when we have a covering $p:Y\to X$ choose $x\in X$ and we have two actions on the fiber $p^{-1}(x)$:

1) The monodromy action of $\pi_1(X,x)$ defined as follows: $[\gamma].\tilde x=\tilde \gamma (1)$ that is; the action gives the endpoint of the unique lift of $\gamma$ starting at $\tilde x$.

2)The Deck group action of the group of Deck transformations $Deck(Y,X)$ on $p^{-1}(x)$ defined as follows: $\phi. \tilde x=\phi(\tilde x)$

When $Y=\tilde X$ is the universal covering we have that the two groups $Deck(\tilde X,X)$ and $\pi_1(X,x)$ are isomorphic.

Now my problem is the following : when defining homology of $X$ with local coefficients in a $\mathbb Z[\pi_1(X)]-$module $A$, many authors say that $\pi_1(X,x)$ acts on $\tilde X$ by Deck transformations and set $C_n(X,A):=C_n(\tilde X,\mathbb Z)\otimes A$

so first why they don't say $\pi_1(X)$ acts on $\tilde X$ by monodromy action as defined above instead of saying that it acts by Deck transformatioins and second why taking the universal cover when they could take any cover of $X$ and we still have monodromy and Deck actions.

  • $\begingroup$ For you second question: The action will fail to be proper so the quotient might not be Hausdorff. For the first question: You get the same answer as you noted. $\endgroup$ – Moishe Kohan Jul 15 at 14:48
  • $\begingroup$ @MoisheKohan where does the quotient $\tilde X/\pi_1(X,x)$ appear in this construction ? $\endgroup$ – palio Jul 15 at 15:46
  • $\begingroup$ It appears as the base of a fibration. But what I wrote refers to a different quotient. As an extreme case, take $X=\tilde{X}$. $\endgroup$ – Moishe Kohan Jul 15 at 15:56
  • $\begingroup$ @MoisheKohan In Hatcher page 72 : the action of the Deck transformation group is properly discontinuous. $\endgroup$ – palio Jul 15 at 16:08
  • 1
    $\begingroup$ Sure, but you are proposing to use the action of $\pi_1(X)$ which is not the same as the deck group, unless you have the universal covering space. Incidentally, your monodromy does not even define an action of $\pi_1(X)$ unless you have a regular covering. I suggest that you work out some examples when $\tilde{X}$ is not simply connected and see what you get. $\endgroup$ – Moishe Kohan Jul 15 at 16:17

Following @MoisheKohan comments, I think I see the answer to my questions:

Given a covering map $p:Y\to X$, the Deck transformations group $Deck(Y,X)$ acts on $Y$ and this action restricts to an action on each fiber $p^{-1}(x)$. The monodromy action however is an action of $\pi_1(X,x)$ on the fiber $p^{-1}(x)$ and this action need not be defined on the whole space $Y$ unless the covering $p:Y \to X$ is a regular covering.

When the covering $Y$ is the universal covering $\tilde X$, we have that $\pi_1(X,x_0)$ is isomorphic to $Deck(\tilde X,X)$ and that's why we talk about the action of $\pi_1(X,x)$ on $\tilde X$ by Deck transformations.


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.