Definition: Let $V\subseteq W$ be two transitive models of $ZFC$. A pseudo-Prikry sequence, $s$, at a cardinal $\kappa$ for $(V, W)$ is an $\omega$-sequence, cofinal at $\kappa$ such that for every club $D\subseteq \kappa$ from $V$, $s\setminus D$ is finite.

In the paper "On squares, outside guessing of clubs and $I_{<f}[λ]$" by Shelah and Dzamonja and independently in the paper "Some results on the nonstationary ideal II" by Gitik, stronger versions of the following theorem are proved:

Theorem: if $V\models \kappa$ is inaccessible, and $2^\kappa = \kappa^+$ and $W\models \kappa^+ = (\kappa^+)^V,\,\text{cf }\kappa=\omega$ then there is a pseudo-Prikry sequence at $\kappa$ in $W$.

I wonder how far can a pseudo-Prikry sequence get from just being the regular Prikry sequence.

For simplicity, let's assume that $\kappa$ is measurable in $V$, $\mathcal{U}$ is a normal measure on $\kappa$ and $W = V[G]$ where $G$ is a generic filter for the Prikry forcing for $\kappa$, using the measure $\mathcal{U}$.

Question: Does $W$ contain a pseudo-Prikry sequence which is not a $\mathcal{U}$-Prikry sequence?


The answer is no, there is no such pseudo-Prikry sequence in the Prikry extension.

To see this, let's first make some observations.

Lemma 1. If $j:V\to M$ is the ultrapower by a normal measure $\mu$ on $\kappa$, then $$\bigcap\{\ j(C)\mid C\subset\kappa\text{ club }\}=\{\kappa\}.$$

Proof. Certainly $\kappa\in j(C)$ for any club $C\subset\kappa$, and by considering final segments of $\kappa$, it is clear that the intersection is contained in $[\kappa,j(\kappa))$. So consider $\kappa<\alpha<j(\kappa)$. Since $\mu$ is normal, it follows that $\alpha=j(f)(\kappa)$ for some function $f:\kappa\to\kappa$. Let $C_f\subset\kappa$ be the club of ordinals $\beta$ with $f''\beta\subset\beta$. It follows that $\alpha\notin j(C_f)$, and so any particular $\alpha$ other than $\kappa$ is cast out of that intersection. QED

Now consider the $\omega$-iteration of $\mu$ $$V\to M_1\to M_2\to\cdots\to M_\omega$$

Lemma 2. Similarly, for the $\omega$-iteration embedding $j_\omega:V\to M_\omega$, we have $$\bigcap \{\ j_\omega(C)\mid C\subset\kappa\text{ club in }V\ \}=\{\ \kappa_n\mid n\in\omega\ \}.$$

Proof. For any particular club $C\subset\kappa$, since it is in $\mu$, we have that $\kappa_n\in j_\omega(C)$ for every $n$. Now suppose that $\kappa_n<\alpha<\kappa_{n+1}$ for some $n$. By standard representation results, we know that $\alpha=j_\omega(f)(\kappa_0,\ldots,\kappa_n)$ for some function $f:\kappa^{n+1}\to \kappa$. Let $C_f$ be the set of $\beta<\kappa$ such that $f''\beta^{n+1}\subset\beta$. This is club in $\kappa$, but by design, $\alpha\notin j_\omega(C_f)$. So every ordinal not on the critical sequence is cast out of the intersection. QED

It is a standard fact that the critical sequence $s=\langle\kappa_n\mid n\in\omega\rangle$ is $M_\omega$-generic for Prikry forcing with $\mu_\omega=j_\omega(\mu)$. Suppose that $t=\langle\delta_n\mid n\in\omega\rangle$ is a pseudo-Prikry sequence in $M_\omega[s]$. In particular, if $C\subset\kappa$ is club in $V$, then $j(C)$ contains a tail of the $\delta_n$. But for any particular ordinal $\alpha$ not on the critical sequence, there is a club $C\subset\kappa$ for which $\alpha\notin j_\omega(C)$. Thus, if infinitely many $\delta_n$ are not on the critical sequence, we can intersect the corresponding clubs to find a single club $C\subset\kappa$ such that $j_\omega(C)$ does not contain a tail of $t$, which would contradict our assumption that $t$ is a pseudo-Prikry sequence. Thus, $t$ must eventually be contained in the critical sequence. It follows that $t$ is a Prikry sequence over $M_\omega$, since any almost-subsequence of a Prikry sequence is a Prikry sequence.

Thus, every pseudo-Prikry sequence in $M_\omega[s]$ is actually a Prikry sequence, and then by elementarity, the same holds in the Prikry forcing extension of $V$.

  • $\begingroup$ Thank you! I think that your arguments generalize to a rather large class of forcing notions for changing cofinalities such as the supercompact Prikry forcing and the Radin forcing. $\endgroup$
    – Yair Hayut
    Feb 5 '15 at 14:25
  • $\begingroup$ It was my pleasure; it was a very interesting question. $\endgroup$ Feb 5 '15 at 14:27
  • $\begingroup$ @JoelDavidHamkins A nice question and a nice answer, I enjoyed the answer, thanks. $\endgroup$ Feb 8 '15 at 5:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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