1
$\begingroup$

In at least two papers (here and here) Jorge Martínez and Eric R. Zenk say that Zorn's Lemma implies that all algebraic frames are spatial. However, I haven't been able to find an actual explanation or proof of this claim anywhere, nor can I think of one myself. Does anyone have an idea or a reference on proving this?

$\endgroup$
4
  • 1
    $\begingroup$ More generally (under AC) any continuous frame is spatial - see e. g. VII.4.3 in Johnstone's "Stone Spaces" (pp. 310-311) $\endgroup$ Dec 15, 2015 at 21:25
  • $\begingroup$ I don't have access to that book out in the sticks unfortunately. Is there a sketch of the proof available online? $\endgroup$ Dec 17, 2015 at 18:11
  • 1
    $\begingroup$ I am not aware of any. If you have access to "Continuous Lattices and Domains" by Gierz, Hofmann, Keimel, Lawson, Mislove and Scott, there on page 126 is Theorem I-4.25 stating that for any $x\not\leqslant y$ in a bounded complete algebraic domain there is a completely irreducible element $p$ with $x\not\leqslant p$ and $y\leqslant p$. $\endgroup$ Dec 17, 2015 at 19:57
  • $\begingroup$ Every distributive continuous lattice is a spatial frame. See "The spectral theory of distributive continuous lattices", Karl H. Hofmann and Jimmie D. Lawson 1978. $\endgroup$ Jul 25, 2019 at 14:52

1 Answer 1

1
$\begingroup$

Let $L$ be a complete lattice with top element $1$. Let ${\cal M}(L)$ be the collection of meet-irreducible elements of $L$.

Recall that a frame is said to be spatial if for all $x\in L$ with $x<1$ we have $x=\bigwedge\{z\in {\cal M}(L): z\geq x\}$.

Lemma: Every complete, algebraic lattice $L$ contains a meet-irreducible element that is stricly smaller than the top-element.

Proof. Let $0$ be the bottom element, and $1$ be the top element of $L$. Since $L$ is compact, $1$ is the join of all compact elements of $L$. Let $c>0$ be a compact element and set $$K = \{k\in L: c\not\leq k\}.$$ Clearly $0\in K$ and the join of any directed subset of $K$ is in $K$ since $c$ is compact. By Zorn's Lemma, $K$ contains a maximal element $m\in K$. By definition, any $x>m$ must contain $c$, so $m\vee c$ is the unique least element of $L$ that properly contains $m$, and so $m$ is meet-irreducible.

Theorem: If $L$ is complete, algebraic, then for all $x\in L$ with $x<1$ we have $x=\bigwedge\{z\in {\cal M}(L): z\geq x\}$.

Proof. Suppose $x < m:=\bigwedge\{z\in {\cal M}(L): z\geq x\}$. Then $[x,m]=\{a\in L: x\leq a \leq m\}$ is a complete algebraic lattice without meet-irreducible elements other than the top-element, contradicting the Lemma.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.