I am trying to understand the descent condition for sheaves from presheaves. Let Presh(S) be the $(\infty,1)$ category of presheaves on an $(\infty,1)$ site S, and Sh(S) be the corresponding category of sheaves. In PreSh(S), we have morphisms W: Y $\rightarrow$ X, the $(\infty,1)$ sheaves A are those $(\infty,1)$ presheaves in PreSh(S) which are local objects with respect to this morphism in this category; these local objects induce bijection between PreSh(S)(X, A) and PreSh(S)(Y, A) in the homotopy category of $\infty$-Top. Y can be regarded as cover or hypercover with respect to the $(\infty,1)$ presheaf X. Thus, the $(\infty,1)$ sheaf condition descends from Y to X. But, I have a doubt here. X and Y are objects in PreSh(S); I do not get how Y can be a hypercover which is a morphism in the category Simplicial PreSh(S), not PreSh(S). Can someone kindly clarify this? I have just completed my undergrad in physics; apologies if this doubt sounds trivial.
1 Answer
If you start with a hypercover represented as a morphism $Z_\bullet \to c_\bullet X$ in simplicial presheaves, where $c_\bullet X$ denotes the constant simplicial presheaf, then your $Y$ is the colimit (= geometric realization) of $Z_\bullet$. Since the colimit of $c_\bullet X$ is just $X$, the colimit of $Z_\bullet \to c_\bullet X$ gives you your morphism $Y\to X$ of presheaves.
The point of taking the colimit is that mapping out of colimits takes them to limits, so you have $\mathrm{PreSh}(S)(Y,A) \simeq \lim \mathrm{PreSh}(S)(Z_\bullet, A)$, where the latter is the more straightforwardly defined space of "descent data" in $A$ over the hypercover $Z$.
-
$\begingroup$ Thank you! Is there any generalization to the $(\infty,n)$ case? I have seen localization only for $(\infty,1)$. Thanks! $\endgroup$ Commented Dec 30, 2023 at 19:43
-
$\begingroup$ @PinakBanerjee: The case of (∞,n)-categories is a special case of sheaves valued in (∞,1)-categories: simply take the (∞,1)-category of (∞,n))-categories as a target. $\endgroup$ Commented Dec 30, 2023 at 20:38
-
$\begingroup$ You mean internalization n times, right? $\endgroup$ Commented Dec 30, 2023 at 21:00
-
1$\begingroup$ No, he means that instead of sheaves of $\infty$-groupoids, you can consider sheaves of objects of $C$ for any $(\infty,1)$-category $C$, and $C$ could be the $(\infty,1)$-category of $(\infty,n)$-categories. But I'm not sure if that answers your question. Perhaps it would be best to ask it as a new Question so that you can give more background and explanation of exactly what you do mean. $\endgroup$ Commented Dec 31, 2023 at 3:55
-