Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I have a partial answer for the following homework and I was wondering if you could tell me if I have it right and help me with part (c) which I'm currently stuck on. Many thanks for your help!

Let the forcing partial order $\mathbb{U} = ([\omega]^\omega / \mathrm{fin} , \leq)$ be defined as follows:

Define an equivalence relation on $[\omega]^\omega$ by stipulating $ x \sim y \iff x \Delta y$ is finite and let $[\omega]^\omega / \mathrm{fin} := \{ [x] \mid x \in [\omega]^\omega \} $. On $[\omega]^\omega / \mathrm{fin}$ define a partial ordering $\leq$ by stipulating $[x] \leq [y] \iff y \setminus x$ is finite.

Now let $V \models ZFC $ and $G$ be an $\mathbb{U}$-generic filter over V. In the model $V[G]$ let $\mathcal{U} := \bigcup G$

(a) Show that $\mathcal{U}$ is a set of infinite subsets of $\omega$, i.e. $\mathcal{U} \subset [\omega]^\omega$.

My answer:

$\bigcup G = [\bigcup_{g \in G} g ]$ where $g \in [\omega]^\omega$ so for any $y \in \bigcup G$, $y$ has cardinality $\omega$.

(b) Show that $\bigcup G$ is a filter

My answer:

(i) $\bigcup G \neq \emptyset$: If $\bigcup G = \emptyset$ then $G = \emptyset$ which would contradict $G$ being a filter.

(ii) Any two elements in $\bigcup G$ are compatible, i.e. for any $p,q$ in $\bigcup G$ there is an $r$ in $\bigcup G$ such that $p \leq r$ and $q \leq r$:

Let $p,q$ be in $\bigcup G$. Then there exists $g_q , g_p$ in $G$ such that $p \in g_p$ and $q \in g_q$. $g_i$ are sets of ordinals and so $p$ and $q$ are ordinals, too. So either $p \subset q$ or $q \subset p$. Set $r := \max_{\text{w.r.t.} \subset} p, q$, then $p \leq r$ and $q \leq r$.

(iii) If $p,q \in \bigcup G$ and $p \in \bigcup G$ and $q \leq p$ then $q \in \bigcup G$:

Let $p,q \in \bigcup G$ and $p \in \bigcup G$ and $q \leq p$. $q \leq p$ means that $|p \setminus q| < \omega $ which implies that $|p \cap q| = \omega$ which means that $|p \Delta q| < \omega$ and hence $[p] = [q]$ and hence $[q] \in \bigcup G$.

(c) Show that for every set $x \in [\omega]^\omega$ in the ground model $V$ there exists an $y \in \bigcup G$ such that either $y \subset x$ or $y \cap x = \emptyset$.

Here I'm stuck. I thought I could do a proof by contradiction by assuming that there exists $x \in [\omega]^\omega$ such that for all $y \in \bigcup G$, $x \subset y$ but this doesn't lead anywhere.

Many thanks for your help!


$p \leq q$ reads as $p$ weaker than $q$.

share|improve this question
There’s definitely something wrong in (b)(iii): $|p\setminus q|<\omega$ does not imply that $|p\Delta q|<\omega$, because $q\setminus p$ could still be infinite. You’re essentially saying that if $[x]\le[y]$, then $[x]=[y]$, which is clearly false in general. –  Brian M. Scott Dec 10 '11 at 21:43
@BrianM.Scott Thanks Brian! By now I can see it myself : ) –  Rudy the Reindeer Feb 13 '12 at 19:33

2 Answers 2

For part (c) try using a density argument. For each $x\in [\omega]^\omega$, define a dense set $D_x$ in such a way that any $y\in D_x\cap G$ implies that either $y\subset x$ or $y\cap x= \varnothing.$

share|improve this answer
up vote 1 down vote accepted

Here is future self to past self:

Before we start manipulating symbols blindly let's sit back and take a leisurely look at what the exercise is asking (it's a very nice exercise):

We start with a model $\mathbf V$ of $ZFC$, a forcing partial order $\mathbb U$ and a $\mathbb U$-generic filter $G$. It's important not to confuse the notion of a filter with that of a generic filter: the elements in $\bigcup G$ are sets and not equivalence classes of sets like the elements in $\mathbb U$ and we are talking about a filter on the set $[\omega]^\omega$ and not a generic filter in $\mathbb U$.

We are asked to show that $\bigcup G$ is a set (part (a)) and what's more, that it is an ultrafilter (parts (b) and (c)). Well, to be precise, part (a) asks you to explain why it is a set and also why its elements are infinite sets -- a property which will come in handy in part (b).

That $\bigcup G$ is a set follows from $G\subseteq \mathbf V [G], G \in \mathbf V [G]$ and $\mathbf V [G] \models ZFC$ -- of course $\mathbf V [G]$ has to satisfy "Union" which says exactly that the union of a set is again a set.

As for part (b), we need to show the following three properties of a filter:

(i) $1 = \omega \in \bigcup G$: To this end observe that $D = \{ [\omega]\}$ is dense in $\mathbb U$ and since $G$ is generic, $G \cap D \neq \varnothing$ so that $[\omega] \in G$ and hence $\omega \in \bigcup G$.

(ii) If $a \subseteq b$ and $a$ is in $\bigcup G$ then $b \in \bigcup G$: If $a$ is in $\bigcup G$ we have $[a] \in G$ and since $a \setminus b = \varnothing$ we have $[a] \geq [b]$. Since $G$ is generic it follows that $[b] \in G$.

(iii) If $a,b \in \bigcup G$ then $a \cap b \in \bigcup G$: Let $a,b \in \bigcup G$. Then since $G$ is generic there is $[r] \in G$ such that $[r] \geq [a]$ and $[r] \geq [b]$. Observe that this means that both $r \setminus b$ and $r \setminus a$ are finite. Hence their union is also finite. But $r \setminus (a \cap b) = r \setminus b \cup r \setminus a$ so that $[r] \geq [a \cap b]$ and hence $[a \cap b] \in G$.

For part (c) define $D_x = \{[y] \mid y \in [\omega]^\omega \text{ such that } y \subseteq x \text{ or } y \cap x = \varnothing \}$. Then $D_x$ is dense in $\mathbb U$: Let $z \in [\omega]^\omega$. Then either $z \cap x = \varnothing$ or $z \cap x \neq \varnothing$. If the former, $[z] \in D_x$ and $[z] \geq [z]$. If the latter, let $y = z \setminus x$. Then $y \setminus z = \varnothing$ or in other words, $[y] \geq [z]$. Since $G$ is generic, there is $y \in G \cap D_x$ which proves the claim.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.