18
$\begingroup$

For a partially ordered set $P$, a set $A\subseteq P$ is cofinal if for each element of $P$ there is a larger element in $A$. The cofinality of $P$, ${\rm cof}(P)$, is the minimal cardinality of a cofinal family in $P$.

Let $\kappa$ be an infinite cardinal number. Consider the set $\mathbb{N}^\kappa$ of all functions from $\kappa$ to $\mathbb{N}$ (equivalently, $\kappa$-sequences of natural numbers), with the pointwise partial order: $f\le g$ if $f(\alpha)\le g(\alpha)$ for all $\alpha<\kappa$.

Let $$f(\kappa):={\rm cof}(\mathbb{N}^\kappa,\le).$$ The cardinal $f(\aleph_0)$ is the well-understood dominating number $\mathfrak{d}$, which is consistently any cardinal of uncountable cofinality that is not larger than the continuum (see Blass).

Basic facts include:

  1. For cardinal numbers $\kappa\le\mu$, we have that $f(\kappa)\le f(\mu)$: A projection of a cofinal family in $\mathbb{N}^\mu$ on the first $\kappa$ corrdinates is cofinal in $\mathbb{N}^\kappa$.
  2. $\kappa<f(\kappa)$: Given $\{h_\alpha:\alpha<\kappa\}\subseteq\mathbb{N}^\kappa$, define $h(\alpha):=h_\alpha(\alpha)+1$ for all $\alpha<\kappa$.
  3. The value of $f(\kappa)$ with "$\le$" replaced by eventual dominance remains the same (see Comfort).

Some deeper results:

  1. Jech-Prikry: Assume that $2^{\aleph_0}$ is regular and smaller than $2^{\aleph_1}$. If $f(\aleph_1)=2^{\aleph_0}$, then there is an inner model with a measurable cardinal.
  2. If $\kappa^{\aleph_0}=\kappa$, then $f(\kappa)=2^\kappa$ (see Hathaway).

I did not conduct an extensive literature search. I would appreciate your pointing out relevant results that are not mentioned here.

Question: Are there additional cardinals $\kappa$ for which $f(\kappa)$ can be evaluated in ZFC?

(The question also applies to large cadrinals. I do not request that the cardinals provably exist.)

Update. Intially, I also asked whether $\kappa^{\aleph_0}=\kappa$ could be weakened to ${\rm cof}([\kappa]^{\aleph_0},\subseteq)=\kappa$. Todd Eisworth reminds in the comments below that the latter hypothesis is true in ZFC for $\aleph_1$.

Strangely, I couldn't find an explicit reference for the following.

Question. Is it consistent (modulo suitable large cardinals) that $f(\aleph_1)<2^{\aleph_1}$? Assuming that (See Assaf Rinot's comment below), is there an explicit reference for that?

Motivation: The values $f(\kappa)$ pop up whenever I study (usually, jointly with colleagues) the character and other cardinal invariants of topological groups. They are unavoidable.

(Comment: By no means do I mean that the symbol $f(\kappa)$ should be reserved for the function thus defined. This is just a short, temporary name for brevity.)

$\endgroup$
  • 2
    $\begingroup$ If the weakened implication in (2) holds then we would have $f(\aleph_1)=2^{\aleph_1}$ in ZFC. $\endgroup$ – Todd Eisworth Oct 18 '15 at 18:38
  • $\begingroup$ @ToddEisworth could you provide some details, or a reference? $\endgroup$ – Boaz Tsaban Oct 18 '15 at 18:58
  • $\begingroup$ A wild guess concerning the consistency of $f(\aleph_1)<2^{\aleph_1}$: have a careful look at Komjath-Shelah model of [F1445]. $\endgroup$ – saf Oct 18 '15 at 20:40
  • 2
    $\begingroup$ Just observe that the cofinality of $([\aleph_1]^{\aleph_0},\subseteq)$ is $\aleph_1$, as witnessed by initial segments. Same holds for all the $\aleph_n$ by induction. $\endgroup$ – Todd Eisworth Oct 18 '15 at 22:28
  • 2
    $\begingroup$ @BoazTsaban: Jech's Chapter 29 mentions that the question of whether $f(\aleph_1)<2^{\aleph_1}$ is consistent is open, and I seem to recall that the question was also mentioned in the original "green version" of his book from 1978. $\endgroup$ – Todd Eisworth Oct 21 '15 at 22:23

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.