0
$\begingroup$

In class (and in parallel with Topology by Munkres), we proved the following theorem:

Let $\mathcal{B}$ be a topological basis on $X$, $\mathscr{C}$ be a topological basis on $Y$. Then, $\mathscr{D} = \{B \times C : B \in \mathcal{B}, C \in \mathscr{C} \}$ is a topological basis for the product topology on $X \times Y$.

I can mostly follow the proofs from class and the book, and the crux is to use the lemma that states:

Let $(X,\mathcal{T})$ be a topological space, and $\mathscr{C}$ be a collection of open sets in X, that satisfies: $\forall U \in \mathcal{T}.\forall x \in U.\exists C \in \mathscr{C}. x \in C \subset U$. Then $\mathscr{C}$ is a topological basis generating $\mathcal{T}$.

From the definition and the lemma, I (think I) follow the proof in Munkres and in my notes that shows that the elements of $\mathscr{D}$ satisfy the condition of lemma 2. However one part of the proof I am simply failing to see is how do we know $\mathscr{D}$ is a collection of open sets in $X\times Y$?

The proof in Munkres doesn't address this because I think it's meant to be an obvious conclusion. In my notes I have written "First, by the definition of the product topology of $X \times Y$, all elements of $\mathscr{D}$ are open."

I understand why the definition implies the basis elements are open (or rather I think it's because the elements of any basis are open), but I don't see how this helps since we haven't yet proved that $\mathscr{D}$ is a basis, let alone the basis generating the product topology? Why are we using the definition to justify assumptions on the set for which we are trying to prove is a bases?

My apologies if I am missing something obvious (I suspect I am...)

$\endgroup$
  • 1
    $\begingroup$ What is your definition of the product topology? For a finite product, it's usually the topology given by $B \times C$ where $B$ is open in $X$ and $C$ is open in $Y$ right? So since $\mathcal{B}$ and $\mathcal{C}$ are bases for their respective topologies, in particular their elements are open, and so $\mathcal{D}$ like definitially consists of opens on $X\times Y$. $\endgroup$ – user113102 Oct 21 '19 at 19:22
  • 1
    $\begingroup$ What is the definition of the product topology you are using. One standard definition is the topology generated by $\{U \times V \mid U \subseteq X \text{ open}, V \subseteq Y \text{ open}\}$. If this is the definition, the openness of the elements of $\mathscr{D}$ is trivial since elements of $\mathscr{B,C}$ are open as they are basis elements. $\endgroup$ – Keefer Rowan Oct 21 '19 at 19:25
  • $\begingroup$ Yes that's exactly the definition I am using! I'll add that into the question when I am off mobile. However I am still struggling to see how it is trivially open. I see how B and C are trivially open, but I don't see why their cartesian product is necessarily open? $\endgroup$ – masiewpao Oct 21 '19 at 19:37
  • $\begingroup$ I thought about the comment and I completely realise why now. The answer also helped me identify it precisely, hank you! $\endgroup$ – masiewpao Oct 22 '19 at 1:00
1
$\begingroup$

Because the product topology is defined as a ( actually the smallest) topology that makes the projections continuous, we know $\mathcal{D}$ is a collection of open sets: $B \in \mathcal{B}$ implies $B$ is open in $X$ and $C \in \mathcal{C}$ implies $C$ is open in $Y$ (the bases on $X$ and $Y$ also consist of open sets of their resp. space!).

Then $$B \times C= \pi_X^{-1}[B] \cap \pi_Y^{-1}[C]$$

is the (finite) intersection of two open subsets of $X \times Y$ in the product topology (using that projections are continuous), and so is open.

| cite | improve this answer | |
$\endgroup$

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.