1
$\begingroup$

From Wikipedia:

OD is not necessarily transitive, and need not be a model of ZFC.

This obviously means that, assuming ZFC is consistent, there is a model $M \models \mathrm{ZFC}$ so that $\mathrm{OD}^M \not \models \mathrm{ZFC}$, and yet $M \models M \models \mathrm{ZFC}$ by absoluteness of $\Delta_0$-formulae, and so $M \models \mathrm{ZFC} + V \neq \mathrm{OD}$.

Now, if $V \neq \mathrm{OD}$, then $V_\alpha \cap \mathrm{OD} \subset V_\alpha$ for some $\alpha$ - if such an $\alpha$, what is it or bounds on it? Can we find models $M$ in which the least such $\alpha$ is unboundedly large (i.e. V can be arbitrarily close to OD while not being equal), or would instead the least $\alpha$ be recursive, countable, etc? If I remember correctly, ordinal definability is not absolute, so what would the least $\alpha$ such that $\mathrm{OD}^{V_\alpha} \neq V_\alpha \cap \mathrm{OD}$?

$\endgroup$
4
  • 2
    $\begingroup$ I think that any model of $V=L$ will have all sets ordinal definable, so you can just force over $L$ to add generic subsets of arbitrarily large $\alpha$ without adding any smaller sets. $\endgroup$ Commented Jun 11, 2023 at 16:28
  • $\begingroup$ For the record, $V=OD$ is equivalent to $V=HOD$, and $HOD$ is always an inner model. $\endgroup$
    – Wojowu
    Commented Jun 11, 2023 at 18:44
  • $\begingroup$ How is $M\vDash M\vDash ZFC$ formulated (as a first-order sentence)? Additionally, being a model of ZFC is not absolute: Hamkins, Yang, "Satisfaction is not absolute". $\endgroup$
    – C7X
    Commented Jul 8, 2023 at 5:37
  • $\begingroup$ Uh... I thought satisfaction was absolute. Uh oh. I have no idea. $\endgroup$
    – Binary198
    Commented Jul 9, 2023 at 12:58

0

You must log in to answer this question.

Browse other questions tagged .