9
$\begingroup$

Let $\mathsf{Top}$ denote the category of pointed spaces having the pointed homotopy type of a pointed CW-complex. Let $\mathsf{Grp}$ denote the category of groups. It is well documented that for groups $G$ and $H$ the fundamental group functor induces a bijection $$ \pi_1:\left[K(G,1),K(H,1)\right]_\mathsf{Top}\stackrel{\sim}{\to} \operatorname{Hom}_\mathsf{Grp}(G,H), $$ where on the left we have pointed homotopy classes of maps between Eilenberg Mac Lane spaces.

I am interested in the following generalisation. Let $\mathsf{Top}^\to$ denote the arrow category, whose objects are morphisms $X\to Y$ in $\mathsf{Top}$, and whose morphisms are commuting squares. The arrow category $\mathsf{Grp}^\to$ is defined similarly. There is a notion of homotopy in $\mathsf{Top}^\to$ in which a homotopy between morphisms from $p: X\to Y$ to $p':X'\to Y'$ is a commuting square $\require{AMScd}$ \begin{CD} X\wedge I_+ @> \tilde{H}>> X'\\ @V p\times\operatorname{Id} V V @VVp'V\\ Y\wedge I_+ @> H>> Y' \end{CD} Then the fundamental group gives a functor $\pi_1:\mathsf{Top}^\to\to \mathsf{Grp}^\to$ which respects homotopies.

Let $\mathsf{Fib}\subseteq \mathsf{Top}^\to$ denote the full subcategory whose objects are surjective fibrations with path-connected fibres. Let $\mathsf{Epi}\subseteq \mathsf{Grp}^\to$ denote the full subcategory whose objects are group epimorphisms.

Claim: Given $p:K(G,1)\to K(H,1)$ and $p':K(G',1)\to K(H',1)$ objects in $\mathsf{Fib}$, the fundamental group functor induces a bijection $$ \pi_1:\left[p,p'\right]_\mathsf{Fib}\stackrel{\sim}{\to} \operatorname{Hom}_\mathsf{Epi}(\pi_1(p),\pi_1(p')). $$

Question 1: Does this result appear in the literature? If so, where?

Question 2: Do I have the right theoretical framework? (It feels like in instance of something much more general.)

$\endgroup$
4
  • 1
    $\begingroup$ Sounds reasonable to me... it's a little hard to say if it's the "right" theoretical framework without some idea of what your goals are. For instance, in a vacuum I'd probably personally set it up in a way to avoid talking about point-set stuff at all, but that's just a style thing. In terms of generalization, maybe the first step would be to talk about the category of $n \pm 1$-connected [hedging because I'm usually off by one in the connectivity of a map], maps of $n$-connective, $n$-trucnated spaces for various $n$ (where here you have $n=1$). $\endgroup$ Commented Apr 28, 2021 at 16:36
  • 1
    $\begingroup$ @TimCampion: I'd be interested in how to set it up without mentioning point-set issues. I'm a bit of a luddite when it comes to simplicial sets, infinity categories and the like. $\endgroup$
    – Mark Grant
    Commented Apr 28, 2021 at 19:35
  • $\begingroup$ Where does being aspherical in the title plays a role? I don’t see any mention in the body of question. $\endgroup$
    – user51223
    Commented Apr 29, 2021 at 16:41
  • $\begingroup$ @user51223: the spaces in the claim are $K(G,1)$ spaces, i.e. aspherical $\endgroup$
    – Mark Grant
    Commented Apr 29, 2021 at 16:54

0

You must log in to answer this question.