2
$\begingroup$

Currently I am reading some basic literature on descriptive set theory and boolean algebras. And this comes out a lot, for example in results like:

Let $X$ and $Y$ be topological spaces, and $f:X \to Y$ be a continuous function such that for every nowhere dense subset $N \subset Y$, the subset $f^{-1}(N)$ is nowhere dense. Then, there is an order-continuous boolean homomorphism between algebras or regular open sets $\pi : \mathbf{RO}(Y) \to \mathbf{RO}(X)$, defined by $\pi(U) = \operatorname{int}\overline{f^{-1}(U)}$.

I wish to call functions like $f$ in the statement by some simple short name. I'd like to call such functions meager or thick maps. Another possible name I thought about is just nowhere dense maps. Basically, I could just use something like regular map or N-map. But those are probably not good. There is some parallel with maps with preimages preserving $\sigma$-ideals of measure zero-sets. So, if there is a conventional notion for them, It would be better to use it, as meager sets can be thought as constituting the zero $\sigma$-ideal for a $\sigma$-algebra of Baire-property sets.

But now I want to ask more knowledgeable people if there is some standard terminology in some books or papers I am not aware of? I understand that there is probably no standard notions fore the whole field, but I believe that somebody already faced this problem before me. So, I am not asking to choose a name, but to share some (possibly indirect) links to existing texts (or even talks).

Sorry if this question is just unanswerable because there are no cases of such terminology.

$\endgroup$
2
  • $\begingroup$ Community wiki? Or is your expectation that there is a unique, or unique best, term out there? \\ I did some proofreading, making only changes that seemed obvious, but in some cases I wasn't sure; please feel free to correct if I got it wrong. \\ TeX note: $\operatorname{int} V$ \operatorname{int} V spaces better than $\mathrm{int} V$ \mathrm{int} V. I changed that while proofreading. $\endgroup$
    – LSpice
    Commented Oct 8, 2021 at 23:13
  • 1
    $\begingroup$ I think this question can be made a community wiki, if it is not too minuscule. I don't hope to find a unique, or a unique best term, as I already knew that there are some people who chose not to use any. And being the best is too subjective. All your corrections are on point. $\endgroup$
    – Nik Bren
    Commented Oct 9, 2021 at 12:37

2 Answers 2

3
$\begingroup$

The property that you are referring to has been called various names such as Skeletal, demi-open, and weakly open [1]. Furthermore, this property has quite a few characterizations, and it generalizes well to point-free topology.

Recall that a frame is a complete lattice that satisfies the identity $$x\wedge\bigvee_{i\in I}y_{i}=\bigvee_{i\in I}(x\wedge y_{i}).$$

If $L$ is a frame, then define $x^{*}=\bigvee\{y\in L\mid x\wedge y=0\}$. If $L$ is a frame, then let $\mathfrak{B}L=\{x^{**}\mid x\in L\}$. Then $\mathfrak{B}L$ is the smallest dense sublocale of $L$. Define a mapping $\beta_{L}:L\rightarrow\mathfrak{B}L$ by letting $\beta_{L}(x)=x^{**}$. Then $\beta_{L}$ is a frame homomorphism.

Theorem: Suppose that $\varphi:L\rightarrow M$ is a frame homomorphism. Then the following are equivalent:

  1. There is some frame homomorphism $\psi:\mathfrak{B}L\rightarrow\mathfrak{B}M$ such that $\beta_{M}\varphi=\psi\beta_{L}$.

  2. $\varphi(a^{**})\leq\varphi(a)^{**}$ for each $a\in L$.

  3. $\varphi(a^{**})^{**}=\varphi(a)^{**}$ for each $a\in L$.

  4. $\varphi(a^{*})^{*}\leq\phi(a)^{**}$ for each $a\in L$.

  5. $\varphi(a^{*})^{*}=\phi(a)^{**}$ for each $a\in L$.

  6. For each dense $a\in L$, the element $\phi(a)$ is dense in $M$.

Furthermore, if $\varphi=\Omega f$ for some $f:X\rightarrow Y$, then the above conditions are equivalent to the following conditions:

  1. For each non-empty open set $U\subseteq X$, the set $(\overline{f[U]})^{\circ}$ is also non-empty.

  2. For each open $U\subseteq X$, there is an open $V\subseteq Y$ such that $\overline{V}=\overline{f[U]}$.

  3. If $N$ is nowhere dense in $Y$, then $f^{-1}[N]$ is nowhere dense in $X$.

  4. If $U$ is a dense open subset of $Y$, then $f^{-1}[U]$ is a dense open subset of $X$.

1 http://www.numdam.org/article/CTGDC_1996__37_1_41_0.pdf B. BANASCHEWSKI. A. PULTR. Booleanization Cahiers de topologie et géométrie différentielle catégoriques, tome 37, no 1 (1996), p. 41-60

$\endgroup$
1
$\begingroup$

In the recent paper Juhász, Soukup, and Szentmiklóssy - Spaces of small cellularity have nowhere constant continuous images of small weight, a continuous function with your required property is called pseudo-open. It makes sense, I think, because the property is equivalent to the fact that $f^{-1}(A)$ is an open, dense set whenever $A$ is open and dense.

Notice, however, that in the paper both the domain and the codomain are assumed to be Hausdorff and without isolated points.

$\endgroup$
1
  • $\begingroup$ Thank a lot. I think this notion makes sense because my function don't 'squish big sets' and open maps 'don't squish' open sets. This perfectly works for me, but I'm still curious about other options. $\endgroup$
    – Nik Bren
    Commented Oct 8, 2021 at 20:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .