TheoremBase

Proof

We prove claim 1. Claim 2 follows by the same argument with the addition of KK replaced by its multiplication, using claim 3 of Extraction of a Term from a Finite Sum or Product in a Field in place of claim 2 of that lemma, and claim 1 of Properties of Finite Products in place of claim 1 of Properties of Finite Sums.

We argue by induction on nn, using the induction principle for the natural numbers with successor map SS. Let P\mathcal{P} be the set of natural numbers nn such that the identity of claim 1 holds for every map a:[n]→Ka:[n]\to K and every bijection σ:[n]→[n]\sigma:[n]\to[n].

Base. [1][1] consists of 11 alone, so σ(1)=1\sigma(1)=1 and both sides equal a1a_{1} by claim 1 of Properties of Finite Sums. Hence 1∈P1\in\mathcal{P}.

Step. Suppose n∈Pn\in\mathcal{P}, and let a:[S(n)]→Ka:[S(n)]\to K and a bijection σ:[S(n)]→[S(n)]\sigma:[S(n)]\to[S(n)] be given. Since σ\sigma is a bijection there is exactly one j∈[S(n)]j\in[S(n)] with σ(j)=S(n)\sigma(j)=S(n).

Let b:[S(n)]→Kb:[S(n)]\to K be given by bk=aσ(k)b_{k}=a_{\sigma(k)}, and let gj:[n]→[S(n)]g_{j}:[n]\to[S(n)] be the gap map of Extraction of a Term from a Finite Sum or Product in a Field. Claim 2 of that lemma, applied to bb at the index jj, gives

∑k=1S(n)aσ(k)=(∑k=1nbgj(k))+bj=(∑k=1naσ(gj(k)))+aS(n).(i)\sum_{k=1}^{S(n)}a_{\sigma(k)}=\Bigl(\sum_{k=1}^{n}b_{g_{j}(k)}\Bigr)+b_{j}=\Bigl(\sum_{k=1}^{n}a_{\sigma(g_{j}(k))}\Bigr)+a_{S(n)} . \tag{i}

Let EE be the set of those l∈[S(n)]l\in[S(n)] with l≠jl\ne j, and let DD be the set of those m∈[S(n)]m\in[S(n)] with m≠S(n)m\ne S(n). By claim 1 of Extraction of a Term from a Finite Sum or Product in a Field, gjg_{j} is a bijection from [n][n] onto EE. By claim 3 of Injectivity, Composition, and Restriction of Bijections the restriction of σ\sigma to EE is a bijection from EE onto its image σ(E)\sigma(E); and σ(E)=D\sigma(E)=D, because σ\sigma is injective by claim 1 of Injectivity, Composition, and Restriction of Bijections with σ(j)=S(n)\sigma(j)=S(n), so no element of EE is sent to S(n)S(n), while every m∈Dm\in D is σ(l)\sigma(l) for some l∈[S(n)]l\in[S(n)] by surjectivity, and that ll lies in EE since σ(j)=S(n)≠m\sigma(j)=S(n)\ne m. Moreover D=[n]D=[n]: if m∈[S(n)]m\in[S(n)] and m≠S(n)m\ne S(n) then m<S(n)m<S(n) and hence m≤nm\le n, by the order facts of Properties of the Order on the Natural Numbers; conversely m≤n<S(n)m\le n<S(n) for m∈[n]m\in[n].

Therefore the map τ:[n]→[n]\tau:[n]\to[n] given by τ(k)=σ(gj(k))\tau(k)=\sigma(g_{j}(k)) is a composite of two bijections, hence a bijection by claim 2 of Injectivity, Composition, and Restriction of Bijections.

Let a′a' be the restriction of aa to [n][n]. Since τ(k)∈[n]\tau(k)\in[n] we have aσ(gj(k))=aτ(k)′a_{\sigma(g_{j}(k))}=a'_{\tau(k)}, so the induction hypothesis applied to a′a' and τ\tau gives

∑k=1naσ(gj(k))=∑k=1naτ(k)′=∑k=1nak′.\sum_{k=1}^{n}a_{\sigma(g_{j}(k))}=\sum_{k=1}^{n}a'_{\tau(k)}=\sum_{k=1}^{n}a'_{k}.

Finally, by claim 1 of Properties of Finite Sums, in its restriction and recursion parts,

∑k=1S(n)ak=(∑k=1nak′)+aS(n).\sum_{k=1}^{S(n)}a_{k}=\Bigl(\sum_{k=1}^{n}a'_{k}\Bigr)+a_{S(n)} .

Combining this with the previous display and with (i) gives ∑k=1S(n)aσ(k)=∑k=1S(n)ak\sum_{k=1}^{S(n)}a_{\sigma(k)}=\sum_{k=1}^{S(n)}a_{k}. Hence S(n)∈PS(n)\in\mathcal{P}, and by induction P\mathcal{P} contains every natural number.

Citations

Loading…

Dependencies

Uses0

Loading…

Comments

Log in to comment.

Loading…