Suppose that $A$ and $B$ are two infinite sets and $|A|<|B|$. The question is that how to prove that $|A∪B|=|B|$. The proof is related to the Axiom of Choice.

  • $\begingroup$ It is easy to show that $|B|<=|A \cup B|$. I tryed to find a injection from B to $|A \cup B|$. Then with the using of Cantor-Bernstein Theorm, it is proved. However, thus map is not fined yet. $\endgroup$ – Yijun Yuan Oct 18 '14 at 3:40
  • $\begingroup$ The injection is $B \to B \subset A \cup B$ The challenge is to have an injection from $A \cup B$ into $B$. $\endgroup$ – Ross Millikan Oct 18 '14 at 4:35

$|A\cup B| \leq |A| + |B| \leq |B| + |B| = |B|$ since $|B|$ is of infinite cardinality and $2\cdot \aleph_k = \aleph _k$ for whichever cardinal $|B|$ happens to be.

Then also $|B|\leq |A\cup B|$ by subadditivity.

Hence $|A\cup B| = |B|$ when $B$ is an infinite set and $|A|\leq |B|$

I don't think you need to use the axiom of choice for the proof, though perhaps I have naive notions on how set operations work on higher sized sets.

  • 2
    $\begingroup$ $|B| + |B| = |B|$ uses Zorn, so it is using the AC. $\endgroup$ – Aram Oct 18 '14 at 3:52
  • $\begingroup$ Is there a simple way to explain $|B|+|B|=|B|$ without the using of the Zorn( but with AC)? $\endgroup$ – Yijun Yuan Oct 18 '14 at 5:15
  • $\begingroup$ @袁轶君 Imho, using Zorn for this is very simple. You can find a proof in Kaplansky's book about metric spaces. $\endgroup$ – Aram Oct 18 '14 at 19:02
  • $\begingroup$ @Aram Thanks a lot. I've got the answer. :-) $\endgroup$ – Yijun Yuan Oct 26 '14 at 9:34

By the axiom of choice all infinite cardinals are alephs, so you can use $n+\aleph_0=\aleph_0$ for finite $|A|=n$, and $2\cdot\aleph_\alpha=\aleph_\alpha$ for infinite $|A|=\aleph_\alpha$.

In the absence of the axiom of choice, there could be an infinite cardinal $m$ such that $m\lt2m$. From $m\lt2m$ it follows (this is not trivial) that $2m\lt3m$. Then, if $A,B$ are disjoint sets with $|A|=m$ and $|B|=2m$, we have $|A|\lt|B|$ while $|A\cup B|=m+2m=3m\gt2m=|B|$.

  • $\begingroup$ "all infinite cardinals are alephs" is based on CH, not AC. If the CH is refused, there is a set A, which $\aleph _{0} <|A|<\aleph _{1}$. $\endgroup$ – Yijun Yuan Oct 18 '14 at 5:18
  • $\begingroup$ @袁轶君 Wrong. If the CH is false, there is a set $A$ such that $\aleph_0\lt|A|\lt|\mathbb R|$. But there is never a cardinal between $\aleph_0$ and $\aleph_1$. $\endgroup$ – bof Oct 18 '14 at 5:34
  • $\begingroup$ Well...It's my fault. $\endgroup$ – Yijun Yuan Oct 18 '14 at 5:48

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.