6
$\begingroup$

Assume $V=L(\mathbb{R})$ and the Axiom of Determinacy. Is every set of reals generated by ordinal-definable sets of reals under the operations of countable union and intersection?

The class of sets generated in this way is Wadge-cofinal and not wellorderable (it contains $\lbrace x\rbrace$ for every $x \in \mathbb{R}$) so there don't seem to be obvious limitations on its extent.

This question came up when I was trying to answer Asaf Karagila's "bonus question" here: Generating family for the Lebesgue $\sigma$-algebra

$\endgroup$
0

1 Answer 1

7
$\begingroup$

This does not answer the question, but for every set $A\subset\mathbb{R}$ in $L(\mathbb{R})$, you can get $A\times\{r\}$ in your algebra, for some real $r$, and this does not use AD.

The reason is that if $V=L(\mathbb{R})$, then every set is definable from an ordinal and a real. So if $A\subset\mathbb{R}$, then $A=\{ x\in\mathbb{R}\mid \varphi(x,\alpha,r)\}$ for some definition $\varphi$ and parameters $\alpha\in\text{Ord}$, $r\in\mathbb{R}$. For any finite binary sequence $t$, let $B_t=\{(x,s)\mid t\subset s,\varphi(x,\alpha,s)\}$, which is an ordinal definable subset of the plane. Finally, observe that $A\times\{r\}=\{(x,r)\mid \varphi(x,\alpha,r)\}$ is precisely $\bigcap_{t\subset r} B_t$, a countable intersection of OD sets.

$\endgroup$
1
  • $\begingroup$ Thanks, yes, I noticed this too. So one way to get an affirmative answer would be to show that the $\sigma$-algebra is closed under $\exists^{\mathbb{R}}$. $\endgroup$ Jul 26, 2012 at 19:58

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.