5
$\begingroup$

$ \newcommand{\Ab}{\mathbf{Ab}} \newcommand{\Sp}{\mathbf{Sp}} $In 0-groupoids (sets), the thick subcategories of the category of abelian groups $\Ab$ are given by the primes $p$ and $0$, which we can identify with $\mathbb F_p$ and $\mathbb Q$.

In $\infty$-groupoids, the correct notion of abelian group is "spectrum", and the thick subcategories of $\Sp$ are given by the Morava $K$-theories $K(n,p)$ for primes $p$ and $0\le n \le \infty$, where by convention $K(0,p) = H\mathbb Q$ and $K(\infty,p)=H\mathbb F_p$ for all $p$.

One frequently calls these thick subcategories "primes" or "characteristics", and thinks of them as the points of the "Balmer spectrum" of $\Ab$ or $\Sp$.

I am curious about the story for $n$-groupoids for $0<n<\infty$. I presume for each $n$ there is an appropriate notion of "$n$-abelian group"; for example, this should be a Picard groupoid for $n=1$. Write $\Sp_n$ for whatever this notion turns out to be. I think this is just spectra with homotopy groups concentrated in $[0,n]$, or maybe $[-n, n]$ for non-connective spectra.

What are the thick subcategories of $\Sp_n$?

$\endgroup$
4
  • $\begingroup$ Yes, it should be spectra with homotopy groups in $[0, n]$. $\endgroup$ Apr 10, 2017 at 23:39
  • $\begingroup$ @QiaochuYuan Cool, do you have a reference? Are you also implicitly replacing n-groupoids with something else? $\endgroup$ Apr 11, 2017 at 11:19
  • $\begingroup$ This is a good question, with another good question inside of it! The subtle "inner question" is, I take it, whether there's some algebraic (categorical) notion of Abelian group object in n-categories which models n-truncated connective spectra in the way that Abelian groups model Eilenberg-Mac Lane spectra. This is a stable version of the Homotopy Hypothesis, and is something I've thought about (along with many others). The main, "outer", question is whether one can classify the thick subcategories in such a category, and this is one I haven't come across before but I think is wonderful! $\endgroup$
    – Niles
    Apr 17, 2017 at 19:25
  • $\begingroup$ One pointer to literature: the notion of "Picard category" (a.k.a. "Picard groupoid", a.k.a. "2-abelian group", etc. ) models 1-truncated connective spectra, and this may be a place where one can answer the question about thick subcategories. An initial guess: two cyclic groups of order p (objects and automorphisms) joined by a nontrivial symmetry (modeling the Postnikov invariant). $\endgroup$
    – Niles
    Apr 17, 2017 at 19:32

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.