Field axioms are numbered as in Field and order axioms as in Ordered Field; the order ≤ on R is a total order, hence reflexive, transitive and antisymmetric. Write (Ok) for claim k of Properties of the Order on the Natural Numbers and (Sk) for claim k of Properties of Finite Sums.
Step 0.
(a) 0≤1. By field axiom 6, 1=0. Let t=∣1∣ be the absolute value of 1. By claim 1 of Properties of the Absolute Value in an Ordered Field we have 0≤t, and t=0, since that claim gives t=0 only for 1=0. By claim 4 of the same lemma with x=y=1,
t=∣1⋅1∣=∣1∣∣1∣=tt.
Multiplying by the inverse t−1 supplied by field axiom 7 and using field axioms 5 and 6,
1=tt−1=(tt)t−1=t(tt−1)=t⋅1=t,
so 0≤t reads 0≤1.
(b) For every x∈R one has x≤x+1, and x+1≤x is false. Order axiom 1 with c=x turns 0≤1 into 0+x≤1+x, that is, x≤x+1. If also x+1≤x, then order axiom 1 with c=−x gives (x+1)+(−x)≤x+(−x)=0, whose left-hand side is 1+(x+(−x))=1; so 1≤0, and antisymmetry with 0≤1 gives 1=0, contradicting field axiom 6.
Step 1: claim 1. By (S1), σ(1)=11(1)=1. Fix n∈N. By (O4), (O5) and (O1) we have 1≤n and n≤S(n), so n∈[S(n)] and the restriction of 1(S(n)) to [n] is 1(n). Hence (S1), used first in its recursion part and then in its restriction part, gives
σ(S(n))=(k=1∑n1k(S(n)))+1S(n)(S(n))=σ(n)+1.
Step 2: claim 2. Note first that p<q makes q≤p false: p<q gives p≤q by (O1), so q≤p would give p=q by (O2), contradicting (O3).
Fix m∈N and let A be the set of those n∈N such that either n≤m, or else both m<n and σ(m)+1≤σ(n). By (O4) we have 1≤m, so 1∈A.
Let n∈A. If S(n)≤m, then S(n)∈A. Otherwise m<S(n), because by (O3) the alternatives S(n)<m and S(n)=m would each give S(n)≤m by (O1). There are two cases.
Case n≤m. From m<S(n) we get m≤S(n) by (O1) and m=S(n) by (O3), so (O5) gives m≤n and hence n=m by (O2). Step 1 then gives σ(S(n))=σ(n)+1=σ(m)+1.
Case m<n and σ(m)+1≤σ(n). Step 1 and step 0(b) give σ(n)≤σ(n)+1=σ(S(n)), so transitivity gives σ(m)+1≤σ(S(n)).
In both cases m<S(n) and σ(m)+1≤σ(S(n)), so S(n)∈A. By Principle of Induction for the Natural Numbers, A=N.
Now let m<n. Then n≤m is false, so σ(m)+1≤σ(n). Step 0(b) gives σ(m)≤σ(m)+1, hence σ(m)≤σ(n) by transitivity; and σ(m)=σ(n) would give σ(m)+1≤σ(m), which step 0(b) excludes.
Step 3: claim 3. Let σ(m)=σ(n). By (O3) exactly one of m<n, m=n, n<m holds, and claim 2 excludes the first and the third. Hence m=n.