MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Basically Stone duality or more general the duality between spatial locales and sober spaces is about enriching the set of morphisms $X \to \{0,1\}$ with an additional structure and then finding appropriate conditions for a map $\mathrm{Hom}(X,\{0,1\}) \to \mathrm{Hom}(Y,\{0,1\})$ to be the pullback of a unique morphism $Y \to X$. For example in the discrete case, we get the following basic set-theoretic observation:

Let $X,Y$ be sets. A map $\mathrm{Hom}(X,\{0,1\}) \to \mathrm{Hom}(Y,\{0,1\})$ is the pullback of a morphism iff it preserves arbitrary suprema and finite infima if we endow $\mathrm{Hom}(X,\{0,1\}) \cong \mathcal{P}(X)$ with the partial order given by inclusion.

Question. Is there a similar statement for the functor $\mathrm{Hom}(-,\{0,...,n\})$, where $n \geq 1$? That is can we endow these sets with additional structure such that we can find a criterion when a map $\mathrm{Hom}(X,\{0,...,n\}) \to \mathrm{Hom}(Y,\{0,...,n\})$ is the pullback of a map $Y \to X$?

First observe that there is a natural isomorphism between this functor and $\mathcal{P}^{[n]}$, which is defined by $\mathcal{P}^{[n]}(X) := \{(A_1,...,A_n) : A_1 \subseteq ... \subseteq A_n \subseteq X\}$ on objects and by preimages on morphisms. It is a partial order (inclusion in each entry) with suprema and infima. But it also has various operations (endofunctors) coming from automorphisms of $\{0,...,n\}$. For example, the permutation $i \mapsto n - i$ yields the operation $K : (A_1,....,A_n) \mapsto (A_n^c,...,A_1^c)$ and the cycle $(0 ~ 1)$ yields the operation $H : (A_1,...,A_n) \to (A_2 \setminus A_1,A_2,...,A_n)$.

For $n = 2$ I can show that a map $\mathcal{P}^{[n]}(X) \to \mathcal{P}^{[n]}(Y)$ is a pullback iff it preserves suprema, finite infima and commutes with $H$ and $K$ (for $n=1$ this is also true, we don't even need $H$ or $K$). Remark that $H^2=K^2=(HK)^3=1$ and that $K$ transforms suprema into infima and vice versa. Is this a known structure? If not, it is possible to endow $\mathrm{Hom}(X,3)$ with a more well-known structure?

For $n>2$ I don't think that $H$ and $K$ suffice. But I guess that the whole set of endofunctors of $\mathcal{P}^{[n]}$ suffices.

share|cite|improve this question
Try looking at the edit to Todd's answer to…. It may be relevant. – Benjamin Steinberg Jan 3 '12 at 14:06
I've looked at it, I think it is not related. – Martin Brandenburg Jan 3 '12 at 14:35
When I see this, I think "Priestly duality". However, I can't recall the topic or some of the recent workers in it, except that one of them (Brian Davey?) used to write songs about the conferences he attended. I suspect the path you suggest has been trodden. Gerhard "Ask Me Not About Dualities" Paseman, 2012.01.03 – Gerhard Paseman Jan 3 '12 at 14:51
There is a book about natural dualities in universal algebra. – Benjamin Steinberg Jan 3 '12 at 20:24
Do you mean "Natural dualities for the working algebraist" by Clark and Davey? – Martin Brandenburg Jan 3 '12 at 20:51

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.