Design a site like this with
Get started

An attempt to motivate the definition of a topology.

Many people first learn about topology as a generalization of a metric space. We can observe that the open sets of a metric space obey three rules:

  • The entire space is open, as well as the empty set
  • The intersection of finitely many open sets is open
  • The arbitrary union of open sets is open

These are the axioms of a topology. But what if we could motivate these without reference to the behavior of a metric space? This is my attempt to do so by making an argument about what a “neighborhood” should be, and trying to argue that the definitions of a topology capture the behavior we want.

Definitions are meant to emulate observed or desired behavior through mathematical formalism. So the first question to ask would be, “what behavior are we trying to capture?” I think a reasonable (or at least sufficient for now) answer would be homeomorphism. We want a way to say that two things have “the same shape,” which is strict enough to capture properties we want, such as intuitive (non-formal for now) notions of “connectedness,” or “compactness,” but loose enough that geometric information such as size, length, curvature, etc. don’t come into play.

The basic unit of information we will be working with should be able to capture the “topological” information around each point. One way is to talk about “neighborhoods” around each point. Informally, if you google “neighborhood definition,” one result will say “the area surrounding a particular place, person, or object.” So we want to think about an object by using information about what it means for two parts of it to be “in a neighborhood” of each other. By talking about these neighborhoods, perhaps we will be able to capture the information we want by specifying which points are “close to” each other in this sense.

So, with this vague starting point, we next ask ourselves: “without knowing anything else about neighborhoods, what rules should they follow?”

First, we might make the very obvious (but very necessary) observation that a set which can be called a neighborhood of a point must certainly contain that point. That’s a starting point. Closely following this would be the assertion that every point ought to have a neighborhood, even if it’s all alone.

A starting point for a definition would be a way to specify neighborhoods. Our rough draft definition will be:

A topology is a set X, equipped with a function N:X\to P(P(X)), which picks out for each point a collection of subsets of X, which are to be designated that points “neighborhoods.” What rules should this function follow?

Next, we want to think about which sets containing a point should or shouldn’t be a neighborhood of a point. Without reference to any other behavior, one immediate property of sets we can establish is whether or not that set contains a neighborhood of that point. Heuristically, neighborhoods should be somewhat local, so we don’t care how a neighborhood of point a treats point b \neq a. Formally, we can say that being a neighborhood and containing a neighborhood are the same.

In other words, for any point x \in X, if A \subseteq B \subseteq X, and A \in N(x), then B \in N(x).

As a sidenote, I think that at this point we have the power to tell if a point is “isolated” from every other point. Specifically, a point x will be called an isolated point if \{x\} \in N(x). The local behavior around that point is the same as the behavior at only that point: nothing exists immediately around it.

Next, we ask ourselves “how should two neighborhoods of a point interact?” Well, if two sets each contain “an area around a point,” then their intersection should, too. Morally, the property of being “around a point” means that if the intersection of two sets does not contain “an area around the point”, that should mean at least one of those two sets was not a neighborhood. This will be our third axiom:

If A, B \in N(x), then A \cap B \in N(x).

This is a nice list of axioms so far. Is there anything else we want?

Well, we wanted to consider “closeness.” So, we should figure out what it should mean for two points to be “in a neighborhood of each other.” Earlier, I said that a point is isolated if it has a neighborhood consisting only of itself. So, if a point is not isolated, it should at least matter a little bit how the points around it behave. After all, if a set is a neighborhood of x, shouldn’t x have some neighbors?

Let’s say that x_0, x_1 \in X are “neighbors in A” if A \in N(x_0), N(x_1). Clearly this is an equivalence relation. Consider some neighborhood A of x, and let [x] denote the equivalence class of elements which are “neighbors in A” to x. Everyone here are neighbors, so shouldn’t this set be a neighborhood? Formally, we say:

If A \in N(x), then there exists M \subseteq A such that for all y \in M, we have M \in N(y).

To conclude this section, we state all the axioms we came up with again:

\textbf{Definition:} A \textit{topology} on a set X, is a function N:X\to P(P(X)) which satisfies the following axioms:

  • If A \in N(x), then x \in A
  • If A \in N(x), and A \subseteq B \subseteq X, then B \in N(x).
  • If A, B \in N(x), then A \cap B \in N(x).
  • If A \in N(x), then there exists an M \subseteq A such that for any y \in M, M \in N(y).

According to wikipedia, these axioms are due to Felix Hausdorff, whose eponymous spaces are very good indeed. We then come to the following important definition:

Given the above information, a subset A \subseteq X is called \textbf{open} if it is a neighborhood of every point it contains. That is, if for all x \in A, we have A \in N(x).

\textbf{EDIT:} Thanks to my friend @kimmellionaire on Twitter, who pointed out that the above is not quite sufficient! We also need the following axiom:

  • Every point has at least one neighborhood. That is, for every x \in X, N(x) \neq \varnothing

Without this axiom, we wouldn’t necessarily have that the entire space is open. Can you figure out why?

Here is an example of how this captures behavior we want. We can tell if two objects are completely distinct – if we have two open sets which are completely disjoint, then anything which lies in solely in one is surely distinct from anything which lies solely in the other. After all, you can go anywhere in the first without being in a neighborhood of the second, and vice versa. This suggests the following definition:

A topological space X is called \textbf{disconnected} if there exist two disjoint nonempty open subsets whose union is X. Conversely, it should be called \textbf{connected} if it is not disconnected.

We can also formalize the notion of “compactness” in a useful way. Informally, compactness should mean that it is in some sense “small,” but while containing a lot. It should be some more general version of being finite, as that is a notion of “smallness” which is easily distinguishable from things which are not small.

We capture this information as follows. Suppose that we have a topological space X, and an open cover U. That is, a U \in P(P(X)) such that every set in U is open, and the union of all sets in U is the entire space X. If something were to be “compact,” then surely we would not require many sets to cover it. The notion of smallness we have is finiteness, so maybe the thing that should be small is any cover. Or at least, any cover should really just be a small cover with some added bonus sets. We arrive at the following definition:

A topological space X is called \textbf{compact} if for any open cover U \in P(P(X)), there exists a finite U'\subset U such that U' is also an open cover.

In the next post, I will try to justify why the definition of continuty we have is correct. Until then, your exercise is to show that this definition of a topology via neighborhoods is equivalent to the definition via open sets. That is, given a neighborhood function N(x), we can recover a unique collection of open sets \tau, and given that same collection \tau, we can recover a unique neighborhood function N(x), which will be exactly the one we started with.


A proof that there exists an infinite collection of sets whose cartesian product is nonempty.

I stumbled upon this succinct proof while discussing topology with some friends.

Suppose that, for any infinite collection of sets A, the cartesian product, i.e. the collection of all choice functions on A, is empty. I claim this implies Tychonoff’s theorem, which states that the product of any collection of compact topological spaces is a compact topological space.

To see this, we note that Tychonoff’s theorem holds easily for any finite collection of topological spaces. Now, for any infinite collection of compact topological spaces, the product of all of these is empty by assumption, hence compact.

Thus, Tychonoff’s theorem is shown to be true. This implies the axiom of choice, which contradicts our hypothesis.

By Modus Tollens, we thus conclude that there exists some infinite collection of sets, the cartesian product of all of which is nonempty.

Using the Seifert-van Kampen theorem to calculate the fundamental group of the torus, aided by illustrations from mspaint

One of my favorite theorems is the Seifert-van Kampen theorem. It’s a very handy result in algebraic topology which allows us to calculate the fundamental group of complicated spaces by breaking them down into simpler spaces. The version of the theorem I’ll be using here can be stated as follows:

\textbf{Theorem (Seifert-van Kampen):} Let X be a path connected topological space, and let U_1, U_2 be subsets of X such that U_1, U_2 are both open and path connected, U_1 \cap U_2 is path connected, and U_1 \cup U_2 = X. If for some x_0 \in U_1 \cap U_2 we have

\pi_1(U_1, x_0) \cong \langle a_1, \dots, a_n \mid R_1 \rangle

\pi_1(U_2, x_0) \cong \langle b_1, \dots, b_m \mid R_2 \rangle

\pi_1(U_1 \cap U_2, x_0) \cong \langle c_1, \dots, c_j \mid R_3 \rangle

where R_1, R_2, R_3 are a set of relations, then we can calculate the fundamental group of U_1 \cup U_2 = X in the following way. If I_1:U_1\cap U_2\to U_1 and I_2:U_1\cap U_2 \to U_2 are the inclusion maps, then we have

\pi_1(U_1 \cup U_2, x_0) \cong \langle a_1, \dots a_n, b_1, \dots, b_m \mid R_1, R_2, R_4\rangle

where R_4 = \{I_1(c_k) = I_2(c_k) \mid k\in \{1,\dots, j\}\}

\textbf{Important corollary:} If U_1 \cap U_2 is simply connected, then the fundamental group of X is the free product of the fundamental groups of U_1 and U_2. This follows by noting that I_i(C_k) = 1 for all i \in \{1,2\}, k \in \{1,\dots, j\}, and hence R_4 consists only of the trivial relation.

Side note: from now on, because X is path connected, the choice of base point does not affect the isomorphism class (this is a classic result). As such, I will be simplifying notation by referring to the fundamental group of X as simply \pi_1(X).

We will use this to calculate \pi_1(T), there T is the torus, which is realized as the following identification space:

We will let U_1 be a small disk in the center, and we will let U_2 be the complement of U_1, plus a little bit, so the intersection will be an open annulus. Here is an illustration:

We can see that U_1 is simply connected, so \pi_1(U_1) is trivial. Further, because the intersection of U_1 and U_2 is an annulus, it has fundamental group isomorphic to \mathbb{Z}\cong\langle a\mid \emptyset \rangle, where a is a clockwise loop which encircles U_1. We next need to find the fundamental group of U_1, which is the punctured torus.

First, we take the “hole” at the center of the square and “stretch” it out:


Now, as in the statement of the Seifert-van Kampen theorem above, we will let U_1 be defined as follows:

Similarly, here is U_2

So, here is U_1 \cap U_2

We see that U_1 and U_2 are both annuli, and hence have fundamental group \mathbb{Z}. Further, their intersection is in an X shape (as in, shaped like an uppercase X), so is simply connected. Thus, the fundamental group of the punctured torus is the free product of \mathbb{Z} with itself, which is isomorphic to the free group on two generators.

As a side note, we can specify what the generators of this group are. A loop which starts at the bottom left of the square and travels along the left edge is not homotopic to a loop which starts at the bottom left and travels along the bottom edge.

So, back to this:

To summarize so far, we have:

\pi_1(U_1) \cong \langle \mid e = 1 \rangle

\pi_1(U_2)\cong \langle a, b \mid \emptyset \rangle

\pi_1(U_1\cap U_2)\cong \langle c \mid \emptyset \rangle

We know the fundamental groups of all the relevant spaces, so we are almost done. All we have to do is calculate R_4, and we can put all the pieces together. We see that I_1(c) = 1. What is I_2(c)?

Here is what I_2(c) looks like as a loop in U_2. We make the looper wider, pressing it towards the boundary of the square as follows:


The loop is clockwise, so when it reaches the edge, we can trace the loop by going up along the left edge, right along the top edge, down along the right edge, and left along the bottom edge. We noted before that if the fundamental group of the punctured torus is \langle a, b \mid \emptyset \rangle, then a can be a loop starting at the bottom left and traversing the left edge, and b can be a loop starting at the bottom left and traversing the bottom edge. Because this is a quotient space, we can see that a is also the same as starting at the top right corner, going “up” to reach the bottom right corner, and traversing the right edge, and b is the same as starting at the top right corner, going “right” to reach the top left corner, and traversing the top edge.

So, I_2(c) is a loop traversing the edges of the square in a clockwise manner. In other words, I_2(c) can be achieved by traversing a, followed by b, followed by a^{-1}, followed by b^{-1}. So, I_2(c) = aba^{-1}b^{-1} = [a, b].

So, here’s the punchline:

\pi_1(T) \cong \langle a, b, e \mid e = 1, I_1(c) = I_2(c) \rangle \cong \langle a, b \mid ab = ba \rangle

In other words, \pi_1(T) \cong \mathbb{Z}^2!

A Classic Joke Proof

While this may stretch the stated point of this blog, I’m ultimately the one in control, and I’ll decide what gets posted here. And I think it’s really funny and would be fun to write up.

I am nowhere near the first person who has ever told this joke.

\textbf{Claim:} For all n > 2, \sqrt[n]{2}\not\in\mathbb{Q}.

\textbf{Proof:} For real numbers a, b \in \mathbb{R}, such that b \neq 0, suppose that (\frac{a}{b})^n = 2. We have that a^n = 2b^n = b^n + b^n.

By Fermat’s last theorem, we have a, b are not both integers, and hence that \frac{a}{b} cannot be rational (it obviously implies a, b aren’t both integers, but how does it imply that they aren’t both rational? Think on this…).

Unfortunately, I have not yet read and understood Wiles’ proof of FLT, so I do not know if this is a circular argument.

Every set has a minimal superset

I credit Mikey and Atal with this proof.

For any set A which is a subset of a (lets say much larger) set K, a minimal superset of A is a set B\subset K such that A\subset B, and if A \subset C, then B\subset C for any superset C of A.

I claim that any set A has a minimal superset. We see this by using Zorn’s Lemma on the elements of \mathcal{P}(K) which contain A (I denote this collection of sets as K_A). The ordering will be the reverse of the typical: if I, J \in K_A such that I\subset J, then we say I > J. If S_1 < S_2 < \dots is a chain in K_A, then this chain is bounded above by \bigcap_{i=1}^\infty S_i. Thus, by Zorn’s Lemma, there is a maximal element B.

I claim this B is a minimal subset. We see that B must contain A by definition, and further B contains no other elements of K_A. Suppose that C is in K_A. We have that A\subset C and A \subset C\cap B, as both C and B are in K_A. Further, C\cap B \subset B. Because C\cap B contains A and is a subset of B, we are forced to conclude, by maximality of B, that B = C\cap B, meaning that B \subset C. Thus, B is a minimal superset of A.

Of course, this whole proof can be greatly simplified by noting that A itself will fulfill the property of being a minimal superset of A, so this is a proof by immediate example. Hey, the definition doesn’t forbid it!

Showing two easy spaces are not homeomorphic.

This one is my own doing. I will be showing the following:

(0,\infty) is not homeomorphic to [0, \infty)

First, we begin with a definition and a lemma. If X = (A, \tau) is any topological space, we define the \textbf{one point compactifiation} of X as the topological space

(A \cup \{ \infty_X \}, \tau \cup \{B \cup \{ \infty_X \} \mid A/B \text{ is a compact closed subset of X} \})

We have the following lemma:

\textbf{Lemma:} if X, Y are two homeomorphic topological spaces, then their one point compactifications are homeomorphic.

\textbf{Proof:} Because X, Y are homeomorphic, there is a homeomorphism f:X\to Y between them. Denote the one point compactifications of X, Y as \hat{X}, \hat{Y}, respectively. Define \hat{f}:\hat{X}\to\hat{Y} such that f(\infty_X)=\infty_Y, and such that \hat{f}(x) = f(x) for any non-infinite x\in X.

I claim that \hat{f} is a homeomorphism. Suppose that U\subset \hat{Y} is an open subset of \hat{Y} which does not contain \infty_Y. Then, by construction, we have that \hat{f}^{-1}(U) is an open subset of \hat{X} which does not contain \infty_X. Suppose that U\subset \hat{Y} is an open subset of \hat{Y} which does contain \infty_Y. By construction, the complement of U is closed and compact. So, f^{-1}(\hat{Y}/U) is a compact and closed subset of \hat{X} by construction, so \hat{f}^{-1}(U) is the complement of a closed compact subset of X and contains \infty_X, so is open. Similarly, \hat{f}^{-1} is continuous, so we are done.

With that out of the way, it is well known that (0, \infty) is homeomorphic to \mathbb{R}, whose one point compactification is homeomorphic to S^1, the unit circle. However, [0, \infty) is homeomorhic to [0, 1), whose one point compactification is homeomorphic to [0, 1]. I claim the following:

[0, 1] is not homeomorphic to S^1

This can be seen by noting that \pi_1([0, 1], 1)\cong\{0\} while \pi_1(S^1, 1)\cong\mathbb{Z}, and clearly these are not isomorphic. By the functoriality of \pi_1, we have that [0, 1] is not homeomorphic to S^1. By the contraposition of the lemma, we thus see that (0, \infty) and [0, \infty) are not homeomorphic.

Addendum to the previous post

The previous post is very inelegant because it invokes machinery such as the one point compactification (not too absurd for a problem of this nature I suppose) and the fundamental group (which is way overkill). The best way (in my opinion) to solve this problem is to use the following lemma:

\textbf{Lemma:} If X, Y are two homeomorphic topological spaces, with a homeomorphism given by f:X\to Y, and if S is any subspace of X, then S is homeomorphic to f(S) as a subspace of Y, with a homeomorphism given by f|_S

\textbf{Proof:} Let U be any open subset of f(S). By definition, this is of the form U' \cap f(S) for some U' which is open in Y. We have the following:

f|_S^{-1}(U) = f|_S^{-1}(U'\cap f(S)) = f^{-1}(U')\cap f^{-1}(f(S)) = f^{-1}(U')\cap S

The right hand side is open in S, so f|_S is continuous. Similarly, f|_S^{-1} is continuous, so we are done.

Now, to use the lemma: suppose that the two spaces mentioned in the previous post were homeomorphic, and let f be such a homeomorphism. For simplicity, the domain of f shall be [0,\infty). Consider S = (0,\infty) as a subset. We see that f(S) = (0,\infty) / f(0). However, in this case S is connected, but f(S) isn’t, a contradiction.

This seemingly dumb trick can be used to prove that many different spaces are not homeomorphic. I leave the following as exercises to the reader using this.

  1. Show that a figure 8 and a circle are not homeomorphic.
  2. Show that an uppercase X and an uppercase T are not homeomorphic.
  3. Show that S^1 and \mathbb{R} are not homeomorphic (see the previous post) without invoking their fundamental groups.

A (mostly) purely algebraic proof of the infinitude of prime numbers

The following proof was provided/explained to me by @grassmanian on twitter. I thank him again for the support. Unfortunately, neither he nor I have found a way to show that each maximal ideal of \mathbb{Z} contains a unique prime number without using the fact that \mathbb{Z} is a PID, hence the mostly in the title.

We begin with a lemma:

\textbf{Lemma:} If A is an integral domain which has finitely many units, and finitely many maximal ideals, then it is a field.

\textbf{Proof:} It is known that any finite integral domain is a field, so we shall assume that A is infinite. Consider the Jacobson radical of A, denoted Jac(A). This is defined as follows:

\text{Jac}(A) := \{x \mid 1 + xy \text{ is a unit for all }y\in A\}

We observe the following: because A is an integral domain, we have that for any nonzero x\in A, and for any y_1\neq y_2 \in A, 1 + xy_1 \neq 1 + xy_2. A is infinite, so if Jac(A) contains any nonzero elements, then A will have infinitely many units. This is forbidden by hypothesis, so we are forced to conclude that \text{Jac}(A) = \{0\}.

Now, because A is an integral domain (in particular because A is commutative), we have the following equivalence of definitions:

\text{Jac}(A) = \bigcap_{M\in\text{MSpec}(A)}M

Where \text{MSpec}(A) denotes all maximal ideals of A.

By hypothesis, A has only finitely many maximal ideals. That is, \text{MSpec}(A) = \{M_1, M_2, \dots, M_n\} for some n\in \mathbb{N}. Any two maximal ideals are coprime, so we may conclude the following by the Chinese Remainder Theorem for rings:

A = A/(0) = A/(M_1\cap M_2\cap \dots \cap M_n) \cong \prod_{i=1}^n A/M_i

Because each M_i is maximal, we have that each A/M_i is in fact a field, and thus A is a product of fields. However, the product of two or more fields is never an integral domain. By hypothesis A is an integral domain, so we are forced to conclude that there is only one such M_i, meaning that A is indeed a field.

With that lemma out of the way, we proceed.

\textbf{Proof of the main claim:} Because \mathbb{Z} is an integral domain with only finitely many units (\{-1, 1\}), but is not a field, we may conclude that \mathbb{Z} has infinitely many maximal ideals. \mathbb{Z} is a principal ideal domain, so we have that each maximal ideal is of the form (p) for some prime p\in\mathbb{Z}. Hence, there are infinitely many primes.

The CoBook

Well, this is my attempt at a blog. This is going to be a mostly for fun repository of miscellaneous math stuff, so please nobody get angry at me if I say something dumb.

The title is a reference to “The Book,” which is a collection of the most elegant proofs to all math problems that Paul Erdős claimed God was keeping from us. It is also a dumb category theory joke, wherein the prefix “co” denotes the dual of some object. In category theory, this means reversing all the arrows, hence the url.

The content of this blog will mostly be my writeups of “inelegant” solutions to problems I come up with, and occasionally I will post something more substantive. My definition of “elegant” is usually “invokes the least,” so that should tell you something about the kind of things I’ll be posting.

I hope you enjoy this, and if you don’t, I hope you forget about it quickly.

I’m doing this basically out of boredom and a desire to practice my skills at writing things up nicely and cleanly, so don’t expect too much out of me!

I’ll post daily, maybe?