Statement(True/False) : Without Axiom of of choice there doesn't exists any set $B$ such that both $B$ and $B^c$ intersect every closed uncountable subsets Or every perfect sets in $\Bbb{R}$.

Any set $B\subset \Bbb{R}$ which intersect every closed uncountable sets is known as Bernstein set.

If $B$ is Bernstein set then $B$ complement is also Bernstein . Also any subset of Bernstein that that are measurable must be meager, hence Bernstein set is not measurable infact has full outer measure.

Since without Axiom of choice it is impossible to construct a non measurable set (here) and Bernstein set are non measurable , without Axiom of Choice existence of Bernstein set is impossible.

My Question: Is my argument correct?

Is there any alternative way to prove non existence of Bernstein set without Axiom of Choice?

  • $\begingroup$ I believe your argument that AC is needed to construct a Bernstein set is correct: The Bernstein set is not Lebesgue measurable and we need Axiom of Choice (or something similar) to assert existence of a set that is not Lebesgue measurable. $\endgroup$
    – Michael
    Jul 6, 2022 at 18:56

1 Answer 1


There is a subtle distinction. If I say that you don't have a cat, does that mean that you have a dog?

No. It doesn't. Without the axiom of choice means just not assuming it, but it means nothing about whether or not we assume its negation. It also doesn't tell you how it fails.

Bernstein sets can exist with the axiom of choice being false, simply because choice might fail "elsewhere", while the real numbers can still be well ordered in the universe.

Indeed, Bernstein sets can exist even if the real numbers cannot be well ordered! That much is consistent.

What you really want to say, though, is that you cannot prove the existence of Bernstein sets in ZF alone. And indeed, your argument is fine. Bernstein sets are non measurable, they are uncountable and without a perfect subset, etc. And indeed it is consistent that all sets are Lebesgue measurable, or have perfect subsets, etc.

  • $\begingroup$ Where can I read about the consistency of a Bernstein set without a well-ordering of the reals? $\endgroup$ Jul 7, 2022 at 6:30
  • $\begingroup$ @AlessandroCodenotti: It's the in the recent works of Ralf Schindler and his coauthors. For example, arxiv.org/abs/1809.10420 $\endgroup$
    – Asaf Karagila
    Jul 7, 2022 at 6:38
  • $\begingroup$ Nice, thanks! I'll ask Ralf about this next time I meet him around $\endgroup$ Jul 7, 2022 at 7:09

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .