why does the union of a generic filter cannot be the zero function?

76 Views Asked by At

I'm trying to understand some concept of the forcing idea, but I'm having trouble with two questions.
assume that M is countable s.t. $M\models ZFC$, $\mathbb{P}$ is cohen forcing, $\mathbb{P}\in M $,and G is an $\mathbb{P}$-genric filter over M.

  1. i already know that $\bigcup G$ is a function $f:\omega \to 2$, why does $f$ cannot be the zero function?
  2. why does $G\notin M$?

for 1 i know that $$D=\left\{ p\in\mathbb{P}:1\in\text{rng}\left(p\right)\right\} $$ is dense in $\mathbb{P}$ but i was unable to proof that $D\in M$. $$$$ Thanks a lot

1

There are 1 best solutions below

0
On

Density, density, density. Almost all forcing arguments of this sort boil down to finding the appropriate dense sets.

For example, why isn't $f$ constant? Because given any $p$, there is an extension of $p$ which takes both $0$ and $1$, which one? Well, for example, just add "on top" of $p$ these two values.

Therefore, the set $\{p\mid \operatorname{rng}(p)=\{0,1\}\}$ is dense. Since it is definable from a formula, this dense set is in the ground model, so $G$ meets it, and therefore some $p\in G$ must have both $0$ and $1$ in its range. Therefore $f$ is not constant.

Similarly, $G$ can't be in $M$ itself, because if $r\in M$, then $\{p\mid\exists n: p(n)\neq r(n)\}$ is a dense set, for the same reason as above. And it is in $M$ because it is definable from $r$, and $M$ is a model of $\sf ZF$, or at least a large enough fragment of it. So this dense set is in $M$, and therefore it meets $G$, and therefore $f$, the generic real, must be different from $r$, for any $r\in M$, and so $f\notin M$.