8
$\begingroup$

For a simplicial complex $X$ with an action of a discrete group $G$, we can impose the following condition, namely that if $g\in G$ stabilizes a given simplex $\sigma\subseteq X$, then $g:\sigma\to\sigma$ is in fact the identity map. This condition is nice because it implies that the quotient $X/G$ is glued out of simplices (rather than simplices modulo a finite group of permutations of their vertices).

Is there a standard term for this condition (or any related one) on the action $G\curvearrowright X$?

$\endgroup$
  • 6
    $\begingroup$ This condition appears in several of my papers, and despite looking I have not found a standard term for it. One potentially useful way of thinking about it is that it is equivalent to saying that you can turn your simplicial complex into a semisimplicial set with a $G$-action. $\endgroup$ – Andy Putman May 27 '19 at 14:45
7
$\begingroup$

I've never stumbled across a standard one in all of the literature, though I've definitely appreciated a remark by Ken Brown in his group cohomology bible: "The hypothesis that $G_\sigma$ fixes $\sigma$ pointwise is not very restrictive in practice. In the case of a simplicial action, for example, it can always be achieved by passage to the barycentric subdivision."

Some references have the pointwise-fixed assumption baked into the definition of a $G$-CW-complex, but not all. Ken Brown's book doesn't, and he calls a $G$-CW-complex "admissible" if we include the pointwise-fixed constraint. This doesn't seem to be standard though. (Now as for definite nonstandard terminology, Farb-Margalit's book "A Primer on Mapping Class Groups" calls it a group action "without rotations".)

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ My vote is to start using "admissible (simplicial) action" often. $\endgroup$ – Chris Gerig Jun 2 '19 at 18:33

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.