2
$\begingroup$

I was wondering if $(1)$ the product of an infinite number of lattices is a lattice and $(2)$ if the product of an infinite number of complete lattices is a complete lattice.


More precisely, suppose that $\{L_{a},a\in A\}$ is a family of lattices, where A is potentially infinity (e.g., $A=\mathbb{N}$ or $A=[0,1]$). Is $L=\times_{a\in A} L_{a}$ a lattice? What if $\{L_{a},a\in A\}$ is a family of complete lattices. Is $L=\times_{a\in A} L_{a}$ a complete lattice as well? I know that both statements hold for finite products, but do they extend to infinite products?


I believe that the answer to the first question is yes. We simply define the "standard" order on $L$ where $x\geq y$ if $x_{a} \geq y_{a}$ for all $a \in A$, $x \vee y = \{x_a \vee y_a \}_{a \in A}$, and $x \wedge y = \{x_a \wedge y_a \}_{a \in A}$.

The second question seems to be more complicated. I am also inclined to say that it is true, but I am less sure of it. In particular, I was thinking if the approach takes from finite products could be extended here. In particular, let $K\subset L$. Define $K_a=\{x_a\in L_a | \exists y\text{ in } K \text{ with } y_a=x_a \}$. Then $K_a$ is a subset of $L_a$ and since $L_a$ is complete then $sup(K_a)$ exists in $L_a$. Then $sup(K)=\{sup(K_a)\}_{a\in A}$ which belongs to $L$ since $sup(K_a)\in L_a$ for all $a \in A$.


$\endgroup$
  • 1
    $\begingroup$ You are right about the first question. For the second, however, you haven't proved that what you call $\sup(K)$ is actually a supremum: you should prove it. $\endgroup$ – Maxime Ramzi May 31 '18 at 10:47
  • $\begingroup$ I see, thanks for pointing out this gap. $\endgroup$ – Mdoc May 31 '18 at 21:51
2
$\begingroup$

Your $K_a$ is called the $a$-th projection of $K$, and you want to prove that $$\bigvee K = \left(\bigvee K_a:a \in A\right).$$

(Notice that you notation, is not correct. When you write $x \vee y = \{x_a \vee y_a\}_{a \in A}$, you saying that the join of $x$ and $y$ is given by that set, but it must be a sequence, because the order matters. So it should be $x \vee y = (x_a \vee y_a)_{a \in A}$ or $x \vee y = (x_a \vee y_a :a \in A)$, or something similar.)

To prove the above equality, let $v_a = \bigvee K_a$, and $v = (v_a:a \in A)$.
Now we prove that $v$ is the join of $K$.
First, let $k \in K$. For each $a \in A$, we have that $k_a \in K_a$, whence $k_a \leq v_a$, yielding $k \leq v$.
Now let $u = (u_a:a\in A)$ be any upper bound of $K$, that is, $k \leq u$ for all $k \in K$.
It follows that $k_a \leq u_a$, for every $k \in K$ and every $a \in A$, whence, by the definition of supremum, $v_a \leq u_a$, for every $a \in A$, and therefore $v \leq u$.
So $v$ is the least upper bound of $K$, QED.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Thank you for the answer! And thank you for pointing out the issue with the notation. I am trying to learn a bit of lattice theory on my own so such small but important details escape my notice. $\endgroup$ – Mdoc May 31 '18 at 21:52
  • $\begingroup$ @Mdoc I'm glad it was helpful. Good luck with those studies! $\endgroup$ – amrsa Jun 1 '18 at 11:10

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.