6
$\begingroup$

$\newcommand{\op}{\mathrm{op}}\newcommand{\Un}{\mathrm{Un}}$Lurie's relative nerve functor $\mathrm{N}^{F}_{\bullet}$ is a simpler version of the unstraightening functor $$\Un_\phi:\mathrm{sPSh}(\mathcal{C}_\bullet)\to\mathsf{sSets}_{/S}$$ (or rather of its opposite; see HTT 3.2.5.1) which "unstraightens the homotopy coherence laws of a simplicial presheaf into a left fibration over $S_\bullet$" (I'm paraphrasing the nLab here)

Spelling out HTT 3.2.5.2, one sees that $\mathrm{N}^F_\bullet$ admits the following very simple description: it is the simplicial set where

  • a vertex of $\mathrm{N}^F_\bullet$ is a pair $(A,x)$ with

    • $A$ an object of $\mathcal{C}$;
    • $x$ a vertex of $F(A)$;
  • an edge of $\mathrm{N}^F_\bullet$ from $(A,x)$ to $(B,y)$ is a pair $(f,e)$ with

    • $f:A\to B$ a morphism of $\mathcal{C}$;
    • $e: F(f)(x)\to y$ an edge of $F(B)$;
  • a $2$-simplex of $\mathrm{N}^F_\bullet$ is a $9$-tuple $(f,g,x,y,z,e_{01},e_{12},e_{02},\sigma)$ with

    • $f: A\to B$ a morphism of $\mathcal{C}$;
    • $g: B\to C$ a morphism of $\mathcal{C}$;
    • $x$ a vertex of $F(A)$;
    • $y$ a vertex of $F(B)$;
    • $z$ a vertex of $F(C)$;
    • $e_{01}: F(f)(x)\to y$ an edge of $F(B)$;
    • $e_{12}: F(g)(y)\to z$ an edge of $F(C)$;
    • $e_{02}: F(g\circ f)(x)\to z$ an edge of $F(C)$;
    • $\sigma: e_{12}\circ F(g)(e_{01})\Rightarrow e_{02}$ a $2$-simplex of $F(C)$ as in the diagram $$\require{AMScd} \require{cancel} \def\diaguparrow#1{\smash{\raise.6em\rlap{\!\!\!\!\!\!\scriptstyle #1} \lower.6em{\cancelto{}{\Space{2em}{1.7em}{0px}}}}} \begin{CD} && F(g)(y)\\ & \diaguparrow{F(g)(e_{01})} @VVe_{12}V \\ F(g\circ f)(x) @>>e_{02}> z \end{CD}$$
  • and so on.

Now, in HTT 3.2.5.5, Lurie introduces a functor $\mathfrak{F}_X$ that is (FWIU) supposed to be a simpler analogue of the straightening functor $\mathrm{St}_\phi$. However, Lurie does so by invoking the adjoint functor theorem, similarly to the definition of $\Un_\phi$.

I gather that nevertheless those functors admit nice constructions (as shown for $\Un_\phi$ in Rezk's notes); so, my question is:

Question: What is a concrete description of Lurie's $\mathfrak{F}_X$ functor?

$\endgroup$
1
  • $\begingroup$ This is completely unrelated to your question, but that way to obtain a diagonal arrow in amscd is clever and painful at the same time. $\endgroup$ – fosco Dec 12 '20 at 13:37
8
+200
$\begingroup$

There is indeed a simple, explicit description of the left adjoint to the relative nerve functor. This left adjoint sends a simplicial set over the category $C$, i.e. a morphism of simplicial sets $p \colon X \to C$, to the functor $C \to \mathbf{sSet}$ that sends an object $c \in C$ to the simplicial set $X/c$ defined by the following pullback square of simplicial sets: $\require{AMScd}$ \begin{CD} X/c @>>> C/c\\ @V V V @VV \mathrm{dom} V\\ X @>>p> C \end{CD} A reference for this description of the left adjoint is the paper

Heuts, Gijs; Moerdijk, Ieke. Left fibrations and homotopy colimits. Math. Z. 279 (2015), no. 3-4, 723--744. doi

in which they denote this left adjoint by $r_!$.

Let me take this opportunity to remind everyone that the relative nerve construction appeared way back in the 1980's in §4 of the following paper written by the category theorist John Gray:

Gray, John W. The representation of limits, lax limits and homotopy limits as sections. Mathematical applications of category theory (Denver, Col., 1983), 63--83, Contemp. Math., 30, Amer. Math. Soc., Providence, RI, 1984. doi

$\endgroup$
2
  • $\begingroup$ Shout out to the MSRI library, where I discovered that paper of Gray earlier this year. $\endgroup$ – Alexander Campbell Dec 8 '20 at 6:01
  • 2
    $\begingroup$ Wow, I had no idea this construction went all the way back to the 80's. Also, this description of $\mathfrak{F}_{X}$ is lovely. Thank you very much! $\endgroup$ – Sofia Dec 8 '20 at 7:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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