Prove that the fundamental group of a torus is $\mathbb{Z} \times \mathbb{Z}$ using a fundamental covering map.

How would one go about doing this?? Is there a specific covering map to use? Or some quotient via the universal cover $\mathbb{R} \times \mathbb{R}$?

  • $\begingroup$ The torus is $S^1 \times S^1$. You want to use the product of the standard covering map $\mathbb R \to S^1$. $\endgroup$ – manthanomen Nov 8 '17 at 6:39
  • $\begingroup$ @manthanomen How does this work explicitly? $\endgroup$ – adfafsdafsd Nov 8 '17 at 6:40

It's easier than that. Or, at least, the intuition is. Take the quotient map $q:\Bbb R\times \Bbb R \to T^2$, where $T^2$ is the torus. Then the inverse image $q^{-1}(t)$ of a single point $t\in T^2$ naturally looks like $\Bbb Z\times \Bbb Z$.

There are a few more details to this if you want an actual proof (like how a class $x\in \pi_1(T^2,t)$ is naturally linked to a single point of $q^{-1}(t)$, after you choose an origin, and how this lets you determine the group structure) but this is the basic idea behind it.

Some details on the above parenthesis: The universal cover map is simply connected, which means that any two paths with the same end points are homotopic with a homotopy fixing the end points. If the two end points are both in $q^{-1}(t)$, then the composition of that path with $q$ becomes a loop with end point $t$.

If we have a loop $\gamma$ in $T^2$ with end point $t$, and choose a specific starting point $t_0\in q^{-1}(t)$, then $\gamma$ is the image of exactly one path in $\Bbb R^2$ that starts at $t_0$. (And if we choose a different starting point, the result is that the entire path is translated.)

Think a little on this correspondence between loops in $T^2$ with end point $t$ and paths in $\Bbb R^2$ with end points in $q^{-1}(t)$, and you will hopefully see that two loops represent the same class in $\pi_1(T^2,t)$ iff their corresponding paths in $\Bbb R^2$ with starting point in $t_0$ has the same end point.

Finally, for the group structure, we can see what happens when we compose loops. If we have two loops $\gamma_1,\gamma_2$ with end points $t$, then we can compose them to get $\gamma_3 = \gamma_1 * \gamma_2$, which is also a loop with end point $t$. What happens to their corresponding paths in $\Bbb R$? If we say they all start at $t_0$, and say that $\gamma_i$ ends at $t_i$, then we get $t_3=t_1+t_2-t_0$. This gives it the group structure of $\Bbb Z^2$, with $(0,0)$ set at $t_0$.

  • $\begingroup$ Could you elaborate a bit on the details? I ask this question for my own knowledge, not for any school work. $\endgroup$ – adfafsdafsd Nov 8 '17 at 6:43
  • $\begingroup$ Wherewhat is this result that the inverse image of a quotient covering map determines the fundamental group structure? $\endgroup$ – adfafsdafsd Nov 8 '17 at 6:46
  • $\begingroup$ @adfafsdafsd I added a few paragraphs. $\endgroup$ – Arthur Nov 8 '17 at 7:09

Here is a bit more general way to see the connection between covering spaces and fundamental groups. Suppose $X$ is any path-connected space admitting a universal (i.e. simply-connected) cover $(\overset{\sim}{X},p)$. Then $\pi_1(X)$ is isomorphic to the group of $\textit{deck transformations}$ $\overset{\sim}{X} \to \overset{\sim}{X}$, i.e. homeomorphisms $f: \overset{\sim}{X} \to \overset{\sim}{X}$ such that $p \circ f = p$.

In our case, a universal cover for $T= \mathbb{S}^1 \times \mathbb{S}^1$ is the space $\mathbb{R}^2$ together with $p=(p_1,p_2): \mathbb{R}^2 \to T$ where the $p_i$'s are the covering maps $\mathbb{R} \to \mathbb{S}^1$. It remains to compute deck transformations $\mathbb{R}^2 \to \mathbb{R}^2$. Since $p_1$ and $p_2$ are the covering maps $\mathbb{R} \to \mathbb{S}^1$, any such deck transformation must consist of a pair of deck transformations $\mathbb{R} \to \mathbb{R}$. But the collection of such transformations is precisely $\pi_1(\mathbb{S}^1) \cong \mathbb{Z}$, so we get $\pi_1(T) \cong \mathbb{Z} \times \mathbb{Z}$.


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.