# Generalizing the coordinate space of a manifold

A manifold is, essentially, something that "looks like" $\mathbb R^n$ locally. That is, a topological space $X$ such that for every $x\in X$ there is a continuous map $\varphi$ from an open set $U_x$ containing $x$ to $\mathbb R^n$.

Suppose I want to consider a possible generalization of this concept, in particular by extending which coordinate spaces I can use. A first naive generalization would be to admit any vector space $V$ instead of $\mathbb R^n$, however, any finite dimensional $V$ is isomorphic to some $\mathbb R^n$, so this is clearly not adding anything new. A possible next attempt would be to admit an arbitrary module $M$ (perhaps over a commutative ring) as the coordinate space.

My question would be: Is this concept one that's been explored? Could it produce an interesting object of study?

• I'm not sure whether this could be a question of enough interest to post on MO as well / instead, or perhaps the answer is trivial. Commented Oct 6, 2017 at 20:45
• What topology would you impose on an arbitrary module? Commented Oct 6, 2017 at 20:53
• @AlexProvost I don't know commutative algebra well enough, but I'd assume there's no guarantee of a nice topology existing on an arbitrary module as there is on $\mathbb R^n$, is this correct? Commented Oct 6, 2017 at 20:59
• Right. The point of using $R^n$ is that its topology is very natural and compatible with its algebraic structure. It's not clear to me how one could come up with a useful generalization the way you suggest. But do note that we can use your vector space idea and extend it to infinite dimensional vector spaces; see for instance Hilbert or Banach manifolds. Commented Oct 6, 2017 at 21:18
• You may also be interested in reading about varieties, schemes, or locally ringed spaces. Commented Oct 6, 2017 at 21:24

As I said in the comments, it is not clear what topology one would put on an arbitrary module in order to do this. The point of using $\mathbb{R}^n$ is that its topology is very natural and compatible with its algebraic structure. But do note that we can use your vector space idea and extend it to infinite dimensional vector spaces; see for instance Hilbert or Banach manifolds.