In ZFC, Dedekind finite set and finite set are same things. So I have a set say A(which is equal to N in ZFC) all Dedekind finite set are equivalent to proper subsets of A and A is well ordered set.

But I just get to know, and there is a model of ZF theory in which there exist an infinite but Dedekind finite set. So I have one question. Can we find a set A such that all Dedekind finite sets are equivalent to its proper subset and A is well ordered.(Of course N will not work)


No. First we note that if $A$ is Dedekind-finite then all its subsets are Dedekind-finite. If $A$ is infinite, then for every $n$ there is some $A_n$ such that $|A|=|A_n|+n$. Therefore the cardinals $\{|A_n|\mid n<\omega\}$ make an infinite decreasing sequence of cardinals below that of $A$.

You might want to ask if the cardinals are even linearly ordered, because they don't have to be. But it is also consistent. If $A$ is an infinite set which cannot be partitioned into two infinite sets we call $A$ amorphous. It is not hard to see that the cardinals below $|A|$ are exactly $\omega\cup\{|A_n|\mid n<\omega\}$ which is a linear order.

| cite | improve this answer | |
  • $\begingroup$ Hi Asaf. Can we actually pick a sequence $(A_n)_n$ with sizes as stated? (Naturally, we can pick the sequence $(|A_n|)_n$.) $\endgroup$ – Andrés E. Caicedo Sep 25 '14 at 18:11
  • $\begingroup$ Andres, no, since $2^A$ might be Dedekind-finite. For example if $A$ is amorphous. $\endgroup$ – Asaf Karagila Sep 25 '14 at 18:36
  • $\begingroup$ Yes, but the $A_n$ don't need to be subsets of $A$, so why does this matter? $\endgroup$ – Andrés E. Caicedo Sep 25 '14 at 20:14
  • $\begingroup$ Right, we are interested in arbitrary sets. I'll think about it some more. $\endgroup$ – Asaf Karagila Sep 25 '14 at 20:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.