If you don't know what a group is, this might be a bit of an unfamiliar read. I recommend checking out some introductory resources on group theory here.
In short, my Tumblr title is the order of the baby monster group $B$ (not to be confused with the BABYMONSTER group). The baby monster group is the second largest sporadic finite simple group, and, as a subquotient of the monster group $M$, is part of the happy family of sporadic groups. It is one of the building blocks of all finite groups, which is why it's so mathematically significant.
While I could give up there and leave a light sketch of the meaning of what I just said, I thought it would be more fun for all of us if I showed you exactly what the finite simple groups are, and why the baby monster group is significant at all.
dense group-theoretic jargon
In order to understand what I'm talking about in the classification of simple groups, there's some complicated definitions related to breaking down a group you should be familiar with.
Some groups contain subgroups within them - that is, a subgroup $H$ of $G$ (written as $H \leq G$)
is a subset of $G$ where that subset is a group itself! (Under the same operation.)
Remember this means:
- $H$ contains all the elements you can get by combinations of its own elements (since it's a group), sort of like a little independent island of elements inside $G$. We say this means the subgroup $H$ is closed under the group operation, because combining elements of $H$ will only ever give you other elements in $H$.
- $H$ has to contain the element $e$, since it's a group.
- Same applies to all the other group properties; these just felt most relevant to mention.
For example, $\mathbb Z$ is a subgroup of $\mathbb R$ under addition because all integers are also real numbers (i.e. it's a subset), and adding integers will always give you other integers (closure).
The trivial group is a subgroup of every group. It's made up of just the identity element, ${e}$. We call it the trivial group because all trivial groups are obviously isomophic.
Sometimes it's useful to take cosets of subgroups. If $H \leq G$, and $g$ is a random element in $G$, we can create the "left coset" $gH$ by taking every element $h \in H$ and left-multiplying it by $g$ to make $gh$.
Since all subgroups are closed, their cosets are either made up of completely different elements or they're exactly the same. This means that if $g'$ is an element in $gH$, $g'H$ and $gH$ are the same coset!
You can also take what are called left and right cosets. A left coset is when you left-multiply by $g$ to get $gH$, and a right coset is when you right-multiply by $g$ to get $Hg$.
Let's look at the dihedral group of order 8 - $D_4$, the symmetry group of a square, which represents all possible orientations you can get from rotating the square clockwise ($r$) and flipping it about the horizontal ($f$):
$$D_4 = \{e, r, r^2, r^3, f, fr, fr^2, fr^3\}$$
One cool thing you might notice is that if you take just the elements $R = \{e, r, r^2, r^3\}$ and multiply them by $f$ (form the coset $fR$) you get the rest of the group $D_4$. In other words, you can actually "factor out" $R$ from $D_4$:
$$ \begin{aligned} D_4 &= \{e, r, r^2, r^3\} \cup \{f, fr, fr^2, fr^3\} \\ D_4 &= R \cup fR \end{aligned} $$
This "factoring" turns $D_4$ into a disjoint union of cosets of our subgroup $R$. (We call $R$ the rotational subgroup of $D_4$ since it's made up of only rotations.) This works for all subgroups, since as we mentioned before all cosets are either disjoint or equal; there's always some way to get the whole supergroup as a set-of-cosets. The number of left cosets you get with this method is called the index of the subgroup.1
Another subgroup is $C = \{e, r^2\}$, since you can factor $D_4$ like this:
$$ D_4 = C \cup rC \cup fC \cup (fr)C $$
so what's a quotient group?
...So, here is the set-of-cosets of $C$:
$$ \begin{aligned} \{C, rC, fC, (fr)C\} \\ = \left\{ \{e, r^2\}, \{r, r^3\}, \{f, fr^2\}, \{fr, fr^3\} \right\} \end{aligned} $$
It turns out that we can "multiply" two cosets together to get another coset by saying that $aC \cdot bC = (ab)C$. For example:
-
$rC \cdot fC = (rf)C = (fr)C$.
-
$(rf)C \cdot fC = (rf^2)C = rC$, since $f^2 = e$).
-
$rC \cdot (rf)C = (r^2f)C = (fr^2)C = fC$, since $r^2f \in fC$; you can work it out for yourself - see what you get when you try to make the coset $(r^2f)C$.
This works no matter how you choose to "label" the cosets. Remember that $r^3C = rC$, since $r^3 \in rC$. Nevertheless, $r^3C \cdot fC = (r^3f)C = (fr^3)C$ still equals $rC$, since $fr^3 \in rC$.
Under this operation, the set-of-cosets is actually a group!!2 We call this group of factored-out cosets the quotient group $D_4/C$.
Let's also check out the reflection subgroup $F = {e, f}$. Its set-of-cosets looks like this:
$$ \begin{aligned} \{F, rF, r^2F, r^3F\} \\ = \{ \{e, r^2\}, \{r, r^3\}, \{f, fr^2\}, \{fr, fr^3\} \} \end{aligned} $$
Let's look at $rF \cdot rF$. One way to do this is to multiply it straight-out; $rF \cdot rF = r^2F$. But, since $r^3 \in rF$, $r^3F = rF$. What happens if we do $r^3F \cdot rF$?
$$r^3F \cdot rF = r^4F = F \neq r^2F$$
Oops. It turns out that our "multiplication" operation is not a well defined operation at all; we got a different result when "multiplying" the same elements! Actually, the set-of-cosets of $F$ doesn't even form a group! This only works for some groups... but why?
Like $C$, some subgroups are more equal than others. Normal subgroups are a special kind of subgroup that give you a little more insight into the overall structure of its supergroup, through this magical quotient group. The important property is that in a normal subgroup, the left cosets and the right cosets produced by the same element are equal - formally, if $N$ is a normal subgroup of $G$ (written as $N \triangleleft G$), then for all $g \in G$, $gN = Ng$.
If we look back at what we did for $D_4/C$, this makes a little bit of sense. To show that $rC \cdot fC = r^3C \cdot fC$ (and therefore the multiplication operation is well defined), we did a bunch of simplifications like $(rf)C = (fr)C$ and $(r^3f)C = (fr^3)C$. Is it just a coincidence that these cosets happened to be the same? Not at all, and it's actually exactly what the definition of a normal subgroup tells us!
But in order to provide a more formal understanding, let's try replicate our setup for $C$ in the general case. We have the set of cosets of $N$ which we'll call $Q$. We have our special multiplication operation defined as
$$aN \cdot bN = (ab)N$$
but, as we saw, this doesn't work for all sets, which means not all subgroups' sets-of-cosets form a group. However, we can show that this is always true for a normal subgroup!
Like we've seen, the coset $aN$ is the same as the coset of any element $xN$ where $x \in aN$, since $N$ is closed. In order for our operation to actually be an operation, it shouldn't matter which $x$ we pick. We can show that if $a$ and $x$ are any two elements in $aN$, and $b$ and $y$ are any two elements in $bN$, $(ab)N = (xy)N$. The freedom of multimedia content has allowed me to attach an animation to demonstrate the little bit of manipulation we have to do.
$$(ab)N = a(bN) = a(yN) = a(Ny) = (aN)y = (xN)y = x(Ny) = x(yN) = (xy)N$$

Make sure you can follow every step: since $y \in bN$, $bN = yN$, so we can swap them out. Since $N$ is a normal subgroup, $yN = Ny$, so we can also switch around the order of our multiplication. This shows that if $N \triangleleft G$, $Q$ is a group!3 We call $Q$ the quotient group of $G$ with respect to $N$, and we write that as $G/N$.
So, if $N \triangleleft G$, then the quotient group $G/N$ is the group formed by all the unique cosets of $N$ under that special multiplication operation. You can think of this as the normal subgroup $N$ representing a key identifable structure within the group, and the quotient group representing how that structure is present throughout the group.4
A really good resource covering normal subgroups and quotient groups is this lovely overview from Wiliams College which you can see I expertly cherrypicked into this very post.
do you even remember what this was about?
And so finally. Simple groups, which I previously very loosely labelled as "the building blocks of a group", are actually very easy to define. A simple group is a group whose only normal subgroups are the trivial group and itself!5 In other words, it has no quotients; it's indivisible. Every (finite) group, if you follow it down, will dissolve into a series of simple groups. In fact, every finite group $G$ can be broken down into a subnormal series:
$$\{e\} = H_0 \triangleleft H_1 \triangleleft H_2 \triangleleft \cdots \triangleleft H_n = G$$
Every finite group actually has a unique composition series of this form if at each step, you take $H_{n-1}$ as the largest normal subgroup of $H_n$.6
As it turns out, almost all simple groups (there are infinitely many) fit into one of four categories:
- Cyclic groups
- Alternating groups (the group of even permutations on $n$ elements)
- Groups of Lie type7
There are exactly 26 exceptions, which we call sporadic groups.
This discovery was actually really, really important. According to Wikipedia, the proof spans "tens of thousands of pages in several hundred journal articles written by about 100 authors, published mostly between 1955 and 2004."
The monster group $M$ is the largest of the 26 sporadic groups; it has $32! \cdot 10! \cdot 4!^2 \cdot 2 \cdot 7 \cdot 13 \cdot 41 \cdot 47 \cdot 59 \cdot 71$ elements. You can see why this is scary - a structure which is apparently fundamental to group theory seems completely random, meaningless and arbitrarily large (?!). Even the construction of the monster group took a long, long time to find, and it's very difficult to define or find a faithful representation.
That said, if you've ever heard of the number $196\ 883$, this is probably why. The monster group can be represented as a "196 883 dimensional commutative nonassociative algebra over the real numbers".
....This number also shows up almost miraculously in non-discrete mathematics. This is what led many mathematicians to see the monster group as a beautiful, transcendental connection in mathematics, but the monstrous moonshine conjecture is way out of the scope of this post. As Conway said, "there's never been any kind of explanation of why [the monster group]'s there, and it's obviously not there just by coincidence. It's got too many intriguing properties for it all to be just an accident." If that quote wasn't dramatic enough, then
$$ \begin{aligned} \text{"I can explain what Monstrous Moonshine is in one sentence, \ it is the voice of God."} \\ \text{- Simon P. Norton} \end{aligned} $$
Moving on, although simple groups have no normal subgroups, they are allowed to have regular old subgroups - and those subgroups have quotient groups. We call those quotient groups subquotients. 19 of the other 25 sporadic groups are actually subquotients of the monster group! We call those groups the happy family because they're all neatly related to each other.
You can picture subquotient relationships of groups in a Hasse diagram, where an arrow $A \rightarrow B$ means $B$ is a subquotient of $A$.
| Jgmoxness, CC BY 4.0, via Wikimedia Commons (original file) |
The black-outlined groups in this diagram are the only sporadic groups not part of the happy family, referred to as the Pariah groups.
As you can see, the baby monster group is a subquotient of the monster group, and 12 of the other 25 sporadic groups are also involved in it as subquotients! It has an order of:
$$ 20! \cdot 6!^2 \cdot 4! \cdot 2^12 \cdot 23 \cdot 31 \cdot 47 \\ = 4\ 154\ 781\ 481\ 226\ 426\ 191\ 177\ 580\ 544\ 000\ 000 $$
Which is the title of my Tumblr blog tree-three. The reason it's the order of the baby monster group specifically is because
- The order of the monster group is so large it doesn't fit within the character limit.
- The baby monster group sounds cuter.
...And that's about all there is to say on the matter. My title is a number related to a mathematical object that is surprisingly significant for rather obscure reasons, but I hope to have shared the joy in that obscurity by my best shot at an explanation of what a sporadic group is.
With all that out of the way though, there's obviously so much more to group theory. To tell you the truth, we're not even done with normal subgroups! (hint hint) In the meantime, I encourage you to check out other resources until I'm back with more on Galois theory!
Continued in Part 2
-
A cool byproduct of this is Lagrange's Theorem (Taylor's version): if you have a finite group $G$ and a subgroup $H \leq G$, then the order of $G$ is equal to the order of $H$ times the index of $H$. This means the order of a subgroup is always a divisor of the order of its supergroup. <
-
The quotient group $D_4/C$ is isomorphic to the Klein group, one of the two only groups on 4 elements. In this group, each element is its own inverse; $rC \cdot rC = C$. The other group it could have been is the cyclic group $C_4$. <
-
#thanksWikipedia (You might want to check out their proof of the converse as well.) <
-
The subgroup $C$ I mentioned before is actually the center of the group $D_4$, which is the set of elements that commute with every other element in the group. The center of a group is always a normal subgroup, which makes sense: a subgroup is normal iff for every $g \in G$ and $n \in N$, $gng^{-1} \in N$ - this is an equivalent condition to the one we used. Since every element $c \in C$ conjugates with every element in $G$, in fact $gcg^{-1} = c$ and so it's normal no matter what, and we can always form a quotient group! <
-
When I say largest normal subgroup, I mean that there is no other normal subgroup of $H_{n}$ that contains $H_{n-1}$. We call this the maximal normal subgroup. <
-
I don't actually know what a Lie group is. Expect a future blog post. <