2
$\begingroup$

I am trying to understand the proof of Corollary 18.34 in Jech's Set Theory:

If $0^\sharp$ does not exist, then if $\kappa$ is a singular cardinal and if there exists a nonconstructible subset of $\kappa$, then some $\alpha < \kappa$ has a nonconstructible subset.

In the course of the proof, he assumes that every $\alpha < \kappa$ has only constructible subsets. He then wants to show that it suffices to show that all subsets of $\kappa$ of cardinality $cf(\kappa)$ are constructible. In order show this, suppose $A \subseteq \kappa$ and $(a_\nu : \nu < cf(\kappa))$ is a cofinal sequence. By the assumption $A \cap a_\nu$ is constructible. Jech then claims that $\mathcal{A} = \{A \cap a_\nu : \nu < cf(\kappa)\}$ is a constructible set.

Why is this set constructible? His reasoning is : "$\mathcal{A}$ is a subset of $L_\kappa$ of size $\leq cf(\kappa)$ and hence constructible". I do not see why the constructibility of $\mathcal{A}$ follows from this.

$\endgroup$
2
$\begingroup$

He goes to explain that in the next paragraph.

First note that we can encode this set as a set of ordinals of size $\leq\operatorname{cf}(\kappa)$, such that the encoding and decoding functions are in $L$. Therefore by the covering lemma (which holds in the absence of $0^\#$) there is a set of ordinals covering this set, whose cardinality is still $<\kappa$ (it is $\operatorname{cf}(\kappa)+\aleph_1$), $Y$.

In $L$ we can find a bijection between $Y$ and some $\alpha<\kappa$ (e.g. the Mostowski collapse), and by the assumption all the subsets of $\alpha$ which are in $V$ are in fact in $L$. In particular the subset encoding the original set of ordinals; therefore its decoding - the sequence of initial segments of $A$ - is also in $L$, as wanted.

$\endgroup$

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.