Let $R$ be a local ring, $m$ its maximal ideal and $k:= R/m$ its residue field. Suppose that a finite cyclic group $G= \mathbb{Z}/ m \mathbb{Z}$ has a linear nontrivial action on $R$. Let $R^G$ be a ring of invariant elements of $R$ by this action.

Let $E:= E_R(k)$ be an injective hull of $k$

Question 1 Is $G$ naturally acts on $E$?

Question 2 Is $E$ injective hull of $k$ as a $R^G$-module?

Actually I'm thinking the case where $R$ is the local ring of an isolated hypersurface singularity $(f=0)$ over $\mathbb{C}$ and $f$ is a $\mathbb{Z}_m$-eigenfunction.

  • $\begingroup$ I assume that $I := E$? $\endgroup$ Commented Feb 1, 2012 at 15:28

1 Answer 1


In the case of a hypersurface of dimension $d$, or any Gorenstein singularity of dimension $d$, $E \cong H^d_{\mathfrak{m}}(R)$ (of course, this isomorphism is up to multiplication by a unit). $G$ should act on $H^d_{\mathfrak{m}}(R)$ directly (you should even be able to do this explicitly via Cech cohomology). This should be enough for question 1 in your setting.

For question 2, I'm pretty sure the answer is no in general. In particular, the socle of $E$ as an $R^G$-module, is probably not 1-dimensional. Note the socle of the injective hull of the residue field is always 1-dimensional (see for example Bruns and Herzog's book).

Let me give an example. Suppose $R = \mathbb{C}[x,y]$, $\mathfrak{m}$ is the origin, and $G = \mathbb{Z}/2$ acts on $R$ by multiplying the variables by ${-1}$. Then $R^G = \mathbb{C}[x^2, xy, y^2]$. The socle (elements killed by the maximal ideal) of $H^2_{\mathfrak{m}}(R)$ as an $R$-module is just the Cech class $[1/(xy)]$. The socle as an $R^G$-module however also includes the elements $[1/(x^2y)], [1/(xy^2)], [1/(x^2y^2)]$ since the relevant maximal ideal of $R^G$ is $(x^2, xy, y^2)$.

The point is that the socle as an $R^G$-module is all elements killed by $x^2, xy$ and $y^2$.

  • $\begingroup$ Karl, I think you want $G$ to act by $-1$, not $i$. $\endgroup$ Commented Feb 1, 2012 at 17:06
  • $\begingroup$ Graham, of course you are right. I'll fix it. $\endgroup$ Commented Feb 2, 2012 at 1:21
  • $\begingroup$ Thank you for the answer. Please take a look at my additional question if possible. mathoverflow.net/questions/87319/… $\endgroup$
    – tarosano
    Commented Feb 2, 2012 at 11:09

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.