well-ordering principle
Well-ordering principle
The well-ordering principle states that every non-empty subset of natural numbers (including zero) contains an element which is less than or equal to any other element in the subset.
Exclusion of zero
Due to the fact that mathematicians still can't decide whether zero is a natural number, some sources will exclude zero from the definition. Regardless, the principle satisfies natural numbers with or without zero.
Proof by completeness axiom
Let some set \[S\] be a non-empty subset of natural numbers, i.e. \[S\subseteq \mathbb{N}\] and \[S\ne\emptyset\]. Since natural numbers itself is a subset of real numbers, \[S\subseteq \mathbb{R}\]. Additionally, for all \[n\in \mathbb{N}\], by definition of natural numbers \[n\ge0\], consequently, for all \[n\in S\], \[n\ge 0\]. Thus, 0 is a lower bound of \[S\].
By the corollary of the completeness axiom, \[S\] has an infimum as it has a lower bound. Hence, let \[b=\inf S\]. By definition, since \[b\] is the greatest lower bound for \[S\], it follows that \[b+1\] can't be a lower bound of \[S\]. So, for some \[n\in S\], \[n<b+1\]. We'll use proof by contradiction to show that \[n\] is the smallest element of \[S\].
Assume \[n\] is not the smallest element of \[S\], then there must exist some \[m\in S\] such that \[b\le m<n<b+1\]. It is obvious that there exist no combination of natural numbers that satisfies the inequality. Mathematically, by rearranging the inequality, \[m<n\implies0<n-m\] (centre inequality); \[b+1\le m+1\implies n<b+1\le m+1\] (left inequality), which follows that \[n<m+1\implies n-m<1\]. We find that \[0<n-m<1\]. However, there exist no natural number between zero and 1, therefore, \[n=b\] must be the smallest element of \[S\].
To make sense of this proof, write down a simple set of natural numbers, e.g. \[S=\left\{ 2,2,3,4 \right\}\], then substitute the numbers while following the proof above.
Proof by induction axiom
The following is a proof by contradiction.
Let \[S\] be a non-empty subset of \[\mathbb{N}\] with no least element. Note that \[0\notin S\] since 0 is the smallest number, and choosing 0 as the smallest number is just purely convention. Though, the following proof still works regardless of choosing 0 or 1 as the smallest number (for those arguing that 0 is not a natural number).
Let \[J\] be the set of all natural numbers that are not in \[S\]. Since 0 is the lower bound of \[\mathbb{N}\] (as it is the smallest natural number) and \[S\] can not have a least element, \[0\notin S\implies 0\in J\]. We will now prove by induction \[0,1,2,\dots,n\in J\] implies that \[n+1\in J\]. Suppose \[0,1,2,\dots,n\in J\]. Then, \[n+1\] can not be in \[S\], as \[n+1\] would be the lowest bound of \[S\] because \[0,1,2,\dots,n\notin S\], which would make \[n+1\] the least element of \[S\]. Thus, since our premise is that \[S\] can not have a least element, \[n+1\in J\].
By induction, we conclude that \[J=\mathbb{N}\], which implies that \[S\] must be an empty set, which contradictions our initial assumption. Therefore, \[S\] must have a least element.