7
$\begingroup$

I want to see if the following result is correct or not:

Let $\varphi(x)$ be a formula in $\mathcal{L}_{\mathrm{ZF}}$ with only bounded quantifiers such that $\exists x\,\varphi(x)$ holds in $V$, then $\exists x\,\varphi(x)$ holds in $H_{\omega_1}$.

I wrote a proof for it using the following reasoning:

$\varphi$ is absolute for all transitive models; let $\exists x\,\varphi(x)$ be true, then $\varphi(a)$ holds for some $a$. Let $\vartheta\in\mathbf{ON}$ such that $a\in V_\vartheta$. Let $M\prec V_\vartheta$ be a countable elementary substructure, $\bar M$ be the Mostowski collapse ($\bar M\cong M$), then $\bar M$ is transitive and countable, hence hereditarily countable, so $\bar M\subseteq H_{\omega_1}$, so there is a witness for $\varphi(x)$ in $H_{\omega_1}$.

My questions are: (1) is the result above correct? (2) Is the proof given correct? (3) If the result is correct and the proof is not correct, what would be a correct proof for the result?

$\endgroup$
  • 2
    $\begingroup$ The result and your proof are correct. But your result can be strengthened to let $\phi$ contain parameters from $H_{\omega_1}$ $\endgroup$ – GME Mar 9 '16 at 21:08
  • $\begingroup$ @GME Thank you, and if $\varphi$ contains parameters from $H_\kappa$ for $\kappa$ uncountable then I can show absoluteness of $\varphi$ for $H_\kappa$? $\endgroup$ – Marc Mar 9 '16 at 21:10
  • $\begingroup$ Yes, that's right! $\endgroup$ – GME Mar 9 '16 at 21:17
3
$\begingroup$

The proof is correct, but I'd change the order of a few things and extend a bit to make it slightly clearer:

We assume that $\exists x\varphi(x)$ is true in $V$, so there is some $a$ such that $\varphi(a)$ holds. Let $\vartheta$ an ordinal such that $a\in V_\vartheta$, since $V_\vartheta$ is transitive and $\varphi$ is a bounded formula, $V_\vartheta\models\varphi(a)$ and therefore $V_\vartheta\models\exists x\varphi(x)$. Let $M$ be a countable elementary submodel of $V_\vartheta$ and let $\overline M$ be its Mostowski collapse, then $\overline M\in H_{\omega_1}$.

Finally, since $M$ is an elementary submodel of $V_\vartheta$, and $M\cong\overline M$ we have that $\overline M\models\exists x\varphi(x)$. So there is some $a'\in\overline M$ such that $\overline M\models\varphi(a')$, but again by absoluteness of bounded formulas between transitive sets, $H_{\omega_1}\models\varphi(a')$ so $H_{\omega_1}\models\exists x\varphi(x)$ as wanted.

Essentially the same proof can be extended to show that $H_\kappa\prec_{\Sigma_1}V$, as remarked by GME in the comments.

$\endgroup$

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.