induction. Hence a Peano system exists.

Facts & Assumptions

Given: N=ω\mathbb{N} = \omega, 0=0 = \varnothing, σ(n)=n{n}\sigma(n) = n \cup \{n\} (The natural numbers N\mathbb{N} (von Neumann)).

[L1]
[L2]

Every natural number is transitive and satisfies nnn \notin n (Every natural number is a transitive set and is not a member of itself).

Proof

technique · direct
1.1

(P3) The induction axiom holds: it is exactly the induction principle.

L1
1.2

(P1) For every nn, σ(n)=n{n}\sigma(n) = n \cup \{n\} contains nn, so σ(n)\sigma(n) is nonempty and σ(n)=0\sigma(n) \neq \varnothing = 0.

given
1.3

(P2) Suppose σ(m)=σ(n)\sigma(m) = \sigma(n), i.e. m{m}=n{n}m \cup \{m\} = n \cup \{n\}; then mn{n}m \in n \cup \{n\} so mnm \in n or m=nm = n, and symmetrically nmn \in m or n=mn = m; if mnm \neq n then mnm \in n and nmn \in m, whence by transitivity [L2] mnm \in n gives mnm \subseteq n, and nmnn \in m \subseteq n gives nnn \in n, contradicting nnn \notin n [L2]; therefore m=nm = n and σ\sigma is injective.

L2given
2.1

P1, P2, P3 all hold, so (N,0,σ)(\mathbb{N}, 0, \sigma) is a Peano system; in particular a Peano system exists.

step 1.1step 1.2step 1.3
TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

The recursion theorem

Statement

Let (N,0,σ)(N,0,\sigma) be a Peano system (Peano system), in particular the natural numbers N\mathbb{N} (The natural numbers N\mathbb{N} (von Neumann)). For any set AA, any element aAa \in A, and any function f:AAf : A \to A, there is a unique function g:NAg : N \to A such that g(0)=ag(0) = a and g(σ(n))=f(g(n))g(\sigma(n)) = f(g(n)) for all nNn \in N.

Facts & Assumptions

Given: a Peano system (N,0,σ)(N,0,\sigma), a set AA, an element aAa \in A, and a function f:AAf : A \to A. Call a set RN×AR \subseteq N \times A closed when (0,a)R(0,a) \in R and, for all nNn \in N and xAx \in A, (n,x)R(n,x) \in R implies (σ(n),f(x))R(\sigma(n),f(x)) \in R.

[L1]

Induction holds in NN: any SNS \subseteq N with 0S0 \in S that is closed under σ\sigma equals NN (axiom P3, The principle of mathematical induction).

[L2]

In a Peano system σ(n)0\sigma(n) \ne 0 (P1) and σ\sigma is injective (P2) (Peano system).

Proof

technique · direct
1.1

The set N×AN \times A is closed, so the collection of closed subsets of N×AN \times A is nonempty; let g:={R:RN×A closed}g := \bigcap \{ R : R \subseteq N \times A \text{ closed} \}, which is itself closed, since if every closed RR contains (0,a)(0,a) then so does gg, and if (n,x)g(n,x) \in g then (n,x)R(n,x) \in R for each closed RR, whence (σ(n),f(x))R(\sigma(n),f(x)) \in R for each such RR, so (σ(n),f(x))g(\sigma(n),f(x)) \in g.

givenconstruct
2.1

Let D={nN:xA, (n,x)g}D = \{ n \in N : \exists x \in A,\ (n,x) \in g \}; then 0D0 \in D since (0,a)g(0,a) \in g, and if nDn \in D with (n,x)g(n,x) \in g then (σ(n),f(x))g(\sigma(n),f(x)) \in g by closure, so σ(n)D\sigma(n) \in D; by induction D=ND = N, hence gg relates every nn to at least one xx.

L1step 1.1
2.2

Suppose (0,b)g(0,b) \in g with bab \ne a; then g{(0,b)}g \setminus \{(0,b)\} still contains (0,a)(0,a) and is closed, since any generated pair (σ(m),f(z))(\sigma(m),f(z)) has σ(m)0\sigma(m) \ne 0 by P1 and so is never equal to (0,b)(0,b); this contradicts the minimality of gg, so aa is the unique second coordinate paired with 00.

L2step 1.1
3.1

Let U={nN:there is exactly one x with (n,x)g}U = \{ n \in N : \text{there is exactly one } x \text{ with } (n,x) \in g \}; then 0U0 \in U by step 2.2.

step 2.2
4.1

Assume nUn \in U with unique value xx, so (n,x)g(n,x) \in g and (σ(n),f(x))g(\sigma(n),f(x)) \in g by closure; if (σ(n),c)g(\sigma(n),c) \in g with cf(x)c \ne f(x), then g{(σ(n),c)}g \setminus \{(\sigma(n),c)\} contains (0,a)(0,a) because σ(n)0\sigma(n) \ne 0, and is closed, because a generated pair (σ(m),f(z))(\sigma(m),f(z)) with (m,z)g{(σ(n),c)}(m,z) \in g \setminus \{(\sigma(n),c)\} can equal (σ(n),c)(\sigma(n),c) only if σ(m)=σ(n)\sigma(m) = \sigma(n) and f(z)=cf(z) = c, but σ(m)=σ(n)\sigma(m) = \sigma(n) forces m=nm = n by injectivity, then z=xz = x by uniqueness at nn, giving f(z)=f(x)cf(z) = f(x) \ne c, a contradiction; hence g{(σ(n),c)}g \setminus \{(\sigma(n),c)\} is closed and smaller than gg, contradicting minimality, so f(x)f(x) is the unique value at σ(n)\sigma(n) and σ(n)U\sigma(n) \in U.

L2step 3.1step 1.1
5.1

By induction U=NU = N, so together with the totality from step 2.1 the relation gg is a function g:NAg : N \to A; it satisfies g(0)=ag(0) = a, and closure with uniqueness gives g(σ(n))=f(g(n))g(\sigma(n)) = f(g(n)) for all nn.

L1step 2.1step 3.1step 4.1
6.1

If h:NAh : N \to A also satisfies h(0)=ah(0) = a and h(σ(n))=f(h(n))h(\sigma(n)) = f(h(n)), then the set {n:g(n)=h(n)}\{ n : g(n) = h(n) \} contains 00 and is closed under σ\sigma, since g(n)=h(n)g(n) = h(n) implies g(σ(n))=f(g(n))=f(h(n))=h(σ(n))g(\sigma(n)) = f(g(n)) = f(h(n)) = h(\sigma(n)), so by induction it equals NN, that is h=gh = g.

L1step 5.1
CorollaryStatement: AI-adaptedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Addition is a well-defined total operation

Statement

For each mNm \in \mathbb{N} there is a unique function sm:NNs_m : \mathbb{N} \to \mathbb{N} with sm(0)=ms_m(0) = m and sm(σ(n))=σ(sm(n))s_m(\sigma(n)) = \sigma(s_m(n)). Writing m+n:=sm(n)m + n := s_m(n), addition +:N×NN+ : \mathbb{N} \times \mathbb{N} \to \mathbb{N} is a well-defined total binary operation satisfying m+0=mm + 0 = m and m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n).

Facts & Assumptions

Given: the natural numbers N\mathbb{N} with successor σ\sigma.

[L1]

The recursion theorem: for a set AA, an element aAa \in A, and a function f:AAf : A \to A there is a unique g:NAg : \mathbb{N} \to A with g(0)=ag(0) = a and g(σ(n))=f(g(n))g(\sigma(n)) = f(g(n)) (The recursion theorem).

[L2]

(N,0,σ)(\mathbb{N}, 0, \sigma) is a Peano system (The von Neumann naturals form a Peano system).

Proof

technique · direct
1.1

Fix mNm \in \mathbb{N}; since (N,0,σ)(\mathbb{N}, 0, \sigma) is a Peano system [L2], apply the recursion theorem with A=NA = \mathbb{N}, a=ma = m, and f=σf = \sigma: there is a unique function sm:NNs_m : \mathbb{N} \to \mathbb{N} with sm(0)=ms_m(0) = m and sm(σ(n))=σ(sm(n))s_m(\sigma(n)) = \sigma(s_m(n)).

L1L2
2.1

Define m+n:=sm(n)m + n := s_m(n); since sms_m is a total function on N\mathbb{N} for every mNm \in \mathbb{N}, the assignment (m,n)m+n(m,n) \mapsto m + n is a well-defined function N×NN\mathbb{N} \times \mathbb{N} \to \mathbb{N}.

step 1.1
3.1

The recursion equations become m+0=sm(0)=mm + 0 = s_m(0) = m and m+σ(n)=sm(σ(n))=σ(sm(n))=σ(m+n)m + \sigma(n) = s_m(\sigma(n)) = \sigma(s_m(n)) = \sigma(m + n), the defining clauses of addition.

step 1.1step 2.1
DefinitionDefinition: Literature-sourcedProof: Not applicableaudited 2026-07-25Open item page →

Addition of natural numbers

Definition

Addition +:N×NN+ : \mathbb{N} \times \mathbb{N} \to \mathbb{N} is defined by recursion on the second argument: for each fixed mNm \in \mathbb{N},

m+0=m,m+σ(n)=σ(m+n).m + 0 = m, \qquad m + \sigma(n) = \sigma(m + n).

Remarks

This is a definition by recursion, and it requires justification: it does not directly say what m+nm + n is, but pins down m+0m + 0 and reduces m+σ(n)m + \sigma(n) to m+nm + n. The recursion theorem (The recursion theorem) guarantees that for each mm there is exactly one function nm+nn \mapsto m + n obeying these two equations, so ++ is a well-defined total binary operation (Addition is a well-defined total operation).

Taking m=n=0m = n = 0 gives 0+0=00 + 0 = 0. The equation m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m+n) with n=0n = 0 gives m+1=m+σ(0)=σ(m+0)=σ(m)m + 1 = m + \sigma(0) = \sigma(m + 0) = \sigma(m), so σ(m)=m+1\sigma(m) = m + 1: the successor is "add one", recovering the familiar reading of the Peano successor. The basic laws (commutativity, associativity, cancellation, and the left identity 0+n=n0 + n = n) are not built in and must be proved by induction (Left identity for addition, Addition is commutative, Addition is associative, Addition is cancellative).

CorollaryStatement: AI-adaptedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Multiplication is a well-defined total operation

Statement

For each mNm \in \mathbb{N} there is a unique function pm:NNp_m : \mathbb{N} \to \mathbb{N} with pm(0)=0p_m(0) = 0 and pm(σ(n))=pm(n)+mp_m(\sigma(n)) = p_m(n) + m. Writing mn:=pm(n)m \cdot n := p_m(n), multiplication :N×NN\cdot : \mathbb{N} \times \mathbb{N} \to \mathbb{N} is a well-defined total binary operation satisfying m0=0m \cdot 0 = 0 and mσ(n)=mn+mm \cdot \sigma(n) = m \cdot n + m.

Facts & Assumptions

Given: N\mathbb{N} with successor σ\sigma and addition ++ (Addition of natural numbers).

[L1]

The recursion theorem (The recursion theorem).

[L2]

For fixed mm, the map addm:xx+m\mathrm{add}_m : x \mapsto x + m is a function NN\mathbb{N} \to \mathbb{N} (Addition of natural numbers).

[L3]

(N,0,σ)(\mathbb{N}, 0, \sigma) is a Peano system (The von Neumann naturals form a Peano system).

Proof

technique · direct
1.1

Fix mNm \in \mathbb{N}; the map addm:NN\mathrm{add}_m : \mathbb{N} \to \mathbb{N}, addm(x)=x+m\mathrm{add}_m(x) = x + m, is a well-defined function.

L2
2.1

Since (N,0,σ)(\mathbb{N}, 0, \sigma) is a Peano system [L3], apply the recursion theorem with A=NA = \mathbb{N}, a=0a = 0, f=addmf = \mathrm{add}_m: there is a unique pm:NNp_m : \mathbb{N} \to \mathbb{N} with pm(0)=0p_m(0) = 0 and pm(σ(n))=addm(pm(n))=pm(n)+mp_m(\sigma(n)) = \mathrm{add}_m(p_m(n)) = p_m(n) + m.

L1L3step 1.1
3.1

Define mn:=pm(n)m \cdot n := p_m(n); as pmp_m is total for each mNm \in \mathbb{N}, the assignment (m,n)mn(m,n) \mapsto m \cdot n is a well-defined function N×NN\mathbb{N} \times \mathbb{N} \to \mathbb{N} with m0=0m \cdot 0 = 0 and mσ(n)=mn+mm \cdot \sigma(n) = m \cdot n + m.

step 2.1
DefinitionDefinition: Literature-sourcedProof: Not applicableaudited 2026-07-25Open item page →

Multiplication of natural numbers

Definition

Multiplication :N×NN\cdot : \mathbb{N} \times \mathbb{N} \to \mathbb{N} is defined by recursion on the second argument, using addition (Addition of natural numbers): for each fixed mNm \in \mathbb{N},

m0=0,mσ(n)=mn+m.m \cdot 0 = 0, \qquad m \cdot \sigma(n) = m \cdot n + m.

We write mnmn for mnm \cdot n when no confusion arises.

Remarks

As with addition, this is a recursion and is legitimate by the recursion theorem (The recursion theorem): for each mm the map nmnn \mapsto m \cdot n is the unique function with m0=0m \cdot 0 = 0 and mσ(n)=mn+mm \cdot \sigma(n) = m \cdot n + m, so \cdot is a well-defined total operation (Multiplication is a well-defined total operation).

The base clause m0=0m \cdot 0 = 0 builds in absorption on the right; that 0n=00 \cdot n = 0 on the left, and 1n=n=n11 \cdot n = n = n \cdot 1, are proved by induction (Zero and one under multiplication). Multiplication distributes over addition (Distributivity and the successor law for multiplication) and is associative and commutative (Multiplication is associative, Multiplication is commutative); with The natural numbers have no zero divisors and Cancellation for multiplication by a nonzero factor this makes (N,+,)(\mathbb{N}, +, \cdot) a commutative semiring in which cancellation holds.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

No natural number equals its own successor

Statement

For every nNn \in \mathbb{N}, nσ(n)n \neq \sigma(n).

Facts & Assumptions

Given: N\mathbb{N} with distinguished element 00 and successor σ\sigma.

[L1]

σ(n)0\sigma(n) \neq 0 for all nn (P1) and σ\sigma is injective (P2) (The von Neumann naturals form a Peano system).

[L2]

The induction principle (The principle of mathematical induction).

Proof

technique · induction
1.1

Base case n=0n = 0: σ(0)0\sigma(0) \neq 0 by P1, so 0σ(0)0 \neq \sigma(0).

baseL1
1.2

Inductive hypothesis: nσ(n)n \neq \sigma(n).

ih
2.1

If σ(n)=σ(σ(n))\sigma(n) = \sigma(\sigma(n)) then injectivity of σ\sigma gives n=σ(n)n = \sigma(n), contradicting the hypothesis; hence σ(n)σ(σ(n))\sigma(n) \neq \sigma(\sigma(n)).

step 1.2L1
3.1

By induction [L2], nσ(n)n \neq \sigma(n) for all nNn \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Every nonzero natural number is a successor

Statement

For every nNn \in \mathbb{N} with n0n \neq 0, there exists mNm \in \mathbb{N} with n=σ(m)n = \sigma(m). Equivalently, every natural number is either 00 or a successor.

Facts & Assumptions

Given: N\mathbb{N} with distinguished element 00 and successor σ\sigma. Let S={nN:n=0 or mN (n=σ(m))}S = \{n \in \mathbb{N} : n = 0 \text{ or } \exists m \in \mathbb{N}\ (n = \sigma(m))\}.

[L1]

The induction principle (The principle of mathematical induction).

Proof

technique · induction
1.1

Base case: 0S0 \in S since 0=00 = 0.

base
1.2

Inductive hypothesis: nSn \in S.

ih
2.1

σ(n)\sigma(n) is the successor of nn, so σ(n)S\sigma(n) \in S (it satisfies m, σ(n)=σ(m)\exists m,\ \sigma(n) = \sigma(m) with m=nm = n).

step 1.2
3.1

By induction [L1], S=NS = \mathbb{N}, so every nNn \in \mathbb{N} is 00 or a successor; if n0n \neq 0 then n=σ(m)n = \sigma(m) for some mm.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Left identity for addition

Statement

For all nNn \in \mathbb{N}: 0+n=n0 + n = n. The right identity n+0=nn + 0 = n is definitional, so together with this law 00 is a two-sided identity for addition.

Facts & Assumptions

Given: Addition with m+0=mm + 0 = m and m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers).

[L1]

Proof

technique · induction
1.1

Base n=0n = 0: 0+0=00 + 0 = 0 by the clause m+0=mm + 0 = m at m=0m = 0.

base
1.2

Inductive hypothesis: 0+n=n0 + n = n.

ih
2.1

Step: 0+σ(n)=σ(0+n)=σ(n)0 + \sigma(n) = \sigma(0 + n) = \sigma(n), using m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) at m=0m = 0, then the hypothesis.

step 1.2given
3.1

By induction [L1], 0+n=n0 + n = n for all nNn \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Left successor law for addition

Statement

For all m,nNm, n \in \mathbb{N}: σ(m)+n=σ(m+n)\sigma(m) + n = \sigma(m + n). (The recursion defines addition on the second argument; this is the matching law for adding on the first.)

Facts & Assumptions

Given: Addition with m+0=mm + 0 = m and m+σ(k)=σ(m+k)m + \sigma(k) = \sigma(m + k) (Addition of natural numbers).

[L1]

Proof

technique · induction
1.1

Base n=0n = 0: σ(m)+0=σ(m)\sigma(m) + 0 = \sigma(m) by x+0=xx + 0 = x, and σ(m+0)=σ(m)\sigma(m + 0) = \sigma(m) by m+0=mm + 0 = m, so σ(m)+0=σ(m+0)\sigma(m) + 0 = \sigma(m + 0).

base
1.2

Inductive hypothesis: σ(m)+n=σ(m+n)\sigma(m) + n = \sigma(m + n).

ih
2.1

Step: σ(m)+σ(n)=σ(σ(m)+n)=σ(σ(m+n))=σ(m+σ(n))\sigma(m) + \sigma(n) = \sigma(\sigma(m) + n) = \sigma(\sigma(m + n)) = \sigma(m + \sigma(n)), applying x+σ(k)=σ(x+k)x + \sigma(k) = \sigma(x + k) at the outer sum, then the hypothesis, then m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) backward inside σ\sigma.

step 1.2given
3.1

By induction [L1], σ(m)+n=σ(m+n)\sigma(m) + n = \sigma(m + n) for all nn, and since mm was arbitrary, for all m,nNm, n \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Addition is associative

Statement

For all a,b,cNa, b, c \in \mathbb{N}: (a+b)+c=a+(b+c)(a + b) + c = a + (b + c).

Facts & Assumptions

Given: Addition with m+0=mm + 0 = m and m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers).

[L1]

Proof

technique · induction
1.1

Base c=0c = 0: (a+b)+0=a+b(a + b) + 0 = a + b and a+(b+0)=a+ba + (b + 0) = a + b, so they are equal.

base
1.2

Inductive hypothesis: (a+b)+c=a+(b+c)(a + b) + c = a + (b + c).

ih
2.1

Step: (a+b)+σ(c)=σ((a+b)+c)=σ(a+(b+c))=a+σ(b+c)=a+(b+σ(c))(a + b) + \sigma(c) = \sigma((a + b) + c) = \sigma(a + (b + c)) = a + \sigma(b + c) = a + (b + \sigma(c)), using m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) twice and the hypothesis.

step 1.2
3.1

By induction [L1], (a+b)+c=a+(b+c)(a + b) + c = a + (b + c) for all cc, hence for all a,b,cNa, b, c \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Addition is commutative

Statement

For all m,nNm, n \in \mathbb{N}: m+n=n+mm + n = n + m.

Facts & Assumptions

Given: Addition (Addition of natural numbers).

[L1]

0+k=k0 + k = k for all kk (Left identity for addition).

[L2]

Left successor law: σ(a)+k=σ(a+k)\sigma(a) + k = \sigma(a + k) for all a,ka, k (Left successor law for addition).

[L3]

Proof

technique · induction
1.1

Base n=0n = 0: m+0=mm + 0 = m and 0+m=m0 + m = m by [L1], so m+0=0+mm + 0 = 0 + m.

baseL1
1.2

Inductive hypothesis: m+n=n+mm + n = n + m.

ih
2.1

Step: m+σ(n)=σ(m+n)=σ(n+m)m + \sigma(n) = \sigma(m + n) = \sigma(n + m) by the recursion and the hypothesis; and the left successor law [L2] gives σ(n)+m=σ(n+m)\sigma(n) + m = \sigma(n + m), so m+σ(n)=σ(n+m)=σ(n)+mm + \sigma(n) = \sigma(n + m) = \sigma(n) + m.

step 1.2givenL2
3.1

By induction [L3], m+n=n+mm + n = n + m for all nn, hence for all m,nNm, n \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Addition is cancellative

Statement

For all m,n,kNm, n, k \in \mathbb{N}: if m+k=n+km + k = n + k then m=nm = n.

Facts & Assumptions

Given: Addition (Addition of natural numbers).

[L1]

σ\sigma is injective (P2) (The von Neumann naturals form a Peano system).

[L2]

Proof

technique · induction
1.1

Base k=0k = 0: if m+0=n+0m + 0 = n + 0 then m=nm = n, since m+0=mm + 0 = m and n+0=nn + 0 = n.

base
1.2

Inductive hypothesis: for all m,nm, n, m+k=n+km=nm + k = n + k \Rightarrow m = n.

ih
2.1

Step: suppose m+σ(k)=n+σ(k)m + \sigma(k) = n + \sigma(k); then σ(m+k)=σ(n+k)\sigma(m + k) = \sigma(n + k) by m+σ(j)=σ(m+j)m + \sigma(j) = \sigma(m + j), so m+k=n+km + k = n + k by injectivity [L1], hence m=nm = n by the hypothesis.

step 1.2L1
3.1

By induction [L2], cancellation holds for all kk, hence for all m,n,kNm, n, k \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: AI-adaptedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Zero and one under multiplication

Statement

For all nNn \in \mathbb{N}: 0n=00 \cdot n = 0, 1n=n1 \cdot n = n, and n1=nn \cdot 1 = n. Here 1=σ(0)1 = \sigma(0), and the clause n0=0n \cdot 0 = 0 is definitional.

Facts & Assumptions

Given: multiplication m0=0m \cdot 0 = 0 and mσ(n)=mn+mm \cdot \sigma(n) = m \cdot n + m (Multiplication of natural numbers), and addition with m+0=mm + 0 = m and m+σ(k)=σ(m+k)m + \sigma(k) = \sigma(m + k) (Addition of natural numbers).

[L1]

0+k=k0 + k = k for all kNk \in \mathbb{N} (Left identity for addition).

[L2]

The induction principle (The principle of mathematical induction).

Proof

technique · induction on $n$
1.1

Right unit n1=nn \cdot 1 = n: n1=nσ(0)=n0+n=0+n=nn \cdot 1 = n \cdot \sigma(0) = n \cdot 0 + n = 0 + n = n, using the multiplication recursion then [L1].

givenL1
1.2

Base of the left inductions: 00=00 \cdot 0 = 0 and 10=01 \cdot 0 = 0 by the clause m0=0m \cdot 0 = 0.

base
1.3

Inductive hypothesis: 0n=00 \cdot n = 0 and 1n=n1 \cdot n = n.

ih
2.1

Step: 0σ(n)=0n+0=0+0=00 \cdot \sigma(n) = 0 \cdot n + 0 = 0 + 0 = 0, using the hypothesis 0n=00 \cdot n = 0 and 0+0=00 + 0 = 0 (the instance k=0k = 0 of [L1]); and 1σ(n)=1n+1=n+1=σ(n)1 \cdot \sigma(n) = 1 \cdot n + 1 = n + 1 = \sigma(n), where n+1=n+σ(0)=σ(n+0)=σ(n)n + 1 = n + \sigma(0) = \sigma(n + 0) = \sigma(n) by the addition recursion m+σ(k)=σ(m+k)m + \sigma(k) = \sigma(m + k) and n+0=nn + 0 = n.

step 1.3givenL1
3.1

By induction [L2], 0n=00 \cdot n = 0 and 1n=n1 \cdot n = n for all nn; together with step 1.1, n1=nn \cdot 1 = n for all nn.

step 1.1step 2.1discharge-induction
LemmaStatement: AI-adaptedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Distributivity and the successor law for multiplication

Statement

For all a,b,c,nNa, b, c, n \in \mathbb{N}: (left distributivity) a(b+c)=ab+aca \cdot (b + c) = a \cdot b + a \cdot c; and (successor-left law) σ(a)n=an+n\sigma(a) \cdot n = a \cdot n + n.

Facts & Assumptions

Given: multiplication m0=0m \cdot 0 = 0, mσ(n)=mn+mm \cdot \sigma(n) = m \cdot n + m and addition m+0=mm + 0 = m, m+σ(k)=σ(m+k)m + \sigma(k) = \sigma(m + k) (Multiplication of natural numbers, Addition of natural numbers); in particular the addition recursion a+σ(k)=σ(a+k)a + \sigma(k) = \sigma(a + k) is available.

[L1]

Addition is associative (Addition is associative) and commutative (Addition is commutative).

[L2]

The induction principle (The principle of mathematical induction).

Proof

technique · induction on $c$ for distributivity, with a second induction on $n$ for the successor law
1.1

Base c=0c = 0: a(b+0)=ab=ab+0=ab+a0a \cdot (b + 0) = a \cdot b = a \cdot b + 0 = a \cdot b + a \cdot 0, using b+0=bb + 0 = b and a0=0a \cdot 0 = 0.

base
1.2

Inductive hypothesis: a(b+c)=ab+aca \cdot (b + c) = a \cdot b + a \cdot c.

ih
1.3

Successor-left law σ(a)n=an+n\sigma(a) \cdot n = a \cdot n + n, by a second induction on nn: base n=0n = 0 gives σ(a)0=0=a0+0\sigma(a) \cdot 0 = 0 = a \cdot 0 + 0; assuming σ(a)n=an+n\sigma(a) \cdot n = a \cdot n + n, the step gives σ(a)σ(n)=σ(a)n+σ(a)=(an+n)+σ(a)=an+(n+σ(a))=an+σ(n+a)=an+σ(a+n)=an+(a+σ(n))=(an+a)+σ(n)=aσ(n)+σ(n)\sigma(a) \cdot \sigma(n) = \sigma(a) \cdot n + \sigma(a) = (a \cdot n + n) + \sigma(a) = a \cdot n + (n + \sigma(a)) = a \cdot n + \sigma(n + a) = a \cdot n + \sigma(a + n) = a \cdot n + (a + \sigma(n)) = (a \cdot n + a) + \sigma(n) = a \cdot \sigma(n) + \sigma(n), using associativity and commutativity [L1] and a+σ(k)=σ(a+k)a + \sigma(k) = \sigma(a + k).

L1given
2.1

Step: a(b+σ(c))=aσ(b+c)=a(b+c)+a=(ab+ac)+a=ab+(ac+a)=ab+aσ(c)a \cdot (b + \sigma(c)) = a \cdot \sigma(b + c) = a \cdot (b + c) + a = (a \cdot b + a \cdot c) + a = a \cdot b + (a \cdot c + a) = a \cdot b + a \cdot \sigma(c), using b+σ(c)=σ(b+c)b + \sigma(c) = \sigma(b + c), the multiplication recursion, the hypothesis, associativity [L1], and aσ(c)=ac+aa \cdot \sigma(c) = a \cdot c + a.

step 1.2L1
3.1

By induction [L2], a(b+c)=ab+aca \cdot (b + c) = a \cdot b + a \cdot c for all cc (hence all a,b,ca, b, c) and σ(a)n=an+n\sigma(a) \cdot n = a \cdot n + n for all a,na, n.

step 2.1step 1.3discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Multiplication is associative

Statement

For all a,b,cNa, b, c \in \mathbb{N}: (ab)c=a(bc)(a \cdot b) \cdot c = a \cdot (b \cdot c).

Facts & Assumptions

Given: multiplication and addition (Multiplication of natural numbers, Addition of natural numbers).

[L1]

Left distributivity a(b+c)=ab+aca \cdot (b + c) = a \cdot b + a \cdot c (Distributivity and the successor law for multiplication).

[L2]

The induction principle (The principle of mathematical induction).

Proof

technique · induction on $c$, with $a, b$ fixed
1.1

Base c=0c = 0: (ab)0=0(a \cdot b) \cdot 0 = 0 and a(b0)=a0=0a \cdot (b \cdot 0) = a \cdot 0 = 0, so the two sides are equal.

base
1.2

Inductive hypothesis: (ab)c=a(bc)(a \cdot b) \cdot c = a \cdot (b \cdot c).

ih
2.1

Step: (ab)σ(c)=(ab)c+ab=a(bc)+ab=a(bc+b)=a(bσ(c))(a \cdot b) \cdot \sigma(c) = (a \cdot b) \cdot c + a \cdot b = a \cdot (b \cdot c) + a \cdot b = a \cdot (b \cdot c + b) = a \cdot (b \cdot \sigma(c)), using the multiplication recursion, the hypothesis, left distributivity [L1], and bσ(c)=bc+bb \cdot \sigma(c) = b \cdot c + b.

step 1.2L1
3.1

By induction [L2], (ab)c=a(bc)(a \cdot b) \cdot c = a \cdot (b \cdot c) for all cc, hence for all a,b,cNa, b, c \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Multiplication is commutative

Statement

For all m,nNm, n \in \mathbb{N}: mn=nmm \cdot n = n \cdot m.

Facts & Assumptions

Given: multiplication (Multiplication of natural numbers).

[L1]

0k=00 \cdot k = 0 for all kNk \in \mathbb{N} (Zero and one under multiplication).

[L2]

The successor-left law σ(n)m=nm+m\sigma(n) \cdot m = n \cdot m + m (Distributivity and the successor law for multiplication).

[L3]

The induction principle (The principle of mathematical induction).

Proof

technique · induction on $n$, with $m$ fixed
1.1

Base n=0n = 0: m0=0m \cdot 0 = 0 and 0m=00 \cdot m = 0 by [L1], so m0=0mm \cdot 0 = 0 \cdot m.

baseL1
1.2

Inductive hypothesis: mn=nmm \cdot n = n \cdot m.

ih
2.1

Step: mσ(n)=mn+m=nm+m=σ(n)mm \cdot \sigma(n) = m \cdot n + m = n \cdot m + m = \sigma(n) \cdot m, using the multiplication recursion, the hypothesis, then the successor-left law [L2].

step 1.2L2
3.1

By induction [L3], mn=nmm \cdot n = n \cdot m for all nn, hence for all m,nNm, n \in \mathbb{N}.

step 1.1step 2.1discharge-induction
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

The natural numbers have no zero divisors

Statement

For all m,nNm,n\in\mathbb{N}: if mn=0m\cdot n=0 then m=0m=0 or n=0n=0. Equivalently, if m0m\ne 0 and n0n\ne 0 then mn0m\cdot n\ne 0.

Facts & Assumptions

Given: multiplication defined by m0=0m\cdot 0=0 and mσ(k)=mk+mm\cdot\sigma(k)=m\cdot k+m; addition defined by m+0=mm+0=m and m+σ(k)=σ(m+k)m+\sigma(k)=\sigma(m+k).

[L1]

Every nonzero natural is a successor: if x0x\ne 0 then x=σ(a)x=\sigma(a) for some aNa\in\mathbb{N} (Every nonzero natural number is a successor).

[L2]

σ(j)0\sigma(j)\ne 0 for all jNj\in\mathbb{N} (Peano axiom P1) (The von Neumann naturals form a Peano system).

Proof

technique · contrapositive
1.1

It suffices to prove the contrapositive: if m0m\ne 0 and n0n\ne 0 then mn0m\cdot n\ne 0.

contrapositive-reduce
1.2

Assume m0m\ne 0 and n0n\ne 0.

assume-hyp
2.1

By [L1] write m=σ(a)m=\sigma(a) and n=σ(b)n=\sigma(b) for some a,bNa,b\in\mathbb{N}.

L1step 1.2
3.1

Then mn=mσ(b)=mb+m=mb+σ(a)=σ(mb+a)m\cdot n=m\cdot\sigma(b)=m\cdot b+m=m\cdot b+\sigma(a)=\sigma(m\cdot b+a), using the multiplication recursion, m=σ(a)m=\sigma(a), and c+σ(a)=σ(c+a)c+\sigma(a)=\sigma(c+a).

step 2.1given
4.1

By [L2], σ(mb+a)0\sigma(m\cdot b+a)\ne 0, so mn0m\cdot n\ne 0; this proves the contrapositive, hence if mn=0m\cdot n=0 then m=0m=0 or n=0n=0.

step 3.1L2discharge-contrapositive
DefinitionDefinition: Literature-sourcedProof: Not applicableverified 2026-08-02 (claude-opus-5)Open item page →

Order on the natural numbers

Definition

For m,nNm, n \in \mathbb{N}, define the order by

mn    kN (m+k=n),m \le n \iff \exists k \in \mathbb{N}\ (m + k = n),

and the strict order by m<n    (mn and mn)m < n \iff (m \le n \text{ and } m \neq n), using addition (Addition of natural numbers).

Remarks

Read mnm \le n as "nn is mm plus something": the gap kk with m+k=nm + k = n is unique when it exists (by commutativity and cancellation, Addition is commutative, Addition is cancellative) and is written nmn - m once subtraction is available. That \le is genuinely a linear order, reflexive, antisymmetric, transitive, and total, is proved in \le is a linear order on N\mathbb{N} and Trichotomy of the order on N\mathbb{N}; compatibility with the arithmetic (Order is compatible with addition, Order is compatible with multiplication) and discreteness (m<n    σ(m)nm < n \iff \sigma(m) \le n, Discreteness: σ(n)\sigma(n) is the immediate successor) follow.

In the von Neumann model this order coincides with membership and inclusion: mn    mnm \le n \iff m \subseteq n and m<n    mnm < n \iff m \in n, matching the picture of each natural number as the set {0,1,,n1}\{0, 1, \dots, n-1\} of its predecessors (The natural numbers N\mathbb{N} (von Neumann)). That coincidence is not proved here; it is derived from the additive definition above on a later page. The least-element property of every nonempty subset, the well-ordering principle (The well-ordering principle), is the deepest consequence.

LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Trichotomy of the order on N\mathbb{N}

Statement

For all m,nNm, n \in \mathbb{N} exactly one of m<nm < n, m=nm = n, m>nm > n holds. In particular the order (Order on the natural numbers) is total.

Facts & Assumptions

Given: The order on N\mathbb{N}, where mnm \le n means k (m+k=n)\exists k\ (m + k = n) and m<nm < n means mnm \le n with mnm \ne n (Order on the natural numbers), and addition with m+0=mm + 0 = m, m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers).

[L1]

0+k=k0 + k = k (Left identity for addition).

[L2]

Left successor law: σ(a)+k=σ(a+k)\sigma(a) + k = \sigma(a + k) (Left successor law for addition).

[L3]

Addition is associative (Addition is associative).

[L4]

Every nonzero natural is a successor: n0n=σ(a)n \ne 0 \Rightarrow n = \sigma(a) for some aa (Every nonzero natural number is a successor).

[L5]

σ(x)0\sigma(x) \ne 0 for every xx (axiom P1 of The von Neumann naturals form a Peano system).

[L6]

Cancellation: x+k=y+kx=yx + k = y + k \Rightarrow x = y (Addition is cancellative).

[L7]

Addition is commutative (Addition is commutative).

Proof

technique · induction
1.1

At most one of the three holds where equality is involved: m<nm < n forces mnm \ne n and m>nm > n forces nmn \ne m, so m=nm = n is incompatible with either strict relation.

given
1.2

The two strict relations are incompatible: if m<nm < n and m>nm > n then m+j=nm + j = n and n+i=mn + i = m with i,j0i, j \ne 0, so m+(j+i)=(m+j)+i=n+i=m=m+0m + (j + i) = (m + j) + i = n + i = m = m + 0; by commutativity [L7], (j+i)+m=0+m(j + i) + m = 0 + m, so j+i=0j + i = 0 by cancellation [L6]; yet j=σ(a)j = \sigma(a) [L4] gives, by the left successor law [L2], j+i=σ(a)+i=σ(a+i)0j + i = \sigma(a) + i = \sigma(a + i) \ne 0 [L5], a contradiction.

givenL2L3L4L5L6L7
1.3

For every nn one has n<σ(n)n < \sigma(n), since n+σ(0)=σ(n+0)=σ(n)n + \sigma(0) = \sigma(n + 0) = \sigma(n) by the recursion, with σ(0)0\sigma(0) \ne 0 [L5].

givenL5
1.4

Base case n=0n = 0: for arbitrary mm, either m=0m = 0, giving m=nm = n, or m0m \ne 0 so 0+m=m0 + m = m [L1] with m0m \ne 0 gives 0<m0 < m, that is n<mn < m; so at least one of m<nm < n, m=nm = n, m>nm > n holds.

baseL1L4
1.5

Inductive hypothesis: fix nn and assume that for every mm at least one of m<nm < n, m=nm = n, m>nm > n holds.

ih
2.1

Successor step: for arbitrary mm, apply the hypothesis; if m<nm < n then n=m+kn = m + k with k0k \ne 0, and the recursion gives m+σ(k)=σ(m+k)=σ(n)m + \sigma(k) = \sigma(m + k) = \sigma(n), so σ(n)=m+σ(k)\sigma(n) = m + \sigma(k) with σ(k)0\sigma(k) \ne 0 [L5], giving m<σ(n)m < \sigma(n); if m=nm = n then m+σ(0)=σ(m+0)=σ(m)=σ(n)m + \sigma(0) = \sigma(m + 0) = \sigma(m) = \sigma(n), so σ(n)=m+σ(0)\sigma(n) = m + \sigma(0) with σ(0)0\sigma(0) \ne 0, giving m<σ(n)m < \sigma(n); if m>nm > n then m=n+im = n + i with i0i \ne 0, so i=σ(a)i = \sigma(a) [L4], and the recursion gives n+σ(a)=σ(n+a)n + \sigma(a) = \sigma(n + a) while the left successor law [L2] gives σ(n)+a=σ(n+a)\sigma(n) + a = \sigma(n + a), so m=n+σ(a)=σ(n+a)=σ(n)+am = n + \sigma(a) = \sigma(n + a) = \sigma(n) + a, whence a=0a = 0 gives m=σ(n)m = \sigma(n) and a0a \ne 0 gives σ(n)<m\sigma(n) < m; in every case at least one of m<σ(n)m < \sigma(n), m=σ(n)m = \sigma(n), m>σ(n)m > \sigma(n) holds.

step 1.5givenL2L4L5
3.1

By the induction principle with base 1.4, hypothesis 1.5, and step 2.1, comparability holds for all m,nm, n, and with the incompatibilities 1.1 and 1.2 exactly one of m<nm < n, m=nm = n, m>nm > n holds, so the order is total.

step 1.1step 1.2step 1.4step 2.1discharge-induction
TheoremStatement: AI-adaptedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

\le is a linear order on N\mathbb{N}

Statement

The relation \le on N\mathbb{N}, where mn    k (m+k=n)m \le n \iff \exists k\ (m + k = n) (Order on the natural numbers), is a linear (total) order: reflexive, antisymmetric, transitive, and total.

Facts & Assumptions

Given: The order mn    k (m+k=n)m \le n \iff \exists k\ (m + k = n) (Order on the natural numbers) and addition with m+0=mm + 0 = m, m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers).

[L1]

0+k=k0 + k = k and addition is associative (Left identity for addition, Addition is associative).

[L2]

Cancellation: m+k=n+km=nm + k = n + k \Rightarrow m = n (Addition is cancellative).

[L3]

Every nonzero natural is a successor: if j0j \ne 0 then j=σ(i)j = \sigma(i) for some ii (Every nonzero natural number is a successor).

[L4]

Trichotomy: exactly one of m<nm < n, m=nm = n, m>nm > n holds (Trichotomy of the order on N\mathbb{N}).

[L5]

Addition is commutative (Addition is commutative).

[L6]

σ(x)0\sigma(x) \ne 0 for every xx (Peano axiom P1) (The von Neumann naturals form a Peano system).

Proof

technique · direct
1.1

Reflexive: m+0=mm + 0 = m, so mmm \le m.

given
1.2

Transitive: if mnm \le n and npn \le p, say m+k=nm + k = n and n+j=pn + j = p, then m+(k+j)=(m+k)+j=n+j=pm + (k + j) = (m + k) + j = n + j = p, so mpm \le p.

givenL1
1.3

A sum is zero only if both summands are: if k+j=0k + j = 0 with j0j \ne 0 then j=σ(i)j = \sigma(i) and k+j=k+σ(i)=σ(k+i)0k + j = k + \sigma(i) = \sigma(k + i) \ne 0 [L6], a contradiction, so k=j=0k = j = 0.

L1L3L6
1.4

Total: by trichotomy one of m<nm < n, m=nm = n, m>nm > n holds, and each of these gives mnm \le n or nmn \le m.

L4
2.1

Antisymmetric: if mnm \le n and nmn \le m, say m+k=nm + k = n and n+j=mn + j = m, then m+(k+j)=(m+k)+j=n+j=m=m+0m + (k + j) = (m + k) + j = n + j = m = m + 0; commuting both sides [L5] gives (k+j)+m=0+m(k + j) + m = 0 + m, and cancelling the common right summand mm [L2] gives k+j=0k + j = 0, hence k=j=0k = j = 0 by step 1.3 and m=nm = n.

givenL1L2L5step 1.3
3.1

Reflexivity, antisymmetry, transitivity, and totality all hold, so \le is a linear order on N\mathbb{N}.

step 1.1step 1.2step 2.1step 1.4
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

Order is compatible with addition

Statement

For all m,n,kNm, n, k \in \mathbb{N}: mn    m+kn+km \le n \iff m + k \le n + k; and m<n    m+k<n+km < n \iff m + k < n + k (Order on the natural numbers).

Facts & Assumptions

Given: The order mn    j (m+j=n)m \le n \iff \exists j\ (m + j = n) (Order on the natural numbers) and addition with m+0=mm + 0 = m, m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers).

[L1]

Addition is associative and commutative (Addition is associative, Addition is commutative).

[L2]

Cancellation: m+k=n+km=nm + k = n + k \Rightarrow m = n (Addition is cancellative).

Proof

technique · direct
1.1

Forward: if mnm \le n, say m+j=nm + j = n, then (m+k)+j=(m+j)+k=n+k(m + k) + j = (m + j) + k = n + k, so m+kn+km + k \le n + k.

givenL1
1.2

Backward: if m+kn+km + k \le n + k, say (m+k)+j=n+k(m + k) + j = n + k, then (m+j)+k=n+k(m + j) + k = n + k, so m+j=nm + j = n by cancellation, giving mnm \le n.

givenL1L2
2.1

The strict forms follow: m<nm < n means mnm \le n and mnm \ne n, while m+k=n+k    m=nm + k = n + k \iff m = n by cancellation, so m<n    m+k<n+km < n \iff m + k < n + k.

step 1.1step 1.2L2
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Order is compatible with multiplication

Statement

For all m,n,kNm, n, k \in \mathbb{N}: if mnm \le n then mknkm \cdot k \le n \cdot k; and if in addition k0k \ne 0 and m<nm < n, then mk<nkm \cdot k < n \cdot k (Order on the natural numbers).

Facts & Assumptions

Given: The order mn    d (m+d=n)m \le n \iff \exists d\ (m + d = n), with m<nm < n meaning mnm \le n and mnm \ne n (Order on the natural numbers); addition with m+0=mm + 0 = m (Addition of natural numbers); and multiplication with m0=0m \cdot 0 = 0, mσ(n)=mn+mm \cdot \sigma(n) = m \cdot n + m (Multiplication of natural numbers).

[L1]

Right distributivity (m+d)k=mk+dk(m + d) \cdot k = m \cdot k + d \cdot k, from left distributivity and commutativity (Distributivity and the successor law for multiplication, Multiplication is commutative).

[L2]

No zero divisors: d0d \ne 0 and k0dk0k \ne 0 \Rightarrow d \cdot k \ne 0 (The natural numbers have no zero divisors).

[L3]

Cancellation for addition: x+k=y+kx=yx + k = y + k \Rightarrow x = y (Addition is cancellative).

[L4]

Addition is commutative: x+y=y+xx + y = y + x (Addition is commutative).

Proof

technique · direct
1.1

If mnm \le n, write m+d=nm + d = n; then nk=(m+d)k=mk+dkn \cdot k = (m + d) \cdot k = m \cdot k + d \cdot k by right distributivity, so mknkm \cdot k \le n \cdot k.

givenL1
2.1

If moreover m<nm < n then d0d \ne 0, for d=0d = 0 would give n=m+0=mn = m + 0 = m, contradicting mnm \ne n; then with k0k \ne 0 we get dk0d \cdot k \ne 0 by [L2], so mk+dk=nkm \cdot k + d \cdot k = n \cdot k with dk0d \cdot k \ne 0; and mknkm \cdot k \ne n \cdot k, since equality would give mk+dk=mk+0m \cdot k + d \cdot k = m \cdot k + 0, hence dk+mk=0+mkd \cdot k + m \cdot k = 0 + m \cdot k by [L4] and dk=0d \cdot k = 0 by [L3], a contradiction; therefore mk<nkm \cdot k < n \cdot k.

step 1.1givenL2L3L4
LemmaStatement: AI-adaptedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Discreteness: σ(n)\sigma(n) is the immediate successor

Statement

For all m,nNm, n \in \mathbb{N}: m<n    σ(m)nm < n \iff \sigma(m) \le n (Order on the natural numbers). Consequently there is no kk with n<k<σ(n)n < k < \sigma(n): the successor σ(n)\sigma(n) is the immediate successor of nn.

Facts & Assumptions

Given: The order mn    j (m+j=n)m \le n \iff \exists j\ (m + j = n), with m<nm < n meaning mnm \le n and mnm \ne n (Order on the natural numbers), and addition with m+0=mm + 0 = m, m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers).

[L1]

Every nonzero natural is a successor: j0j=σ(i)j \ne 0 \Rightarrow j = \sigma(i) (Every nonzero natural number is a successor).

[L2]

σ(m)+i=σ(m+i)\sigma(m) + i = \sigma(m + i) (the left successor law, Left successor law for addition) and m+σ(i)=σ(m+i)m + \sigma(i) = \sigma(m + i) (the addition recursion).

[L3]

Cancellation: x+k=y+kx=yx + k = y + k \Rightarrow x = y (Addition is cancellative).

[L4]

σ(x)0\sigma(x) \ne 0 for every xx (Peano axiom P1) (The von Neumann naturals form a Peano system).

[L5]

Addition is commutative: x+y=y+xx + y = y + x (Addition is commutative).

Proof

technique · direct
1.1

Forward: if m<nm < n then mnm \le n and mnm \ne n, so m+j=nm + j = n for some jj, with j0j \ne 0 since j=0j = 0 would give m=nm = n; then j=σ(i)j = \sigma(i) [L1] and n=m+σ(i)=σ(m+i)=σ(m)+in = m + \sigma(i) = \sigma(m + i) = \sigma(m) + i [L2], hence σ(m)n\sigma(m) \le n.

givenL1L2
1.2

Backward: if σ(m)n\sigma(m) \le n then σ(m)+i=n\sigma(m) + i = n, so n=σ(m)+i=σ(m+i)=m+σ(i)n = \sigma(m) + i = \sigma(m + i) = m + \sigma(i) [L2], hence mnm \le n; and mnm \ne n, since m=nm = n would give m+σ(i)=m=m+0m + \sigma(i) = m = m + 0, so σ(i)+m=0+m\sigma(i) + m = 0 + m by [L5] and σ(i)=0\sigma(i) = 0 by [L3], contradicting [L4]; therefore m<nm < n.

givenL2L3L4L5
2.1

No kk satisfies n<k<σ(n)n < k < \sigma(n): if n<kn < k then σ(n)k\sigma(n) \le k by the equivalence, so kσ(n)k \ge \sigma(n), contradicting k<σ(n)k < \sigma(n); thus σ(n)\sigma(n) is the immediate successor.

step 1.1step 1.2
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Cancellation for multiplication by a nonzero factor

Statement

For all m,n,kNm,n,k\in\mathbb{N} with k0k\ne 0: if mk=nkm\cdot k=n\cdot k then m=nm=n.

Facts & Assumptions

Given: multiplication defined by m0=0m\cdot 0=0 and mσ(j)=mj+mm\cdot\sigma(j)=m\cdot j+m; the order mn    j (m+j=n)m\le n\iff\exists j\ (m+j=n), with m<nm<n meaning mnm\le n and mnm\ne n.

[L1]

Trichotomy: for all m,nNm,n\in\mathbb{N} exactly one of m<nm<n, m=nm=n, m>nm>n holds (Trichotomy of the order on N\mathbb{N}).

[L2]

Strict monotonicity: if k0k\ne 0 and m<nm<n then mk<nkm\cdot k<n\cdot k (Order is compatible with multiplication).

Proof

technique · contradiction
1.1

Suppose mk=nkm\cdot k=n\cdot k with k0k\ne 0 but mnm\ne n.

assume-contra
2.1

By trichotomy [L1], mnm\ne n forces m<nm<n or n<mn<m.

step 1.1L1
3.1

If m<nm<n then mk<nkm\cdot k<n\cdot k by [L2] (since k0k\ne 0); if n<mn<m then nk<mkn\cdot k<m\cdot k by [L2]; either way mknkm\cdot k\ne n\cdot k.

step 2.1L2
4.1

This contradicts mk=nkm\cdot k=n\cdot k, so the assumption mnm\ne n is impossible and m=nm=n.

step 1.1step 3.1discharge-contradiction
TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

The well-ordering principle

Statement

Every nonempty subset SNS \subseteq \mathbb{N} has a least element: there is S\ell \in S with s\ell \le s for all sSs \in S.

Facts & Assumptions

Given: A nonempty subset SNS \subseteq \mathbb{N}. Define T={nN:every m with mn satisfies mS}T = \{\, n \in \mathbb{N} : \text{every } m \text{ with } m \le n \text{ satisfies } m \notin S \,\} (informally, no element of SS is n\le n).

[L1]

Induction principle: a subset of N\mathbb{N} that contains 00 and is closed under σ\sigma equals N\mathbb{N} (The principle of mathematical induction).

[L2]

m<n    σ(m)nm < n \iff \sigma(m) \le n; consequently no kk satisfies n<k<σ(n)n < k < \sigma(n) (Discreteness: σ(n)\sigma(n) is the immediate successor).

[L3]

Exactly one of s<ns < n, s=ns = n, n<sn < s holds (Trichotomy of the order on N\mathbb{N}); and n<s    σ(n)sn < s \iff \sigma(n) \le s (Discreteness: σ(n)\sigma(n) is the immediate successor).

[L4]

σ\sigma is injective (Peano axiom P2) (The von Neumann naturals form a Peano system).

[L5]

σ(a)+i=σ(a+i)\sigma(a) + i = \sigma(a + i) for all a,ia, i (the left successor law, Left successor law for addition).

[L6]

0+k=k0 + k = k for all kk (Left identity for addition).

Proof

technique · contradiction
1.1

Suppose SS is nonempty but has no least element.

assume-contra
1.2

If m<σ(n)m < \sigma(n) then mnm \le n: by [L2] applied to mm and σ(n)\sigma(n) we get σ(m)σ(n)\sigma(m) \le \sigma(n), say σ(m)+i=σ(n)\sigma(m) + i = \sigma(n); then σ(m+i)=σ(m)+i=σ(n)\sigma(m + i) = \sigma(m) + i = \sigma(n) [L5], so m+i=nm + i = n by injectivity [L4], hence mnm \le n.

L2L4L5
2.1

First, 0T0 \in T: if 0S0 \in S then 0s0 \le s for all sSs \in S, since 0+s=s0 + s = s [L6], making 00 a least element, contrary to the assumption; hence 0S0 \notin S; moreover m0m \le 0 forces m=0m = 0, because m0m \ne 0 would give 0<m0 < m (as 0+m=m0 + m = m [L6] with m0m \ne 0) while m0m \le 0 gives m<0m < 0, and [L3] admits at most one of these; so every m0m \le 0 lies outside SS.

step 1.1L3L6given
2.2

Next, TT is closed under σ\sigma: assume nTn \in T; if σ(n)S\sigma(n) \in S then no sSs \in S satisfies s<σ(n)s < \sigma(n) (else sns \le n by step 1.2, so sSs \notin S because nTn \in T, a contradiction), hence σ(n)s\sigma(n) \le s for all sSs \in S by [L3], making σ(n)\sigma(n) a least element, contrary to the assumption; therefore σ(n)S\sigma(n) \notin S, and any mm with mσ(n)m \le \sigma(n) satisfies m=σ(n)m = \sigma(n) or mσ(n)m \ne \sigma(n); in the second case mσ(n)m \le \sigma(n) together with mσ(n)m \ne \sigma(n) gives m<σ(n)m < \sigma(n) by the definition of the strict order, hence mnm \le n by step 1.2 and mSm \notin S because nTn \in T, while in the first case m=σ(n)Sm = \sigma(n) \notin S; so every mσ(n)m \le \sigma(n) lies outside SS, giving σ(n)T\sigma(n) \in T.

step 1.1step 1.2L3
3.1

By the induction principle [L1], from 0T0 \in T and closure under σ\sigma, T=NT = \mathbb{N}; then for every nn, taking m=nnm = n \le n shows nSn \notin S, so S=S = \varnothing, contradicting nonemptiness; therefore SS has a least element.

step 2.1step 2.2L1discharge-contradiction
TheoremStatement: AI-adaptedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Strong (complete) induction

Statement

Let PP be a property of naturals such that for every nNn \in \mathbb{N}, if P(m)P(m) holds for all m<nm < n then P(n)P(n). Then P(n)P(n) holds for all nNn \in \mathbb{N}. (At n=0n = 0 the hypothesis is vacuous, so P(0)P(0) is forced.)

Facts & Assumptions

Given: A property PP satisfying the strong-induction hypothesis HH: for every nn, (P(m) for all m<n)P(n)\big(P(m) \text{ for all } m < n\big) \Rightarrow P(n).

[L1]

Well-ordering: every nonempty subset of N\mathbb{N} has a least element (The well-ordering principle).

[L2]

Exactly one of x<yx < y, x=yx = y, y<xy < x holds (Trichotomy of the order on N\mathbb{N}).

Proof

technique · contradiction
1.1

Suppose PP fails somewhere; let S={nN:P(n) is false}S = \{\, n \in \mathbb{N} : P(n) \text{ is false} \,\}, a nonempty subset of N\mathbb{N}.

assume-contra
2.1

By well-ordering [L1], SS has a least element \ell, so s\ell \le s for every sSs \in S; hence every m<m < \ell lies outside SS, since mSm \in S would give m\ell \le m, which [L2] does not allow alongside m<m < \ell; that is, P(m)P(m) holds for all m<m < \ell.

step 1.1L1L2
3.1

By the hypothesis HH applied at \ell, P()P(\ell) holds, so S\ell \notin S.

step 2.1given
4.1

But S\ell \in S by construction, a contradiction; hence SS is empty and P(n)P(n) holds for all nNn \in \mathbb{N}.

step 2.1step 3.1discharge-contradiction
TheoremStatement: Literature-sourcedProof: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Categoricity: the natural numbers are unique up to unique isomorphism

Statement

Let (N,0,σ)(N, 0, \sigma) and (N,0,σ)(N', 0', \sigma') be Peano systems (Peano system). There is a unique bijection h:NNh : N \to N' with h(0)=0h(0) = 0' and h(σ(n))=σ(h(n))h(\sigma(n)) = \sigma'(h(n)) for all nNn \in N, an isomorphism of Peano systems. Thus the natural numbers are determined up to a unique isomorphism.

Facts & Assumptions

Given: Two Peano systems (N,0,σ)(N, 0, \sigma) and (N,0,σ)(N', 0', \sigma').

[L1]

Recursion theorem, valid in any Peano system since its proof uses only the axioms P1 to P3: for a set AA, an element aAa \in A, and f:AAf : A \to A, there is a unique g:NAg : N \to A with g(0)=ag(0) = a and g(σ(n))=f(g(n))g(\sigma(n)) = f(g(n)) (The recursion theorem).

[L2]

Induction holds in each system (axiom P3): a subset containing the base point and closed under the successor is the whole system (Peano system).

Proof

technique · direct
1.1

Apply the recursion theorem [L1] in (N,0,σ)(N, 0, \sigma) with A=NA = N', a=0a = 0', f=σf = \sigma': there is a unique h:NNh : N \to N' with h(0)=0h(0) = 0' and h(σ(n))=σ(h(n))h(\sigma(n)) = \sigma'(h(n)).

L1
1.2

Symmetrically, apply [L1] in (N,0,σ)(N', 0', \sigma') with A=NA = N, a=0a = 0, f=σf = \sigma: there is a unique h:NNh' : N' \to N with h(0)=0h'(0') = 0 and h(σ(n))=σ(h(n))h'(\sigma'(n')) = \sigma(h'(n')).

L1
2.1

The composite hh:NNh' \circ h : N \to N satisfies (hh)(0)=h(0)=0(h' \circ h)(0) = h'(0') = 0 and (hh)(σ(n))=h(σ(h(n)))=σ(h(h(n)))=σ((hh)(n))(h' \circ h)(\sigma(n)) = h'(\sigma'(h(n))) = \sigma(h'(h(n))) = \sigma((h' \circ h)(n)); the identity idN\mathrm{id}_N satisfies the same two equations, so by induction [L2] the set {n:(hh)(n)=n}\{\, n : (h' \circ h)(n) = n \,\} is all of NN, i.e. hh=idNh' \circ h = \mathrm{id}_N.

step 1.1step 1.2L2
2.2

Likewise hh=idNh \circ h' = \mathrm{id}_{N'} by the same argument carried out in NN'.

step 1.1step 1.2L2
3.1

Hence hh is a bijection with inverse hh', and it preserves the base point and the successor, so it is an isomorphism of Peano systems.

step 2.1step 2.2
4.1

Uniqueness: any h1:NNh_1 : N \to N' with h1(0)=0h_1(0) = 0' and h1(σ(n))=σ(h1(n))h_1(\sigma(n)) = \sigma'(h_1(n)) satisfies the recursion clauses defining hh in step 1.1, so h1=hh_1 = h by the uniqueness in [L1] (equivalently {n:h1(n)=h(n)}\{\, n : h_1(n) = h(n) \,\} contains 00 and is closed under σ\sigma, hence is all of NN by induction [L2]).

step 1.1L1L2

5 · Examples, counterexamples and false statements

ExampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

1+1=21+1=2 from the recursion

Example

With 1=σ(0)1 = \sigma(0) and 2=σ(1)=σ(σ(0))2 = \sigma(1) = \sigma(\sigma(0)), the recursive definition of addition gives 1+1=21 + 1 = 2. This is the archetypal proof by unfolding a recursive definition.

Facts & Assumptions

Given: addition m+0=mm + 0 = m, m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers); numerals 1=σ(0)1 = \sigma(0), 2=σ(1)2 = \sigma(1) (The natural numbers N\mathbb{N} (von Neumann)).

Verification

technique · direct
1.1

By definition 1=σ(0)1 = \sigma(0) and 2=σ(1)=σ(σ(0))2 = \sigma(1) = \sigma(\sigma(0)).

given
1.2

1+0=11 + 0 = 1, by m+0=mm + 0 = m with m=1m = 1.

given
2.1

1+1=1+σ(0)=σ(1+0)1 + 1 = 1 + \sigma(0) = \sigma(1 + 0), by m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) with m=1m = 1, n=0n = 0.

step 1.1given
3.1

Hence 1+1=σ(1+0)=σ(1)=21 + 1 = \sigma(1 + 0) = \sigma(1) = 2.

step 2.1step 1.2
ExampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passjudge pass (openai/gpt-5.4)audited 2026-07-25Open item page →

2+2=42+2=4 from the recursion

Example

With 2=σ(σ(0))2 = \sigma(\sigma(0)) and 4=σ(σ(σ(σ(0))))4 = \sigma(\sigma(\sigma(\sigma(0)))), the recursion gives 2+2=42 + 2 = 4, by unfolding the successor twice down to 2+0=22 + 0 = 2.

Facts & Assumptions

Given: addition m+0=mm + 0 = m, m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) (Addition of natural numbers); 2=σ(σ(0))2 = \sigma(\sigma(0)), 4=σ(σ(σ(σ(0))))4 = \sigma(\sigma(\sigma(\sigma(0)))) (The natural numbers N\mathbb{N} (von Neumann)).

Verification

technique · direct
1.1

Write 2=σ(σ(0))2 = \sigma(\sigma(0)); then 2+2=2+σ(σ(0))2 + 2 = 2 + \sigma(\sigma(0)).

given
1.2

2+0=22 + 0 = 2, by m+0=mm + 0 = m.

given
2.1

2+σ(σ(0))=σ(2+σ(0))=σ(σ(2+0))2 + \sigma(\sigma(0)) = \sigma(2 + \sigma(0)) = \sigma(\sigma(2 + 0)), applying m+σ(n)=σ(m+n)m + \sigma(n) = \sigma(m + n) twice.

step 1.1given
3.1

Hence 2+2=σ(σ(2))=σ(σ(σ(σ(0))))=42 + 2 = \sigma(\sigma(2)) = \sigma(\sigma(\sigma(\sigma(0)))) = 4.

step 2.1step 1.2
CounterexampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

The induction axiom is independent of the other Peano axioms

Statement refuted

Refuted claim: any triple (N,0,σ)(N, 0, \sigma) satisfying (P1) 00 is not a successor and (P2) σ\sigma is injective automatically satisfies (P3) induction. The witness is N=NZN = \mathbb{N} \sqcup \mathbb{Z} (a disjoint union of a copy of the naturals and a copy of the integers), with base point 00 the zero of the N\mathbb{N}-copy and σ\sigma the usual successor nn+1n \mapsto n + 1 on each copy separately.

Facts & Assumptions

Given: N=NZN = \mathbb{N} \sqcup \mathbb{Z} with σ\sigma acting as successor within each copy; 0=0 = the N\mathbb{N}-copy's zero. Write NN,NZN_{\mathbb{N}}, N_{\mathbb{Z}} for the two copies.

[L1]

The Peano axioms P1, P2, P3 (Peano system).

[L2]

N\mathbb{N} itself is the standard model (The natural numbers N\mathbb{N} (von Neumann)).

Counterexample

technique · direct
1.1

P1 holds: σ\sigma is the successor within each copy; 00 is the N\mathbb{N}-copy zero, which is not the successor of any element (nothing in NNN_{\mathbb{N}} maps to it, and σ\sigma maps NZN_{\mathbb{Z}} into NZN_{\mathbb{Z}}), so σ(x)0\sigma(x) \neq 0 for all xx.

givenL1
1.2

P2 holds: σ\sigma is injective on NNN_{\mathbb{N}} and on NZN_{\mathbb{Z}} separately, and σ\sigma maps each copy into itself, so σ\sigma is injective on NN.

givenL1
1.3

P3 fails: let S=NNS = N_{\mathbb{N}}, the N\mathbb{N}-copy; then 0S0 \in S and σ\sigma maps SS into SS (σ(NN)NN\sigma(N_{\mathbb{N}}) \subseteq N_{\mathbb{N}}), so SS contains 00 and is closed under σ\sigma; the N\mathbb{N}-copy is the standard model [L2].

givenL1L2
2.1

But S=NNNS = N_{\mathbb{N}} \neq N, since the Z\mathbb{Z}-copy NZN_{\mathbb{Z}} is disjoint from SS and nonempty.

step 1.3
3.1

Thus (N,0,σ)(N, 0, \sigma) satisfies P1 and P2 but not P3, refuting the claim: induction is independent of P1 and P2 and cannot be dropped.

step 1.1step 1.2step 1.3step 2.1
CounterexampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

Injectivity of the successor is independent of the other Peano axioms

Statement refuted

Refuted claim: (P2) injectivity of σ\sigma follows from (P1) 00 is not a successor together with (P3) induction. The witness is the three-element system N={0,1,2}N = \{0, 1, 2\} with σ(0)=1\sigma(0) = 1, σ(1)=2\sigma(1) = 2, σ(2)=2\sigma(2) = 2.

Facts & Assumptions

Given: N={0,1,2}N = \{0, 1, 2\}, base point 00, σ(0)=1\sigma(0) = 1, σ(1)=2\sigma(1) = 2, σ(2)=2\sigma(2) = 2.

[L1]

The Peano axioms (Peano system).

Counterexample

technique · direct
1.1

P1 holds: the values of σ\sigma are σ(0)=1\sigma(0) = 1, σ(1)=2\sigma(1) = 2, σ(2)=2\sigma(2) = 2, none of which is 00, so σ(x)0\sigma(x) \neq 0 for all xx.

givenL1
1.2

P3 holds: if SNS \subseteq N with 0S0 \in S and SS closed under σ\sigma, then 0S0 \in S forces 1=σ(0)S1 = \sigma(0) \in S, then 2=σ(1)S2 = \sigma(1) \in S, so S={0,1,2}=NS = \{0, 1, 2\} = N; thus induction holds.

givenL1
1.3

P2 fails: σ(1)=2=σ(2)\sigma(1) = 2 = \sigma(2) but 121 \neq 2, so σ\sigma is not injective.

givenL1
2.1

So (N,0,σ)(N, 0, \sigma) satisfies P1 and P3 but not P2, refuting the claim; note the failure permits the pathology σ(2)=2\sigma(2) = 2 (a number that is its own successor) and a finite model in which distinct numerals collapse.

step 1.1step 1.2step 1.3
False statementConstruction: AI-adaptedVerification: AI-adaptedprecheck passverified 2026-08-02 (claude-opus-5)Open item page →

FALSE: successor-closure alone forces a set to be all of N\mathbb{N}

Statement

False statement. If a set SNS \subseteq \mathbb{N} is nonempty and closed under the successor (nSσ(n)Sn \in S \Rightarrow \sigma(n) \in S), then S=NS = \mathbb{N}. (That is, the induction principle would hold without its base case 0S0 \in S.)

Facts & Assumptions

Given: the claim above.

[L1]

The induction principle requires 0S0 \in S (The principle of mathematical induction).

[L2]

σ(n)0\sigma(n) \neq 0 for all nn (P1) (The von Neumann naturals form a Peano system).

Refutation

technique · direct
1.1

Take S=N{0}={nN:n0}S = \mathbb{N} \setminus \{0\} = \{n \in \mathbb{N} : n \neq 0\}, the set of nonzero naturals; it is nonempty (for instance 1=σ(0)S1 = \sigma(0) \in S).

given
1.2

SS is closed under σ\sigma: for any nn, σ(n)0\sigma(n) \neq 0 by P1 [L2], so σ(n)S\sigma(n) \in S.

L2
2.1

But 0S0 \notin S, so SNS \neq \mathbb{N}; the nonempty successor-closed set SS is not all of N\mathbb{N}, refuting the claim.

step 1.1step 1.2
3.1

The base case 0S0 \in S is therefore indispensable in the induction principle [L1]; successor-closure and nonemptiness do not suffice.

step 2.1L1

Sources