Processing math: 100%
Table of Contents

Cardinal numbers

Consider the set of natural numbers N={0,1,2,3,}. There is a countably 0

Claim: 01

Proof: We

Consider some set X and its power set 2X. Let the cardinality of |X|=0

Define f to be some set function, which associates a subset of X to each element xX. xf(x). To show 01, we have to show that there is no bijection between the set and its powerset (since 1 is identified as the cardinality of the powerset of a set of cardinality 0).

We give a proof by contradiction. Assume that a bijection f exists, and consider the subset of X which consists of all elements which are not contained in their image under f. Y={xX | xf(x)},(1) which means that: yY,yf(y).(2)

Now if f is a bijection, then there must be some ˜yX that is the pre-image of Y, such that: f(˜y)=Y.

Now we ask the question: “is ˜yY? ”
If yes, then we get a contradiction with (2). If no, then this means ˜ynf(˜y) and from the definition of the set Y in (1), ˜y should have been included in Y. So we have a contradiction, and f must not exists.

On the other hand, we have the trivial injection g which assigns to each element of a set the set with the single element: g(x)={x},xX.

Therefore we have shown that there is an injection from X to its powerset, but there is no bijection, hence the cardinality of the set X must differ from the cardinality of its powerset. By definition, we call this next cardinality 1.

Sidenote

There is a bijection between the reals (0,1) and subsets of the Naturals. Let r=0.r1r2r3r4 be some real numbers written in binary, where rj{0,1}. Now consider the set SrN, which corresponds to the number r. We define S to consist of all numbers j such that rj=1.

Note that there is also a bijection between the interval (0,1)R and the whole R.

Acknowledgments

I want to thank Paul Raymond-Robichaud for showing me these results, concisely and cleanly.