MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Suppose $S \subset \omega_{1}$ is stationary co-stationary. Then there is a forcing notion $P_{S}$ which shoots a closed unbounded $C \subset S$ without collapsing cardinals (or changing cofinalities).

Let $P$ = $Levy$($\aleph_{0}$, $< \aleph_{1}$). So P is essentially adding $\aleph_{1}$ Cohen reals. If $G_{P} \subset P$ is (directed and) generic over V, then $G_{P}$ is also generic over $L$ (the constructible universe) as $P \in L$ and also over $L[S]$. By Theorem I 6.7 the forcing P satisfies the countable chain condition and $\aleph_{1}^{L[G_{P}]} = \aleph_{1}^{V} = \aleph_{1}^{V[G_{P}]}$ and $V$, $V[G_{P}]$ have the same cardinals and cofinalities. Let

$Q$ = {$C$ : $C$ is a closed bounded subset of $S$ which belongs to $L[S, G_{P}]$} and $C_{1} < C_{2}$ iff $C_{1} = C_{2} \cap (Max(C_{1}) + 1)$.

Clearly $Q$ is a forcing notion of power $\aleph_{1}$, so it cannot collapse cardinals or regularity of cardinals except possibly $\aleph_{1}$ (all finite subsets of $S$ belong to $Q$). So we shall prove that $P * \dot{Q}$ does not collapse $\aleph_{1}$. So let (in $V$) $N \prec (H(\lambda), \epsilon)$, $N$ countable, $(p, \dot{q}) \in P * \dot{Q} \in N$, $(p, \dot{q}) \in N$ , $\delta := N \cap \omega_{1} \in S$ and suppose $G_{P} \subset P$ is generic over $V$ and $p \in G_{P}$. Note that as $S$ is a stationary subset of $\omega_{1}$, there is such $N$ (in $V$). So it is enough to find $(p',\dot{q'}) \geq (p, \dot{q})$ which is $(N, P * \dot{Q})$-generic. As $P$ satisfies the countable chain condition, p is $(N, P)$-generic (by 2.9), hence $N[G_{P}] \cap V = N$. Clearly $\dot{Q}[G_{P}] \in L[S, G_{P}] \subset V[G_{P}]$, now $N[G_{P}]$ does not necessarily belong to $L[S, G_{P}]$, but $N[G_{P}] \cap L[S, G_{P}]$ is $N[G_{\gamma}] \cap L_{\delta}[S, G_{P}] = L_{\delta}[S, G_{P}] \in L[S, G_{P}]$ and is a countable set in $L[S, G_{P}]$.

In $L[S, G_{P}]$ we have an enumeration of $\dot{Q}[G_{P}] \cap N[G_{P}]$ (of length $\omega$), say $\langle q_{n} : n < \omega \rangle$ (but not of the set of dense subsets); in fact we have it even in $L[S, G_{P} \upharpoonright(\delta + 1)]$ (and as we use $Levy(\aleph_{0}, < \aleph_{1})$ not $Levy(\aleph_{0}, < \kappa)$ even in $L[S, G_{P} \upharpoonright \delta]$). Now in $L[S, G_{P}]$ there is a Cohen generic real over $V[G_{P} \upharpoonright (\delta + 1)]$ say $r^{*} \in$ $^{\omega}{\omega}$ and we use it to construct a sequence $\bar{C} = \langle C_{n} : n < \omega \rangle$ such that $C_{n} \in \dot{Q}[G_{P}]$ in $L[S, G_{P}]$ i.e. $\bar{C} \in L[S, G_{P}]$, e.g. we choose $C_{n}$ by induction on $n$; we let $C_{0} = \dot{q}[G_{P}]$ and we let $C_{n+1}$ be $q_{m(n)}$ where $m(n)$ is the first natural number $m$ such that: $m \geq r^{*}(n)$ and $\dot{Q}[G_{P}] \models$ "$C_{n} \leq q_{m}$". Let

$G$ := {$q$ : $q \in \dot{Q}[G_{P}]$, and $q \in N[G_{P}]$ and for some $n$, $\dot{Q}[G_{P}] \models$ "$q \leq C_{n}$"}.

Clearly $G \subset N[G_{P}] \cap \dot{Q}[G_{P}]$ is generic over $V[G_{P} \upharpoonright (\delta + l)]$. So $q^{\dagger} = (\bigcup_{n} C_{n}) \cup \{\delta\} \in \dot{Q}[G_{P}]$ is $(N[G_{P}], \dot{Q}[G_{P}])$-generic. Going to names we finish.

A few questions regarding the proof above (Theorem 4.4 of Shelah's PIF). Here '$p \geq q$' means '$p$ extends $q$'.

First let $A \in N[G_{P}]$ be a maximal antichain of $\dot{Q}[G_{P}]$.

  1. How do we ensure that the $G$ constructed at the end of the proof intersect $N[G_{P}] \cap A$?
  2. If we cannot ensure the above, how do we show that $q^{\dagger}$ is ($N[G_{P}]$, $\dot{Q}[G_{P}]$)-generic?

I think there is some crucial properties of Cohen reals that I am missing here, such as how Cohen-genericity translate to $\dot{Q}[G_{P}]$-genericity, if at all.

Much thanks in advance.

share|cite|improve this question
up vote 3 down vote accepted

Sorry this is sketchy, but I hope it helps!

A Cohen condition $p$ of length $n$ is going to determine $\langle C_m:m<n\rangle$. We know there's going to be an $i<\omega$ such that $q_i$ extends both $C_{n-1}$ and a member of $N[G_P]\cap A$. Extend $p$ to a Cohen condition $p'$ with $p'(n)=i$. If $r^*$ is any Cohen generic extending $p'$, then running the construction with $r^*$ guarantees that $C_n$ extends a member of $N[G_P]\cap A$. So a density argument tells us the construction is forced to work.

The point is that in Shelah's construction, $m(n)$ is going to be equal to $r^*(n)$ very often because $r^*$ is Cohen.

share|cite|improve this answer
Answered this quickly as I'm leaving work. If on the way home I realize I've said something silly, I'll edit! – Todd Eisworth May 9 '13 at 20:59
Ah I get it! Thank you so much. Just one more question. To conclude that $P * \dot{Q}$ is proper from the argument in the proof, I reckon we would need {$N$ : $N$ is a countable subset of $H(\lambda, \epsilon)$ and $N \cap \omega_{1} \in S$} to contain a club. But I cannot seem to show this. Is this true or am I going in a wrong direction altogether? – Zoorado May 10 '13 at 14:12
It won't be proper if $S$ is co-stationary (as the stationarity of the complement of $S$ is destroyed), but it will preserve $\omega_1$ because you get generics for "enough" elementary submodels. – Todd Eisworth May 10 '13 at 15:19
Ok I see. Thanks again! – Zoorado May 11 '13 at 11:51

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.