14
$\begingroup$

I've got a simplicial model category $M.$ I'd like to get my hands on the (infinity,1) category of homotopy limit preserving functors from M to Spaces in order to compare it to another simplicial model category. So it would be convenient if I could have a simplicial model category model for the functor category.

I imagine doing something like the following (sketch):

1) find a model category which models the (infinity,1) category: $\textrm{Fun}(N^{hc}_{\bullet}(M^{cf}),\textrm{SSet})$. I'll call such a model category $\textrm{Fun}(M,\textrm{SSet})$

2) use Bousfield localization on the collection of morphisms S = {for each functor F, the comparison maps F(lim d) ---> lim Fd } in $\textrm{Fun}(M,\textrm{SSet})$ to get a model category structure which models the category of homotopy limit preserving functors.

So my questions are

Question 1: Given a simplicial model category $M$, what model category models the functors from $M$ to Spaces?

Question 2: Given a simplicial model category $M$, what model category models the homotopy-limit-preserving functors from $M$ to Spaces?

Answers to my question don't need to address my sketch, but I am curious about whether that will work. References would also be appreciated.

$\endgroup$

2 Answers 2

7
$\begingroup$

Suppose that the dual $M^{\mathrm{op}}$ of your original simplicial model category $M$ is combinatorial, so that its associated $\infty$-category $\mathcal{M}^{\mathrm{op}}$ is presentable. Then what you are looking at is the $\infty$-category of presheaves of spaces $\mathcal{P}(\mathcal{M}^{\mathrm{op}}) := \operatorname{Fun}(\mathcal{M}, \mathcal{S})$ on $\mathcal{M}^{\mathrm{op}}$, which is again presentable. There are a couple of different model categories that present $\mathcal{P}(\mathcal{M}^{\mathrm{op}})$, one of them being the projective model structure on a category of simplicial presheaves. You can find the details in section 5.1.1 of Lurie's Higher Topos Theory.

The $\infty$-category $\operatorname{Fun}^R(\mathcal{M}, \mathcal{S})$ of limit-preserving functors is equivalent to $\mathcal{M}^{\mathrm{op}}$ itself through the Yoneda embedding $\mathcal{M}^{\mathrm{op}} \to \mathcal{P}(\mathcal{M}^{\mathrm{op}})$; i.e., limit-preserving functors $\mathcal{M} \to \mathcal{S}$ are corepresentable. This is proposition 5.5.2.2 of Higher Topos Theory.

$\endgroup$
5
  • $\begingroup$ This totally answers Question 2. Do you know what model category models the presheaves on M (or M^op) ? Do the projective / injective model structures on Fun(M^op, SSet) do this? How can I see that? Also, do you know how to answer Question 2 if instead of commuting with all limits, the functors only commute with limits of a certain shape (ie, fix the diagram category for the limit)? P.s. I don't really know the etiquette on mathoverflow. Should I select this answer and then post new questions for the ones in this comment? $\endgroup$
    – Joey Hirsh
    Dec 26, 2012 at 21:59
  • $\begingroup$ @Joey: I edited the answer to include the concrete reference to the model structures that present $\mathcal{P}(\mathcal{M}^{\mathrm{op}})$. I don't know how to characterize functors that preserve only certain limits; of course, corepresentable ones do, but there are more. With respect to MO etiquette, you should mark the checkbox next to an answer if it does indeed answer your question. You can ask for clarifications in comments; for follow-up questions I think it's better to post them as new questions. In particular, the characterization of functors that preserve only certain limits might... $\endgroup$ Dec 26, 2012 at 23:33
  • $\begingroup$ ...make an interesting question, and it would hopefully attract the attention of an expert. $\endgroup$ Dec 26, 2012 at 23:34
  • 2
    $\begingroup$ Done! mathoverflow.net/questions/117304/… $\endgroup$
    – Joey Hirsh
    Dec 27, 2012 at 23:54
  • $\begingroup$ @JoeyHirsh Since it looks like you are still/again active, it might be nice if for posterity you would unaccept this answer, which Ilan corrected a few years ago. $\endgroup$ Jan 8, 2020 at 4:24
7
$\begingroup$

Even if the dual $M^{\mathrm{op}}$ of your original simplicial model category $M$ is combinatorial, so that its associated $\infty$-category $\mathcal{M}^{\mathrm{op}}$ is presentable, it is not true that the $\infty$-category of presheaves of spaces $\mathcal{P}(\mathcal{M}^{\mathrm{op}}) := \operatorname{Fun}(\mathcal{M}, \mathcal{S})$ on $\mathcal{M}^{\mathrm{op}}$ is again presentable. This only holds if $\mathcal{M}$ is small which is typically not true for a model category. Also the results of Section 5.1.1 of Lurie's Higher Topos Theory giving a model for the $\infty$-category of presheaves only speak about the small case.

In the case you are interested in one should consider only small functors from $\mathcal{M}$ to $\mathcal{S}$ in order to get an $\infty$-category with small mapping spaces. This is what Lurie does when considering the pro category of spaces as a full subcategory of the $\infty$-category of accessible functors from spaces to spaces. (See Definition 7.1.6.1 in Higher Topos Theory. In this case small is equivalent to accessible.)

You should thus begin with the category of small simplicial functors from $M$ to simplicial sets, and endow it with the projective model structure that was constructed by Chorny and Dwyer: http://arxiv.org/abs/math/0607117. Then you should left Bousfield localize this model structure in such a way that the fibrant objects in the localized model structure would be the projectively fibrant functors that preserve homotopy limits. Note that you cannot use the usual theorems on Bousfield localizations such as those of Hirschhorn or Smith-Lurie because your model category is not cellular nor combinatorial (in fact, it is not even cofibrantly generated). You will have to use techniques such as in http://arxiv.org/abs/1409.8525, which rely on the generalized small object argument due to Chorny.

The localized model structure above would indeed model the $\infty$-category $\operatorname{Fun}^R(\mathcal{M}, \mathcal{S})$ of limit-preserving functors, which is equivalent to $\mathcal{M}^{\mathrm{op}}$ itself through the Yoneda embedding by proposition 5.5.2.2 of Higher Topos Theory, as Alberto García-Raboso mentioned.

$\endgroup$

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.