
TL;DR
This paper investigates how certain topological properties are preserved under $G_\delta$-refinements in specific classes of scattered spaces, addressing open questions about tightness in $G_\delta$-refinements of $\sigma$-products.
Contribution
It establishes preservation results for metacompactness, paralindel"ofness, metalindel"ofness, and linear lindel"ofness in $ ext{SP}$-scattered spaces under $G_\delta$-refinements, and explores related generalizations.
Findings
Preservation of several covering properties in $ ext{SP}$-scattered spaces.
Analysis of $G_\delta$-refinements in $ ext{N}$-scattered and $ ext{ω}$-scattered spaces.
Addressing the question of tightness in $G_\delta$-refinements of $\sigma$-products.
Abstract
In this work we deal with the preservation by -refinements. We prove that for -scattered spaces the metacompactness, paralindel\"ofness, metalindel\"ofness and linear lindel\"ofness are preserved by -refinements. In this context we also consider some other generalizations of discrete spaces like -scattered and -scattered. In the final part of this paper we look at a question of Juh\'asz, Soukup, Szentmikl\'ossy and Weiss concerning the tighness of the -refinement of a -product.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsFuzzy and Soft Set Theory · Digital Image Processing Techniques · Advanced Algebra and Logic
-refinements
Robson A. Figueiredo
Instituto de Matemática e Estatística da Universidade de São Paulo
Rua do Matão, 1010, Cidade Universitária, CEP 05508-090, São Paulo, SP, Brazil
Abstract.
In this work we deal with the preservation by -refinements. We prove that for -scattered spaces the metacompactness, paralindelöfness, metalindelöfness and linear lindelöfness are preserved by -refinements. In this context we also consider some other generalizations of discrete spaces like -scattered and -scattered. In the final part of this paper we look at a question of Juhász, Soukup, Szentmiklóssy and Weiss concerning the tighness of the -refinement of a -product.
1. Preliminaries
Definition 1.1**.**
For any space , the topology obtained by letting every subset of be open is called the -topology and the space so obtained is denoted by .
Definition 1.2**.**
Let be a set and let be a cardinal such that . We say that is cofinal in if, for all , there exists such that . For cardinals , we define as the least cardinality of a cofinal family in . Given a infinite cardinal , we define .
Theorem 1.3** (Passos [Pas2007]).**
*Let be a infinite cardinal such that . Given a set of cardinality , there exists an -covering elementary submodel such that and . *
Recall that a subset of is -closed, where is an infinite cardinal iff whenever and then . It is well known that iff every -closed set in is closed.
2. -scattered spaces
Definition 2.1**.**
A point in a topological space is called a strong -point if it has a neighborhood consisting of -points. The set of all strong -points of is denoted by .
Observe that .
Definition 2.2** ([HRW2007]).**
Recursively, define:
- •
and ;
- •
for any ordinal ;
- •
, if is a limit ordinal.
In [HRW2007], Henriksen, Raphael and Woods proved the following generalizations of the well known theorems 5.1 and 5.2 of [LR1981], respectively:
Theorem 2.3**.**
If is a Lindelöf -scattered regular space, then is Lindelöf.
Theorem 2.4**.**
If is a paracompact -scattered Hausdorff space, then is paracompact.
In the same article they asked:
Question 2.5**.**
If is a metacompact -scattered regular space, so is a metacompact space?
In this section, we will see that not just the metacompactness, but also the paralindelöfness, the metalindelöfness and the linear lindelöfness are preserved by -refinements on the class of -scattered regular spaces.
Theorem 2.6** ([Mis1972]).**
A regular -space is paralindelöf if, and only if, it is paracompact.
Proposition 2.7** ([HRW2007]).**
If is a regular space, then the following are equivalent:
- (1)
* is -scattered;* 2. (2)
if is nonempty, then \mathrm{int}_{A}\{\,a:\text{aPA}\,\}\neq\emptyset.
Theorem 2.8**.**
If is a regular -scattered paralindelöf space, then is paracompact.
Proof.
By the theorem 2.6, it is enough to show that is paralindelöf. Let be an open cover of . Let be the set of all points such that for some locally countable open partial refinement of in .
If then, for each , there exists a locally countable open partial refinement of in such that . Since is paralindelöf, the open cover admits a locally countable open refinement , with whenever . For each , take such that . So
[TABLE]
is locally countable open refinement of in .
Now it remains to show that in fact . Suppose this is not the case. Since is -scattered, by the proposition 2.7, there exists a and an open neighborhood of in such that is a -subspace of . Take such that . We can suppose that , where, for each , and . So is an open subset of , for is a subspace of the -space . As is regular, has an open neighborhood in such that .
Fix . Let . Since , for each , there exists a locally countable open partial refinement of in such that . As is paralindelöf and is closed, admits a locally countable open partial refinement that covers , where whenever . For each , choose such that . Consider the family
[TABLE]
The family is a locally countable open cover of in . Indeed, let . Since is a locally countable open family in , there exists an open neighborhood of in such that is countable. For each , take an open neighborhood of in such that is countable. Consider the following open neighborhood of em :
[TABLE]
Note that for each , is countable. Seeing that
[TABLE]
the family is countable.
Thus,
[TABLE]
is a locally countable open partial refinement of in such that , contradicting the fact that . ∎
Theorem 2.9**.**
If is a regular -scattered metalindelöf space, then is metalindelöf.
Proof.
Let be an open cover of and consider the set whose elements are all those such that for some pointwise countable open partial refinement of in .
If then, proceeding in the same way as in the proof of theorem 2.8, we can obtain a pointwise countable open refinement of in .
In order to complete the proof, it is enough to show that . Suppose on the contrary that . As is regular and -scattered, by the proposition 2.7, there exist a point and an open neighborhood of in such that is a -subspace of . Choose a such that . We can suppose that , where for each , and . Note that is an open subset of . Hence, from the regularity of it follows that there exist an open neighborhood of in such that .
Fix . Note that . Then, for each , there is a pointwise countable open partial refinement of in such that . Because is metalindelöf and is closed then has a pointwise countable open partial refinement , where whenever . For each , choose such that . Consider the family
[TABLE]
It is easily checked that each is a pointwise countable open partial refinement of which covers . So,
[TABLE]
is a pointwise countable open partial refinement of in such that , contradicting the fact that . Thus, . ∎
Theorem 2.10**.**
If is a regular -scattered metacompact space, then is metacompact.
Proof.
Let be an open cover of and consider the set whose elements are all such that for some pointwise finite open partial refinement of in .
Similarly to what it has been done in theorem 2.8, we can get, from the assumption , a pointwise finite open refinement of in .
We complete the proof by showing that . Suppose that . Since is -scattered and regular, by the proposition 2.7, there are a point and an open neighborhood of in such that is a -subspace of . Take a such that . We can suppose that , where for each , and . Note that is an open subset of . Once is regular, has an open neighborhood in such that .
As , for each , there exists a pointwise finite open partial refinement of in such that . Since is metacompact and is closed then has a pointwise finite open partial refinement which covers . We can suppose that , where whenever . For each , choose such that . Consider the family
[TABLE]
Note that each is a pointwise finite open partial refinement of in such that . Therefore,
[TABLE]
is a pointwise finite open partial refinement of in such that , contradicting the fact that . ∎
Theorem 2.11**.**
If is a regular -scattered linearly Lindelöf space, then is linearly Lindelöf.
Proof.
Let be an open cover of , where is an uncountable regular cardinal. For each , let
[TABLE]
Define
[TABLE]
Claim**.**
**
Proof of claim
Suppose on the contrary that . As is a -scattered regular space, by the proposition 2.7, there are and an open neighborhood of in such that is a -subspace of . Choose such that . We can suppose that , where, for each , and . Note that is an open subset of . Once is regular, has an open neighborhood in such that .
Fix . Let . Note that is a closed subset of and so it is linearly Lindelöf. Moreover, ; this implies that, for each , we can take such that . Then, is a family of open subsets of which covers and it is linearly ordered by inclusion. Therefore, there is a countable subset such that covers . So, e
[TABLE]
Then and, thus, . This is a contradiction.
By the claim above, is an open cover of . Since is a linearly Lindelöf space, has a subcover whose cardinality is less than . Because is regular, for some . Thus, is a subcover of whose cardinality is less than . ∎
3. Other generalizations of scattered
Clearly, if a regular Lindelöf space is a countable union of scattered closed subspaces, then is Lindelöf. As we shall see, at least consistently, this is not the case when it is not required that the subspaces are closed.
A space is -scattered if it is an union of a countable family of scattered subspaces.
Example 1**.**
Assuming , there exists a regular -scattered Lindelöf space whose -refinement is not Lindelöf.
Proof.
It is enough to take a Luzin subset of the real line containing the rational numbers and consider it as a subspace of the Michael line. ∎
Question 3.1**.**
Is there a regular -scattered Lindelöf space whose -refinement is not Lindelöf?
Question 3.2**.**
Is there a regular -scattered paracompact space whose -refinement is not paracompact?
Hdeib and Pareek introduced in [HP1989] the following natural generalization of scattered spaces: a space is -scattered if, for each non-empty subset of , there exist a point and an open neighborhood of such that is countable.
Every scattered space is -scattered, but the reverse is not true: the set of rational numbers with the usual topology is -scattered and non-scattered.
The theorem 3.12 of [HP1989] states that in the class of regular -scattered spaces the Lindelöf property is preserved by -refinements. However, this is not true once the space of the example 1 is -scattered.
Question 3.3**.**
Is there a Hausdorff -scattered paracompact space such that is not paracompact?
A space is -scattered if every nowhere dense subset of is a scattered subspace of . The next example was noticed by Santi Spadaro.
Example 2**.**
Assuming , there exists a -scattered Lindelöf space whose -refinement is not Lindelöf.
Proof.
Let the family of all Lebesgue measurable subsets of the real line. For each , define
[TABLE]
Then
[TABLE]
is a topology on stronger than that usual, well known as density topology. Denote by the topological space . By corollary 4.3 of [Tal1976], implies that has a hereditarily Lindelöf, non-separable, regular and Baire subspace . By theorem 2.7 of [Tal1976], every nowhere dense subset of is discrete (and closed). Therefore, is -scattered. On the other side, the pseudocharacter of is countable, for is Hausdorff and hereditarily Lindelöf. Then is discrete and uncountable and, thus, it is not Lindelöf. ∎
Question 3.4**.**
Is there a Hausdorff paracompact -scattered space whose -refinement is not a paracompact space?
4. The tightness of -refinement of -products
Given a family of topological spaces and a point , define
[TABLE]
where . The -product of at is the set equipped with the topology induced by the Tychonoff product .
In [JSSW], Juhász, Soukup, Szentmiklóssy and Weiss proved:
Theorem 4.1**.**
Let and be cardinals, with . Let be the one point lindelöfication of a discrete space of cardinality by a point and let , where for all . Then has tighness .
In the same article, it was asked:
Question 4.2**.**
Assume that is a Lindelöf -space such that . Is it true that
[TABLE]
for all cardinal ?
We will see that the answer is positive.
Lemma 4.3** ([DW1986]).**
If is a finite family of regular locally Lindelöf -spaces, then
[TABLE]
Lemma 4.4**.**
If is a -product of regular locally Lindelöf -spaces, then
[TABLE]
Proof.
Let . Let be a non-closed subset of and let . For each , let
[TABLE]
Since and is a -space, there exists a such that . Now, , where is the natural projection from in . Since by the corollary 4.3 the tightness of is , there exists of cardinality such that . Then and since , we have . ∎
Lemma 4.5**.**
Let be a infinite cardinal. Let . If for each countable subset , has tightness , with , then has tightness .
Proof.
Let be a non-closed subset of and let . By the theorem 1.3, there exists a -covering elementary submodel of cardinality such that . We are going to show that . Suppose that
[TABLE]
where is a countable subset of and each is an open subset of . Since is -covering, there exists , a countable subset of such that . Now note that and ; besides belongs to and, by hypothesis, its tightness is . So by the elementarity there exists , a subset of whose cardinality is at most , such that . Let . Note that , because, since and has cardinality at most and , . So by the elementarity there exists such that .
We claim that . Indeed, since , , it follows that if so . On the other side, if so and thus . Therefore, . ∎
Theorem 4.6**.**
If , with each being a Lindelöf -space such that , with , then
[TABLE]
In particular, if is a Lindelöf -space whose tightness is then the tightness of is .
As a corollary of the previous theorem we have that, for a regular Lindelöf -space,
[TABLE]
It remains to be seen whether:
Question 4.7**.**
Is there a Lindelöf -space such that , with , and
[TABLE]
Question 4.8**.**
Assuming that , is there a Lindelöf -space such that and
[TABLE]
Based on the theorem 3.1 from [DW1986], we have:
Lemma 4.9**.**
Let be a infinite cardinal. If is a -product of regular Lindelöf -spaces, then
[TABLE]
Proof.
Let . For each , let . Suppose that is -closed, and .
Note that, for each countable subset , . Indeed, let . Then there exists such that . If is an basic neighborhood of in , then is an open neighborhood of . So and, thus, . Therefore, .
Then, for each countable subset , . By lemma 4.4, , then we can take of cardinality such that . For each choose such that , and let .
Now, let be a cofinal family in and let
[TABLE]
Note that . Then . So, it remains to be proved that . Let be an basic neighborhood of in . Let such that . Since , then ; so . Therefore, . ∎
In the same way we have proved the lemma 4.4, we can show the following result for the cases in which :
Lemma 4.10**.**
If is a Lindelöf -space then
[TABLE]
Proof.
Let . Suppose that . Note that . Let be a non-closed subset of and let . For each , let
[TABLE]
Since and is a -space, there exists a such that . Now, , where is the natural projection from in . Since by the theorem 4.9 the tightness of is , there exists of cardinality such that . Then and since , we have . ∎
References
