I recently edited an answer of mine on math.SE which discussed the implication of the two assertions:

  • $AH(0)$ which is $2^{\aleph_0}=\aleph_1$, and
  • $CH$ which says that if $A\subseteq 2^{\omega}$ and $\aleph_0<|A|$ then $|A|=2^{\aleph_0}$.

We know they are indeed equivalent under the axiom of choice (and actually much less). It is also trivial to see that $AH(0)\Rightarrow CH$. However the converse is not true, indeed in Solovay's model (or in models of AD) there are no $\aleph_1$ many reals, but $CH$ holds since every uncountable set of reals has a perfect subset.

While revising my answer I tried to find a reference whether or not in the Feferman-Levy model, in which the continuum is a countable union of countable sets, satisfies the continuum hypothesis (we already know that it does not satisfy $AH(0)$).

To my surprise the answer is negative. There exists a set whose cardinality is strictly between the continuum and $\omega$, the construction is described in A. Miller's paper [1] in which he remarks that in the Feferman-Levy the constructed set cannot be put in bijection with the continuum.

I was wondering whether or not this is always true in models in which the continuum is a countable union of countable sets, or is this just one of the peculiarities of the Feferman-Levy model.


  1. Let $V$ be a model of $ZF$ in which $2^{\omega}$ can be written as the countable union of countable sets. Does $CH$ fail in $V$?

  2. Suppose that $V$ is a model of $\omega_1\nleq2^\omega$ and $CH$, does this imply that $\omega_1$ is regular (which means inaccessible in $L$)?


  1. Miller, A. A Dedekind Finite Borel Set. Arch. Math. Logic 50 (2011), no. 1-2, 1--17.
  • $\begingroup$ I have added a second question which seems to be equivalent of the first. It may be an easier formulation... $\endgroup$
    – Asaf Karagila
    Aug 1, 2012 at 23:01
  • $\begingroup$ The recent trend of completely arbitrary downvotes seems really strange. $\endgroup$
    – Asaf Karagila
    Jul 9, 2013 at 4:54
  • $\begingroup$ So "Borel sets are analytic" is not a theorem of ZF? Interesting. $\endgroup$ Dec 26, 2015 at 23:38
  • $\begingroup$ Paul, depends on how you define "Borel" and "analytic". $\endgroup$
    – Asaf Karagila
    Dec 27, 2015 at 5:38
  • 2
    $\begingroup$ That's interesting too. $\endgroup$ Dec 27, 2015 at 10:11

1 Answer 1


The answer for the second question is no. Truss proved in [1] that if we repeat Solovay's construction from a limit cardinal $\kappa$, we obtain a model in which the following properties:

  1. Countable unions of countable sets of real numbers are countable;
  2. Every well-orderable set of real numbers is countable;
  3. Every uncountable set of reals has a perfect subset;
  4. DC holds iff $\omega_1$ is regular iff $\kappa$ is inaccessible in the ground model;
  5. Every set of real numbers is Borel.

This shows that it is possible to have $CH+\aleph_1\nleq2^{\aleph_0}+\operatorname{cf}(\omega_1)=\omega$. However it does not answer the original (first) question.


  1. Truss, John, Models of set theory containing many perfect sets. Ann. Math. Logic 7 (1974), 197–219.
  • $\begingroup$ Interesting. Does Truss use an inaccessible? Item 2 implies that $\omega_1$ is inaccessible to reals, so I suppose he must use one somewhere. Or am I missing something? $\endgroup$ Apr 8, 2016 at 13:35
  • $\begingroup$ Yes, it is inaccessible go reals, which only means it is a limit cardinal in L. But in Truss' models he can have $\omega_1$ singular also. $\endgroup$
    – Asaf Karagila
    Apr 8, 2016 at 14:14
  • $\begingroup$ Ah! That makes sense and 4 is even more interesting now. $\endgroup$ Apr 8, 2016 at 16:54
  • $\begingroup$ I find (1) and (4) in conjunction to be the interesting part. :) $\endgroup$
    – Asaf Karagila
    Apr 8, 2016 at 17:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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