## My 2018 Mathematics A To Z: Commutative

Today’s A to Z term comes from Reynardo, @Reynardo_red on Twitter, and is a challenge. And the other A To Z posts for this year should be at this link.

# Commutative.

Some terms are hard to discuss. This is among them. Mathematicians find commutative things early on. Addition of whole numbers. Addition of real numbers. Multiplication of whole numbers. Multiplication of real numbers. Multiplication of complex-valued numbers. It’s easy to think of this commuting as just having liberty to swap the order of things. And it’s easy to think of commuting as “two things you can do in either order”. It inspires physical examples like rotating a dial, clockwise or counterclockwise, however much you like. Or outside the things that seem obviously mathematical. Add milk and then cereal to the bowl, or cereal and then milk. As long as you don’t overfill the bowl, there’s not an important different. Per Wikipedia, if you’re putting one sock on each foot, it doesn’t matter which foot gets a sock first.

When something is this accessible, and this universal, it gets hard to talk about. It threatens to be invisible. It was hard to say much interesting about the still air in a closed room, at least before there was a chemistry that could tell it wasn’t a homogenous invisible something, and before there was a statistical mechanics that it was doing something even when it was doing nothing.

But commutativity is different. It’s easy to think of mathematics that doesn’t commute. Subtraction doesn’t, for all that it’s as familiar as addition. And despite that we try, in high school algebra, to fuse it into addition. Division doesn’t either, for all that we try to think of it as multiplication. Rotating things in three dimensions doesn’t commute. Nor does multiplying quaternions, which are a kind of number still. (I’m double-dipping here. You can use quaternions to represent three-dimensional rotations, and vice-versa. So they aren’t quite different examples, even though you can use quaternions to do things unrelated to rotations.) Clothing is a mass of things that can and can’t be put on first.

We talk about commuting as if it’s something in (or not in) the operations we do. Adding. Rotating. Walking in some direction. But it’s not entirely in that. Consider walking directions. From an intersection in the city, walk north to the first intersection you encounter. And walk east to the first intersection you encounter. Does it matter whether you walk north first and then east, or east first and then north? In some cases, no; famously, in Midtown Manhattan there’s no difference. At least if we pretend Broadway doesn’t exist.

Also if we don’t start from near the edge of the island, or near Central Park. An operation, even something familiar like addition, is a function. Its domain is an ordered pair. Each thing in the pair is from the set of whatever might be added together. (Or multiplied, or whatever the name of the operation is.) The operation commutes if the order of the pair doesn’t matter. It’s easy to find sets and operations that won’t commute. I suppose it’s for the same reason it’s easier to find rectangular rather than square things. We’re so used to working with operations like multiplication that we forget that multiplication needs things to multiply.

Whether a thing commutes turns up often in group theory. This shouldn’t surprise. Group theory studies how arithmetic works. A “group”, which is a set of things with an operation like multiplication on it, might or might not commute. A “ring”, which has a set of things and two operations, has some commutativity built into it. One ring operation is something like addition. That commutes, or else you don’t have a ring. The other operation is something like multiplication. That might or might not commute. It depends what you need for your problem. A ring with commuting multiplication, plus some other stuff, can reach the heights of being a “field”. Fields are neat. They look a lot like the real numbers, but they can be all weird, too.

But even in a group, that doesn’t have to have a commuting multiplication, we can tease out commutativity. There is a thing named the “commutator”, which is this particular way of multiplying elements together. You can use it to split the original group in the way that odds and evens split the whole numbers. That splitting is based on the same multiplication as the original group. But its domain is now classes based on elements of the original group. What’s created, the “commutator subgroup”, is commutative. We can find a thing, based on what we are interested in, which offers commutativity right nearby.

It reaches further. In analysis, it can be useful to think of functions as “mappings”. We describe this as though a function took a domain and transformed it into a range. We can compose these functions together: take the range from one function and use it as the domain for another. Sometimes these chains of functions will commute. We can get from the original set to the final set by several paths. This can produce fascinating and beautiful proofs that look as if you just drew a lattice-work. The MathWorld page on “Commutative Diagram” has some examples of this, and I recommend just looking at the pictures. Appreciate their aesthetic, particularly the ones immediately after the sentence about “Commutative diagrams are usually composed by commutative triangles and commutative squares”.

Whether these mappings commute can have meaning. This takes us, maybe inevitably, to quantum mechanics. Mathematically, this represents systems as either a wave function or a matrix, whichever is more convenient. We can use this to find the distribution of positions or momentums or energies or anything else we would like to know. Distributions are as much as we can hope for from quantum mechanics. We can say what (eg) the position of something is most likely to be but not what it is. That’s all right.

The mathematics of finding these distributions is just applying an operator, taking a mapping, on this wave function or this matrix. Some pairs of these operators commute, like the ones that let us find momentum and find kinetic energy. Some do not, like those to find position and angular momentum.

We can describe how much two operators do or don’t commute. This is through a thing called the “commutator”. Its form looks almost playfully simple. Call the operators ‘f’ and ‘g’. And that by ‘fg’ we mean, “do g, then do f”. (This seems awkward. But if you think of ‘fg’ as ‘f(g(x))’, where ‘x’ is just something in the domain of g, then this seems less awkward.) The commutator of ‘f’ and ‘g’ is then whatever ‘fg – gf’ is. If it’s always zero, then ‘f’ and ‘g’ commute. If it’s ever not zero, then they don’t.

This is easy to understand physically. Imagine starting from a point on the surface of the earth. Travel south one mile and then west one mile. You are at a different spot than you would be, had you instead travelled west one mile and then south one mile. How different? That’s the commutator. It’s obviously zero, for just multiplying some regular old numbers together. It’s sometimes zero, for these paths on the Earth’s surface. It’s never zero, for finding-the-position and finding-the-angular-momentum. The amount by which that’s never zero we can see as the famous Uncertainty Principle, the limits of what kinds of information we can know about the world.

Still, it is a hard subject to describe. Things which commute are so familiar that it takes work to imagine them not commuting. (How could three times four equal anything but four times three?) Things which do not commute either obviously shouldn’t (add hot water to the instant oatmeal, and eat it), or are unfamiliar enough people need to stop and think about them. (Rotating something in one direction and then another, in three dimensions, generally doesn’t commute. But I wouldn’t fault you for testing this out with a couple objects on hand before being sure about it.) But it can be noticed, once you know to explore.

## The Summer 2017 Mathematics A To Z: Jordan Canonical Form

I made a mistake! I thought we had got to the end of the block of A To Z topics suggested by Gaurish, of the For The Love Of Mathematics blog. Not so and, indeed, I wonder if it wouldn’t be a viable writing strategy around here for me to just ask Gaurish to throw out topics and I have two weeks to write about them. I don’t think there’s a single unpromising one in the set.

# Jordan Canonical Form.

Before you ask, yes, this is named for the Camille Jordan.

So this is a thing from algebra. Particularly, linear algebra. And more particularly, matrices. Matrices are so much of linear algebra that you could be forgiven thinking they’re all of linear algebra. The thing is, matrices are a really good way of describing linear transformations. That is, where you take a block of space and stretch it out, or squash it down, or rotate it, or do some combination of these things. And stretching and squashing and rotating is a lot of what you’d ever want to do. Refer to any book on how to draw animated cartoons. The only thing matrices can’t do is have their eyes bug out huge when an attractive region of space walks past.

Thing about a matrix is if you want to do something with it, you’re going to write it as a grid of numbers. It doesn’t have to be a grid of numbers. But about all the matrices anyone does anything with are grids of numbers. And that’s fine. They do an incredible lot of stuff. What’s not fine is that on looking at a huge block of numbers, the mind sees: huh. That’s a big block of numbers. Good luck finding what’s meaningful in them. To help find meaning we have a set of standard forms. We call them “canonical” or “normal” or some other approving term. They rearrange and change the terms in the matrix so that more interesting stuff is more obvious.

Now you’re justified asking: how can we rearrange and change the terms in a matrix without changing what the matrix is? We can get away with doing this because we can show some rearrangements don’t change what we’re interested in. That covers the “how dare we” part of “how”. We do it by using matrix multiplication. You might remember from high school algebra that matrix multiplication is this agonizing process of multiplying every pair of numbers that ever existed together, then adding them all up, and then maybe you multiply something by minus one because you’re thinking of determinants, and it all comes out wrong anyway and you have to do it over? Yeah. Well, matrix multiplication is defined hard because it makes stuff like this work out. So that covers the “by what technique” part of “how”. We start out with some matrix, let me imaginatively name it $A$. And then we find some transformation matrix for which, eh, let’s say $P$ is a good enough name. I’ll say why in a moment. Then we use that matrix and its multiplicative inverse $P^{-1}$. And we evaluate the product $P^{-1} A P$. This won’t just be the same old matrix we started with. Not usually. Promise. But what this will be, if we chose our matrix $P$ correctly, is some new matrix that’s easier to read.

The matrices involved here have to follow some rules. Most important, they’re all going to be square matrices. There’ll be more rules that your linear algebra textbook will tell you. Or your instructor will, after checking the textbook.

So what makes a matrix easy to read? Zeroes. Lots and lots of zeroes. When we have a standardized form of a matrix it’s nearly all zeroes. This is for a good reason: zeroes are easy to multiply stuff by. And they’re easy to add stuff to. And almost everything we do with matrices, as a calculation, is a lot of multiplication and addition of the numbers in the matrix.

What also makes a matrix easy to read? Everything important being on the diagonal. The diagonal is one of the two things you would imagine if you were told “here’s a grid of numbers, pick out the diagonal”. In particular it’s the one that goes from the upper left to the bottom right, that is, row one column one, and row two column two, and row three column three, and so on up to row 86 column 86 (or whatever). If everything is on the diagonal the matrix is incredibly easy to work with. If it can’t all be on the diagonal at least everything should be close to it. As close as possible.

In the Jordan Canonical Form not everything is on the diagonal. I mean, it can be, but you shouldn’t count on that. But everything either will be on the diagonal or else it’ll be one row up from the diagonal. That is, row one column two, row two column three, row 85 column 86. Like that. There’s two other important pieces.

First is the thing in the row above the diagonal will be either 1 or 0. Second is that on the diagonal you’ll have a sequence of all the same number. Like, you’ll get four instances of the number ‘2’ along this string of the diagonal. Third is that you’ll get a 1 above all but the row above first instance of this particular number. Fourth is that you’ll get a 0 in the row above the first instance of this number.

Yeah, that’s fussy to visualize. This is one of those things easiest to show in a picture. A Jordan canonical form is a matrix that looks like this:

 2 1 0 0 0 0 0 0 0 0 0 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 3 1 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 4 1 0 0 0 0 0 0 0 0 0 0 0 4 1 0 0 0 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 -2 1 0 0 0 0 0 0 0 0 0 0 0 -2

This may have you dazzled. It dazzles mathematicians too. When we have to write a matrix that’s almost all zeroes like this we drop nearly all the zeroes. If we have to write anything we just write a really huge 0 in the upper-right and the lower-left corners.

What makes this the Jordan Canonical Form is that the matrix looks like it’s put together from what we call Jordan Blocks. Look around the diagonals. Here’s the first Jordan Block:

 2 1 0 0 0 2 1 0 0 0 2 1 0 0 0 2

Here’s the second:

 3 1 0 3

Here’s the third:

 4 1 0 0 4 1 0 0 4

Here’s the fourth:

 -1

And here’s the fifth:

 -2 1 0 -2

And we can represent the whole matrix as this might-as-well-be-diagonal thing:

 First Block 0 0 0 0 0 Second Block 0 0 0 0 0 Third Block 0 0 0 0 0 Fourth Block 0 0 0 0 0 Fifth Block

These blocks can be as small as a single number. They can be as big as however many rows and columns you like. Each individual block is some repeated number on the diagonal, and a repeated one in the row above the diagonal. You can call this the “superdiagonal”.

(Mathworld, and Wikipedia, assert that sometimes the row below the diagonal — the “subdiagonal” — gets the 1’s instead of the superdiagonal. That’s fine if you like it that way, and it won’t change any of the real work. I have not seen these subdiagonal 1’s in the wild. But I admit I don’t do a lot of this field and maybe there’s times it’s more convenient.)

Using the Jordan Canonical Form for a matrix is a lot like putting an object in a standard reference pose for photographing. This is a good metaphor. We get a Jordan Canonical Form by matrix multiplication, which works like rotating and scaling volumes of space. You can view the Jordan Canonical Form for a matrix as how you represent the original matrix from a new viewing angle that makes it easy to recognize. And this is why $P$ is not a bad name for the matrix that does this work. We can see all this as “projecting” the matrix we started with into a new frame of reference. The new frame is maybe rotated and stretched and squashed and whatnot, compared to how we started. But it’s as valid a base. Projecting a mathematical object from one frame of reference to another usually involves calculating something that looks like $P^{-1} A P$ so, projection. That’s our name.

Mathematicians will speak of “the” Jordan Canonical Form for a matrix as if there were such a thing. I don’t mean that Jordan Canonical Forms don’t exist. They exist just as much as matrices do. It’s the “the” that misleads. You can put the Jordan Blocks in any order and have as valid, and as useful, a Jordan Canonical Form. But it’s easy to swap the orders of these blocks around — it’s another matrix multiplication, and a blessedly easy one — so it doesn’t matter which form you have. Get any one and you have them all.

I haven’t said anything about what these numbers on the diagonal are. They’re the eigenvalues of the original matrix. I hope that clears things up.

Yeah, not to anyone who didn’t know what a Jordan Canonical Form was to start with. Rather than get into calculations let me go to well-established metaphor. Take a sample of an unknown chemical and set it on fire. Put the light from this through a prism and photograph the spectrum. There will be lines, interruptions in the progress of colors. The locations of those lines and how intense they are tell you what the chemical is made of, and in what proportions. These are much like the eigenvectors and eigenvalues of a matrix. The eigenvectors tell you what the matrix is made of, and the eigenvalues how much of the matrix is those. This stuff gets you very far in proving a lot of great stuff. And part of what makes the Jordan Canonical Form great is that you get the eigenvalues right there in neat order, right where anyone can see them.

So! All that’s left is finding the things. The best way to find the Jordan Canonical Form for a given matrix is to become an instructor for a class on linear algebra and assign it as homework. The second-best way is to give the problem to your TA, who will type it in to Mathematica and return the result. It’s too much work to do most of the time. Almost all the stuff you could learn from having the thing in the Jordan Canonical Form you work out in the process of finding the matrix $P$ that would let you calculate what the Jordan Canonical Form is. And once you had that, why go on?

Where the Jordan Canonical Form shines is in doing proofs about what matrices can do. We can always put a square matrix into a Jordan Canonical Form. So if we want to show something is true about matrices in general, we can show that it’s true for the simpler-to-work-with Jordan Canonical Form. Then show that shifting a matrix to or from the Jordan Canonical Form doesn’t change whether the thing we’re interested in is true. It exists in that strange space: it is quite useful, but never on a specific problem.

Oh, all right. Yes, it’s the same Camille Jordan of the Jordan Curve and also of the Jordan Curve Theorem. That fellow.

## A Leap Day 2016 Mathematics A To Z: Quaternion

I’ve got another request from Gaurish today. And it’s a word I had been thinking to do anyway. When one looks for mathematical terms starting with ‘q’ this is one that stands out. I’m a little surprised I didn’t do it for last summer’s A To Z. But here it is at last:

## Quaternion.

I remember the seizing of my imagination the summer I learned imaginary numbers. If we could define a number i, so that i-squared equalled negative 1, and work out arithmetic which made sense out of that, why not do it again? Complex-valued numbers are great. Why not something more? Maybe we could also have some other non-real number. I reached deep into my imagination and picked j as its name. It could be something else. Maybe the logarithm of -1. Maybe the square root of i. Maybe something else. And maybe we could build arithmetic with a whole second other non-real number.

My hopes of this brilliant idea petered out over the summer. It’s easy to imagine a super-complex number, something that’s “1 + 2i + 3j”. And it’s easy to work out adding two super-complex numbers like this together. But multiplying them together? What should i times j be? I couldn’t solve the problem. Also I learned that we didn’t need another number to be the logarithm of -1. It would be π times i. (Or some other numbers. There’s some surprising stuff in logarithms of negative or of complex-valued numbers.) We also don’t need something special to be the square root of i, either. $\frac{1}{2}\sqrt{2} + \frac{1}{2}\sqrt{2}\imath$ will do. (So will another number.) So I shelved the project.

Even if I hadn’t given up, I wouldn’t have invented something. Not along those lines. Finer minds had done the same work and had found a way to do it. The most famous of these is the quaternions. It has a famous discovery. Sir William Rowan Hamilton — the namesake of “Hamiltonian mechanics”, so you already know what a fantastic mind he was — had a flash of insight that’s come down in the folklore and romance of mathematical history. He had the idea on the 16th of October, 1843, while walking with his wife along the Royal Canal, in Dublin, Ireland. While walking across the bridge he saw what was missing. It seems he lacked pencil and paper. He carved it into the bridge:

$i^2 = j^2 = k^2 = ijk = -1$

The bridge now has a plaque commemorating the moment. You can’t make a sensible system with two non-real numbers. But three? Three works.

And they are a mysterious three! i, j, and k are somehow not the same number. But each of them, multiplied by themselves, gives us -1. And the product of the three is -1. They are even more mysterious. To work sensibly, i times j can’t be the same thing as j times i. Instead, i times j equals minus j times i. And j times k equals minus k times j. And k times i equals minus i times k. We must give up commutivity, the idea that the order in which we multiply things doesn’t matter.

But if we’re willing to accept that the order matters, then quaternions are well-behaved things. We can add and subtract them just as we would think to do if we didn’t know they were strange constructs. If we keep the funny rules about the products of i and j and k straight, then we can multiply them as easily as we multiply polynomials together. We can even divide them. We can do all the things we do with real numbers, only with these odd sets of four real numbers.

The way they look, that pattern of 1 + 2i + 3j + 4k, makes them look a lot like vectors. And we can use them like vectors pointing to stuff in three-dimensional space. It’s not quite a comfortable fit, though. That plain old real number at the start of things seems like it ought to signify something, but it doesn’t. In practice, it doesn’t give us anything that regular old vectors don’t. And vectors allow us to ponder not just three- or maybe four-dimensional spaces, but as many as we need. You might wonder why we need more than four dimensions, even allowing for time. It’s because if we want to track a lot of interacting things, it’s surprisingly useful to put them all into one big vector in a very high-dimension space. It’s hard to draw, but the mathematics is nice. Hamiltonian mechanics, particularly, almost beg for it.

That’s not to call them useless, or even a niche interest. They do some things fantastically well. One of them is rotations. We can represent rotating a point around an arbitrary axis by an arbitrary angle as the multiplication of quaternions. There are many ways to calculate rotations. But if we need to do three-dimensional rotations this is a great one because it’s easy to understand and easier to program. And as you’d imagine, being able to calculate what rotations do is useful in all sorts of applications.

They’ve got good uses in number theory too, as they correspond well to the different ways to solve problems, often polynomials. They’re also popular in group theory. They might be the simplest rings that work like arithmetic but that don’t commute. So they can serve as ways to learn properties of more exotic ring structures.

Knowing of these marvelous exotic creatures of the deep mathematics your imagination might be fired. Can we do this again? Can we make something with, say, four unreal numbers? No, no we can’t. Four won’t work. Nor will five. If we keep going, though, we do hit upon success with seven unreal numbers.

This is a set called the octonions. Hamilton had barely worked out the scheme for quaternions when John T Graves, a friend of his at least up through the 16th of December, 1843, wrote of this new scheme. (Graves didn’t publish before Arthur Cayley did. Cayley’s one of those unspeakably prolific 19th century mathematicians. He has at least 967 papers to his credit. And he was a lawyer doing mathematics on the side for about 250 of those papers. This depresses every mathematician who ponders it these days.)

But where quaternions are peculiar, octonions are really peculiar. Let me call a couple quaternions p, q, and r. p times q might not be the same thing as q times r. But p times the product of q and r will be the same thing as the product of p and q itself times r. This we call associativity. Octonions don’t have that. Let me call a couple quaternions s, t, and u. s times the product of t times u may be either positive or negative the product of s and t times u. (It depends.)

Octonions have some neat mathematical properties. But I don’t know of any general uses for them that are as catchy as understanding rotations. Not rotations in the three-dimensional world, anyway.

Yes, yes, we can go farther still. There’s a construct called “sedenions”, which have fifteen non-real numbers on them. That’s 16 terms in each number. Where octonions are peculiar, sedenions are really peculiar. They work even less like regular old numbers than octonions do. With octonions, at least, when you multiply s by the product of s and t, you get the same number as you would multiplying s by s and then multiplying that by t. Sedenions don’t even offer that shred of normality. Besides being a way to learn about abstract algebra structures I don’t know what they’re used for.

I also don’t know of further exotic terms along this line. It would seem to fit a pattern if there’s some 32-term construct that we can define something like multiplication for. But it would presumably be even less like regular multiplication than sedenion multiplication is. If you want to fiddle about with that please do enjoy yourself. I’d be interested to hear if you turn up anything, but I don’t expect it’ll revolutionize the way I look at numbers. Sorry. But the discovery might be the fun part anyway.