I understand the technical definition of local degree in a theoretical sense via isomorphism of $H_n(U, U_i \backslash \{x_i\}) \cong H_n(S^n)$ etc. However, I'm a bit confused by how exactly this correlates to how local degree is used in the calculation of cellular homologies. For example, I have been studying the example of $\mathbb{RP}^2$ which has the cell structure: $e^0 \cup_{\phi} e^1 \cup_{\varphi} e^2$.

I'd like to calculate the degree of the composition: $$S^{n-1} \xrightarrow{\varphi} X^{n-1} \rightarrow X^{n-1}/X^{n-2} \rightarrow \large \vee_j S^{n-1}\xrightarrow{q_j} S^{n-1}$$ Where $q_j$ is the map which sends all the $S^{n-1}$'s except the $j^\text{th}$ to the base point.

So in our example of $\mathbb{RP^2}$, to calculate the degree of the attaching map $\varphi: \partial D^2 \cong S^1 \rightarrow X^1$ we need to actually know how the space is constructed, not just its cell structure.

The map $\varphi$ will take the circle along the lone 1-cell $e^1$ twice, so to compute its degree via local degrees, I can choose a point in the one-cell* say $y \in interior(e^1)$ and then look at the discrete set $\varphi^{-1}(y) \in S^1$. Since we go along the same 1-cell twice, this will be a set of two points. $$\varphi^{-1}(y) = \{x_1, x_2\} \subset S^1$$

As I understand it, I then should look at how open neighbourhoods around these $x_1, x_2$ "look" compared to the open neighbour hood of $y$. I find that locally $\varphi$ will act just like a map sending a circle to a circle, in particular, it locally looks like either the identity map or the map which reverses the orientation of the circle. In this case, local to both $x_1$ and $x_2$ it will look like the identity which has degree 1.

Therefore, $$\deg{\varphi} = \deg(\varphi\vert_{U/{x_1}}) + \deg(\varphi\vert_{\tilde{U}/{x_2}}) = 1+1 = 2$$

My questions regarding this are the following:

  1. To what extend does my choice of orientation matter, could I not have given $S^1$ the reverse orientation and end up with $\deg(\varphi) = -2$
  2. If the 1-skeleton were to contain two 1-cells, but otherwise the construction was the same and I chose $y$ in this 1-cell which has nothing attached and $\varphi^{-1}(y) = \phi$ is the degree just 0?
  3. If I'm attaching, say, a three cell or even a cell of higher dimension, it becomes less clear how I should envisage this orientation argument. How would one practically cope with this?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.