The Alexander Horned Sphere: is it nonsense?

Modern topology is full of contentious issues, but no-one seems to pay any notice. There are many weird, even absurd, “constructions” and “arguments” which really ought to generate vigorous debate. People should have differences of opinions. Alternatives ought to be floated. The logical structure of the entire enterprise ought to be called into question.

But not in these days of conformity and meekness, amongst pure mathematicians anyway. Students are indoctrinated, not by force of logic, clarity of examples and the compelling force of rigorous computations, but by being browbeaten into thinking that if they confess to “not understanding”, then they are tacitly admitting failure. Why don’t you understand? Don’t you have what it takes to be a professional pure mathematician?

Let’s have a historically interesting example: the so-called “Alexander Horned Sphere”. This is supposedly an example of a “topological space” which is “homeomorphic”… actually do you think I could get away with not putting everything in quotes here? Pretty well everything that I am now going to be talking about ought to be in quotes, okay?

Right, so as I was saying, the Alexander Horned sphere is supposedly a topological space which is homeomorphic to a two-dimensional sphere. It was first constructed (big quotation marks missing on this one!) by J. W. Alexander in 1924, who was interested in the question about whether it was possible for the complement of a simply-connected surface to not be simply connected.

Simply-connected means that any loop in the space can be continuously contracted to a point. The two-dimensional sphere is simply connected, but the one-dimensional sphere (a circle) is not. Alexander’s weird construction gives a surface which is topologically a two-sphere, but its complement is like the complement of a torus: if we take a loop around the main body of the sphere, then we cannot contract it to a point. And why not? Because there is a nested sequence, an infinitely nested sequence of entanglements that our contracting loop can’t get around.


This image was made by Ryan Dahl, Creative Commons license.

Here is a way of imagining what is (kind of) going on. Put your two arms in front of you, so that your hands are close. Now with both hands, make a near circle with thumb and index finger, almost touching, but not quite, and link these two almost loops. Now imagine each of your fingers/thumbs as being like a little arm, with two new appendage finger/thumb pair growing from the end of each, also almost enclosing each other. And keep doing this, as the diagram suggests better than I can explain.

At any finite stage, none of the little almost loops is quite closed, so we could still untangle a string that was looped around say one of your arms, just by sliding it off your arm, past the finger and thumb, around the other arms finger and thumbs, and also navigating around all the little fingers and thumbs that you have grown, something like Swamp Thing.

Yes…but Alexander said “Let’s go to infinity!” And most of the topologists chorused” Yes, let’s go to infinity!” And most of their students dutifully repeated: “Yes, let’s go to infinity, … I guess!” And lo… there was the Alexander Horned Sphere!

But of course, it doesn’t really make sense, does it? Because it blatantly contravenes a core Law of Logic, in fact the one we enunciated two days ago, called the Law of (Logical) Honesty:

Don’t pretend that you can do something that you can’t.

The construction doesn’t work because it requires us to grow, or create, or construct, an infinite number of pairs of littler and littler fingers, and you just can’t do that!! All that we can logically contemplate is a finite version, as shown actually in the above diagram. And for any finite version, the supposed property that Alexander thought he constructed disintegrates.

The Alexander Horned Sphere: but one example of the questionable constructs that abound in modern pure mathematics.


16 thoughts on “The Alexander Horned Sphere: is it nonsense?

  1. Saada

    oh my god. You actually succeded explaining a complete novice, why modern topology is blatant nonsense. What a deed, as far as I am that novice …

  2. merulusfrindens

    Professor Wildberger, I would like to see some constructions that you think can actually be made with regard to topology. Also, given you are so knowledgeable about the history of Mathematics, I would be interested to know where your ideas hail from. Something of where you stand, for example, with respect to, say, Kronecker or existence proofs in the style of Hilbert.

  3. Gerhard X. Ritter

    Thus, using Wildberger’s understanding of mathematics the possible convergence of an infinite series, the Cantor set, the integral in calculus, etc are all nonsense. Fake math, as Trump would say.

  4. Anonymous

    I’m not bothered by an object with an infinite amount of arms. If we let fractals exist, let’s let it exist. However, why should it be homeo to a sphere? It seems like making a homeo map from the real line to a circle. PS I enjoy your Youtube videos

  5. Robert

    There was one of these drawn on a greenboard in one of the Maths Inst corridors when I was at University in the 70s. There was also a splendid steam locomotive drawn on a board nearby. Apparently the artist returned every few years to put back the details that accidentally got rubbed out.
    Don’t know the fate of the horned sphere.

    BTW, you are nothing like Trump, nor Boris Johnson et al. Please don’t stop.

  6. Grinsekotze

    Okay, so your argument is that the Horned Sphere is nonsense… because you say so.
    You are just stating your finitistic view as fact, which is exactly what you are criticizing in other mathematicians’ work. Maybe you should read up on the philosophy of mathematics, instead of pretending you solved it.

      1. Anonymous

        Is the irony lost on you that ‘clearly’ is the word you so condemn in the 2nd paragraph?

  7. Courtney anderson

    I absolutely love this. I was very interested in the horned ball but couldn’t find anyone who thought it was nonsense aswell. Amazingly written.

  8. Paul Counter

    Trying to simplify your argument so that my little brain can understand it, you seem to be saying that if it is not possible to actually make a thing then it should not be relevant to topology. Is that correct?

  9. Cargo

    Hmm… My intuition says that if a compact sphere is mapped in a bijective, continuous and convergent manner, then also the limit map is a homeomorphism.

    Just calculate and the faith will follow 🙂


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s