My Favorite Theorem

My Favorite Theorem

By Kevin Knudson & Evelyn LambScienceMathematics
Download on the App Store

My Favorite Theorem episodes

  • Episode 55 - Rebecca Garcia
    Episode 55 - Rebecca Garcia
    Kevin Knudson & Evelyn Lamb
    Download

    Evelyn Lamb: Hello and welcome to my favorite theorem. Math podcast. I'm one of your hosts Evelyn Lamb. I'm a freelance math and science writer in Salt Lake City, Utah. And this is your other host.

    Kevin Knudson: Hi. I’m Kevin Knudson, professor of mathematics at the University of Florida. How's it going?

    EL: All right, it is a bright sunny winter day today, so I really like—I mean, I'm from Texas originally, so I'm not big on winter in general, but if winter has to exist, sunny winter is better than cloudy winter.

    KK: Sure, sunny winter is great. I mean, it's a sunny winter day in Florida, too, which today means it is currently, according to my watch, 81 degrees.

    EL: Oh, great. Yeah.

    KK: Sorry to rub it in.

    EL: Fantastic. It is a bit cooler than that here.

    KK: I’d imagine so.

    EL: So yeah. Anything new with you?

    KK: No, no. Well, actually so so my I might be going to visit my son in a couple of weeks because he's studying music composition, right? And the the orchestra at his at his university is going to play one of his pieces, and so kind of excited about that.

    EL: Very exciting! Yeah, that's awesome.

    KK: Yeah, but that's about it. Otherwise, you know, just dealing with downed to trees in the neighborhood. Not in our yard, luckily, but yeah, stuff like that. That's it.

    EL: Yeah. Well, we are very happy today to have Rebecca Garcia as a guest. Hi, Rebecca. How are you?

    Rebecca Garcia: Hi, Evelyn. Håfa ådai, I should say, håfa ådai, Evelyn, and håfa ådai, Kevin. Thanks for having me on the program.

    EL: Okay, and what—håfa ådai, did you say?

    RG: Yeah, that's right. That's how we, that's our greeting in Chamorro.

    EL: Okay, so you are originally from Guam, and is Chamorro the name of a language or the name of a group of people, or I guess, both?

    RG: It’s both actually. Yes. That's right. And so Chamorro is the native language in the island. But people there speak English mostly, and as far as I'm able to tell I think I'm the first Chamorro PhD in pure mathematics.

    EL: Well, you’re definitely the first Chamorro guest on our show. I think the first Pacific Island guest also.

    KK: I think that's correct. Yeah.

    EL: So yeah, how did you—so you currently are not in Guam. You actually live in Texas, right?

    RG: I do. I'm a professor at Sam Houston State University, which is in Huntsville, Texas, north of Houston. And I'm also one of five co-directors of the MSRI undergraduate program.

    EL: Oh, nice. That seems like it is a great program. So how did you how how did you get from Guam to Huntsville?

    RG: Oh my goodness. Wow. That is a that is a long, long journey.

    KK: Literally.

    RG: I started out as a as a undergraduate at Loyola Marymount University, and I had the thought of becoming a medical doctor. And so I thought we were supposed to do some, you know, life science or you know, chemistry or biology or something along those lines. And so I started out as one of those majors and had to take calculus and fell in love with calculus and the professors in the math department. And I was drawn to mathematics. And that's how I ended up on the mathematics side. And one of the things that I learned in my undergraduate career was these really crazy math facts about the rational numbers. And so that's one of the things that interested me in mathematics, was just the different types of infinities the concept of countable, uncountable, those sorts of things.

    EL: Yeah, those those seem to be the kinds of facts that draw a lot of people into this rich world of creativity and math that you might not initially think of as related to math when you're going through school. So I think this brings us to your favorite theorem, or at least the favorite theorem you want to talk about today.

    KK: Sounds like it.

    EL: Yeah, so what’s that?

    RG: Yeah. So it’s more, I would say, more of a fun fact of mathematics that the rationals first of all are countable, meaning they are in one-to-one correspondence with the natural numbers. And so you can kind of, you know, label them, there's a first one and a second one in some way, not necessarily in the obvious way. But then, at the same time, they are dense in the real numbers. So that to me, just blows my mind, that between any two real numbers, there's a rational number.

    EL: And yeah, so you can't like take a little chunk of the real line and miss all the rational numbers.

    RG: That’s right.

    KK: Right.

    RG: That to me just blows my mind. Because—and then you just sort of start, you know, your brain just starts messing with you, you know, between zero and one there are infinitely many rational numbers and yet they're still countable. And it just, it just starts to mess with your mind a little bit. Right?

    EL: Yeah. Well, and we were we were talking about this a little bit before and it's this weird thing. Like, yeah, there's, like a countable is like a smallness thing. And yet dense is like, they're, you know, they fill up the whole interval this way. I mean, it is really weird. So where did you first encounter this?

    RG: This was in a class in real analysis. And, yeah, so that's where I started to…I thought I was going to be a functional analyst. I thought I was that's what I wanted to do this. I love real analysis. That didn't happen either. But it was in that class where we were talking about just these strange facts, like the Cantor set: that set is a subset of the reals that is uncountable and yet it’s sparse.

    KK: Totally disconnected, as the topologists say.

    RG: Totally disconnected. There you go. Yeah. Right. And so then all these weird things are happening. And you're just in this world where you thought you understood the real line, and then they throw these things at you like, the reals are dense. I mean, the rationals are dense in the reals, you have these weird uncountable sets that are totally disconnected. What's going on? Yeah, so that's where I started to hear about all these weird things happening.

    KK: Right. So one of two things happens when people learn these things, right? It either blows their minds so much they can't keep going. Or it intrigues them so much that you want to learn more. But not be an analyst. Right?

    RG: [laughing] That’s right. At some point I fell in love with computational algebraic geometry and these Gröbner bases, and how you can really get your hands on some of these things and their applications to combinatorics. So I ended up, I had an algebraist’s heart, but I was exposed to some really good analysts early in my career. And so I was very confused. But I've always, I stay true to my algebraic heart and follow that mostly.

    EL: And so is this a fact that you get to teach to your students now ever?

    RG: So no, this is not, but I do like to talk about the the different infinities and things along those lines. And I like to, before class I come in early, and I'll have a little chat with them about just the fact that—you know, they they don't understand that math is not “done.” So, there's still so much to do. And they have no idea that, you know, there's what, what is research like? What does that mean? And so I talk about open questions. And I bring some of that in the beginning of class. And these concepts that had also drawn me in, about the different kinds of infinities and these weird concepts about the rationals being dense and, you know, just things like that. I do get to talk about it, but it's not in a class that I would teach the material on.

    EL: Yeah, just going back to this idea that you've got the rationals that are dense, so it's this, like, measure zero small set, but it's like everywhere. And then you've got the Cantor set, which is uncountable and sparse. It's like, we've got these various ways of measuring these sets. And you think that they line up in some natural way. And yet they don't. It's just like, you know, the density is measuring a different type of property of the numbers than the measure is.

    RG: That’s exactly.

    EL: And actually, I guess countability is a different thing. Also, I mean, it's, yeah, it's so weird. And it's hard to keep all these things straight. My husband does a lot of analysis and like has, yeah, all of these, like, what kinds of sets are what.

    RG: And what properties they have. And yeah, I don’t have that completely straight.

    KK: This is why I’m a topologist.

    EL: But I mean, topology is like,

    KK: Oh, it's weird too.

    EL: It’s secretly analysis.

    KK: Well…

    EL: Analysis wishes it was topology, maybe.

    KK: So my old undergraduate advisor—who passed away last summer, and I was really sad about that—but he always he always referred to topology as analysis done right.

    EL: Shots fired.

    KK: Which is cheap, of course, right? Because you prove all this stuff in topology Oh, the image of a connected set is connected. Yeah, that's easy now go off to the real line and prove that the connected sets are the intervals. That's the hard part. Right? So yeah, he's being disingenuous, but it was. It's a good line. Right.

    RG: Right.

    EL: So you said that you ran into this, was this an undergraduate class where you first saw these notions of countability and everything?

    RG: Right, it was an undergraduate class where I ran into those notions and I was a junior, well, I guess it was in my second semester as a junior, where we were talking about these strange sets. And that's when I had also thought about going on to graduate school and wanting to do mathematics for the rest of my life. I mean, I was a major by then, of course, but I just didn't know what I was going to do. But it wasn't until then, when I learned about, well, this is this could be a career for you. This may be something you like to do. And of course, this was many, many years ago. And nowadays, you can do so much more with mathematics, obviously. I mean, we know that we can do so much more, I should say. We've always been able to do so much more. We just haven't been able to share that with our students so much. We never really spent the time to let them know there's so many careers and mathematics that one can do. But anyway, at that, at that time I was I was drawn into really thinking about becoming a mathematician, and that was one of the experiences that that made me think that there's so much more to this than than I originally thought.

    EL: Yeah, well, I talk to a lot of people, you know, in my job writing and doing podcasts and stuff about math, and there's so many people who don't realize that, like, math research is a career you can do.

    RG: Right.

    EL: And the more we can share these kinds of “aha” moments and insights, the better and, you know, just show like, well, you can use, you know, kind of the logic and the rules of the game to like, find out these really surprising aspects of numbers.

    RG: Right. And I think also, one of the experiences that I've had as an undergrad that really just sort of sealed the deal—I’m going to go into mathematics—was doing an undergraduate research program as a student. Well wasn't really at the time an undergraduate research program, it was just another summer program. This is many years ago, almost before all of that. And I had the chance to spend a summer just thinking about mathematics at a higher level with a cohort of other students who were like-minded as well, you know. And it was really—it was it was like, “Oh, I can do this for the rest of my life? Like how amazing is that?” And so, I was part of a summer program as an undergrad. And then when I was a graduate student, my lifelong mentor, Herbert Medina, was running a program in Puerto Rico and asked me to be a TA while I was a grad student. And so these were some of the things that led me to do what I do now, working with undergraduates, doing research and mathematics.

    EL: And so that ties in to the MSRI program that you are part of, right?

    RG: Right.

    EL: I guess it I've seen it written like MSRI-UP. So I guess that's undergraduate program?

    RG: Yes. Undergraduate Program. That's right. Yeah. Well, that that's sort of like, a different stage that I'm at now. But yeah, before that, I started my own undergraduate research program together with colleagues in Hawaii, at the University of Hawaii at Hilo. And we ran an undergraduate research program called PURE Math, and that was Pacific Undergraduate Research Experience in Mathematics. And we ran that for five years. And then, and then I ended up moving into the co-director role at MSRI-UP.

    EL: Nice.

    KK: That’s a great program.

    EL: Yeah. So the other thing we like to ask our guests to do, is to pair their theorem with something. You know, just like the right wine can enhance that meal, you know, what would you recommend enjoying the density of the rationals with?

    RG: Well, I did think about this a bit. And one of the things that I think, you know, you think the rationals are dense but they really shouldn't be? So, I think of foods that are dense, but they really shouldn't be, and one of those foods that comes to mind, especially being here in Texas, but also being married to a mathematician who is from Mexico, is tamales. So tamales really should not be dense. They should be fluffy and sumptuous, but here in Texas, you find really dense the most, unfortunately. But it It was strange to also discover that growing up in Guam, we also have our own version of tamales, and a lot of the foods are related in some way to foods from Mexico. So I feel like there's this huge rich connection between myself being from Guam, my husband being Mexican and there's just this strange richness that we share this culture, that I don't know, it just blows my mind too. So the same way that the rational is being dense in the reals blows my mind.

    EL: All right, well, I have to ask more about this tamale like creation from in traditional Guam cuisine. What, is that wrapped in, like, banana leaves or something like that?

    RG: It ought to be, and maybe traditionally it was. I think that nowadays it's not that way. They usually serve it in aluminum foil, and it's made—it's a mixture like tamales. So tamales in Mexico are made with corn, right?

    KK: I was about to ask this. What are they made of in Guam?

    RG: Yeah, yeah. And so in Guam we actually use, like, a rice product.

    EL: Okay.

    RG: It's ground up just like corn. And so instead of corn, we're using rice, and it's flavored in different ways.

    KK: Interesting.

    EL: All right. I have kind of in my mind because I'm more familiar with this like almost, is it kind of like a mochi texture? Because, I mean, that's a rice product, but maybe it's not maybe that's like more gelatinous than this would be.

    RG: Yeah, I guess mochi is really pounded and yeah, so yeah, that's more chewy. I think that the tamal, well, you wouldn't say it like that, but the tamales in Guam are very soft and, gosh, I don't know how to describe it. But it's a very soft textured food.

    KK: I would imagine the rice could be softer, and I mean, corn can get very dense, especially when you start to put lard in it and things like that.

    RG: Yes.

    KK: I mean, it’s delicious.

    RG: It is delicious. And oh my, I can’t get enough tamales. Oh, well.

    KK: Yeah, maybe you can.

    RG: Yeah, I should learn.

    EL: Yeah, well, nice. I unfortunately, we do have a couple restaurants in Salt Lake that are Pacific Island restaurants, but we have more people from Samoa and Tonga here. I don't know if we have a lot of people from Guam here. Yeah, there's actually like a surprising number of like, Samoans who live in Salt Lake. Who knew?

    RG: Right.

    EL: But yeah, it's it's because of like the history of Mormon missionaries.

    KK: That’s what I was gonna say.

    EL: Yeah, the world is very interesting, but yeah I don't know if I've seen this kind of food there. I will just have to, you know, if I'm ever in Huntsville I’ve got to get you to make me some of this. I’m just inviting myself over for dinner now. Hope you don't mind.

    RG: That would be great. It would be wonderful to have you here.

    EL: Is there anything else you'd like to share? We'd like to give our guests a chance to like, share, you know if they've got a website or blog or book or anything, but also if you want to share information about MSRI-UP, application information, anything like that for students? Anything you'd like to share?

    RG: Oh, wow. That's a lot of stuff.

    EL: Yeah, I know. I just rattled off a ton of things.

    RG: Well, yes, I do have, I guess I would like to say for the undergraduate listeners in the audience, please consider applying to our MSRI-UP program, and just in general apply to a research program in the summer. These are paid opportunities for you to expand your mind and do some mathematics in a great environment, and so I highly recommend considering applying for that. And so this is the time right now of course by the time the listeners hear this, I’m sure it will be over, but consider doing some undergraduate research or using your summer wisely.

    KK: I parked cars in the summer in college. I did.

    EL: Well, you never know the connections that might happen though because I was talking to someone one time who basically his big break to get to go to grad school came because, like, somehow he was involved in like parking enforcement somewhere, and some math professor called in to complain about, like, getting a ticket, and one thing led to another and then he ended up in grad school. So really, you never know. Maybe that's not the ideal route to take. There are more direct routes, but yeah, there are many paths.

    RG: Yes, there are. And there's also another, I guess another thing to flag would be, well, contributed to a book that Dr. Pamela Harris and others have put together on undergraduate research. So that just I guess that was just released. I'm not entirely sure now. I think it was accepted, and I don't know if if one is able to purchase it, but if you if you consider working with your students on undergraduate research, this is a great resource to use to get you going, I guess.

    KK: Great.

    EL: Oh, awesome. So this is like a resource for like faculty who want to work with undergraduates? Oh, that's great.

    RG: Yes.

    EL: We will find a link to that and put that in the show notes for people.

    RG: That sounds good.

    EL: Okay, great. Thanks so much for joining us.

    KK: It’s been great.

    RG: Thank you so much.

    On this episode of My Favorite Theorem, we were happy to talk with Rebecca Garcia, a mathematician at Sam Houston State University, about the density of the rational numbers in the reals. Here are some links you might find helpful.

    Her website
    A biography of Garcia for SACNAS
    MSRI-UP 
    A Project-Based Guide to Undergraduate Research in Mathematics, the book she mentioned contributing to

    25 min
  • Episode 54 - Steve Strogatz
    Episode 54 - Steve Strogatz
    Kevin Knudson & Evelyn Lamb
    Download

    Kevin Knudson: Welcome to My Favorite Theorem, a math podcast and so much more. I'm Kevin Knudson, professor of mathematics at the University of Florida, and here is your other host.

    Evelyn Lamb: Hi, I'm Evelyn Lamb. I'm a freelance math and science writer based in Salt Lake City, where it is snowy, but I understand not as snowy as it is for our guest.

    KK: I know, and we've been trying to make this one happen for a long time. So I'm super excited that this is finally going to happen. So today we are pleased to welcome Professor Steve Strogatz. Steve, why don't you introduce yourself?

    Steve Strogatz: Well, wow, thank you. Hi, Kevin. Hey, Evelyn. Thanks for having me on. Yeah, I've wanted to be on the show for a very long time. And I think it's true what Evelyn just said, we have a very big snowstorm here today in not-so-sunny Ithaca, New York, upstate. I just took my dog out for a walk, and the snow was over my boots and going into them and making my feet wet.

    KK: See, I have a Florida dog. She wouldn't know what to do. Actually, we were in North Carolina a few years ago at Christmas, and it snowed, and she was just alarmed. She had no idea what to do. And she's small, too, she just couldn't take it.

    SS: Yeah, well, it would be more like tunneling than running.

    KK: Right.

    EL: Yeah. So we actually met quite a few years ago at this point — actually, I know the exact date because it was, like, two days before my brother's wedding the first time we met because you were on the thesis committee for my sister in law, who is a physicist, many years ago, and so we have this weird, it was when I had just moved to New York to work at Scientific American for the first time. So it was at the very beginning of my life as a math writer. And I remember just being floored by how generous you were with being willing to meet with a nobody like me.

    SS: Well that’s nice.

    EL: At this time when I was first starting.

    SS: But actually, I had a crystal ball, and I knew you were going to become the voice of mathematics for the country, practically. I mean, so I let me brag on Evelyn’s behalf a little bit. If you go on Twitter, you—I wonder if you know this, Kevin, do you know this little factoid I'm going to unreel?

    KK: I bet I do.

    SS: You know where I'm going. On Twitter, if you ask “What mathematician do other mathematicians follow?” I think Evelyn is the number one person the last time I checked.

    KK: She is indeed number one. That's right.

    SS: Yeah.

    EL: I like to say I'm the queen of math, Twitter, although I don't actually like to say this because it feels really weird.

    SS: Well that’s okay. You didn't say it. But yeah, I do remember our meeting that day in my office. And right, it was on this happy occasion of a family, of a wedding. Okay, sorry, I interrupted you, Kevin.

    KK: Oh, I don't know. I was going to say with the Twitter thing. I think you're not far behind, right? Like, aren't you number two, probably?

    SS: I think the last time I looked I was number two.

    KK: Yeah.

    SS: So look at that. Okay, so look at that, the two tweet monsters here.

    KK: And now the funny thing is I'm not even on that list. So here we go.

    SS: Okay. Yeah, well you could catch up. I'm sure you'll be coming right on our heels.

    KK: Maybe. I have over 1000 followers now, but apparently not that many mathematicians. So this is how this goes. Anyway, what weird times we live in, right?

    SS: It's very weird. I mean, I don't know what this can get us, a cup of coffee or what.

    KK: Maybe, maybe. Okay. Let's talk theorems. So Steve, you must have a favorite theorem. What is it?

    SS: Yeah, I have a very sentimental attachment to a theorem and complex analysis called Cauchy’s theorem, or sometimes called Cauchy’s integral theorem.

    KK: Oh, I love that theorem.

    SS: It’s a fantastic theorem. And so I don't know. I mean, I feel like I want to say what I like about it mathematically and what I like about it personally. Does that work?

    EL: Yeah, that’s exactly what we want.

    SS: Well, okay. So then, the scene is, it's my sophomore year of college. Maybe I'll start with the emotional.

    KK: Okay.

    SS: It’s my sophomore year of college. I've just gotten very demoralized in my freshman year, taking the the honors linear algebra course that a lot of universities offer as a kind of first introduction to what college math is really going to be like. You know, a lot of kids in high school have done perfectly well in their precalculus and calculus courses, and then they get to college and suddenly it's all about proofs and abstraction. And it can be—I mean, we sometimes call it a transition course, right? It's a transition into the rigorous world of pure math. And so it was a shock for me. I had a lot of trouble with that course. I couldn't read the book very well, it didn't have pictures. And I'm kind of visual. And so I was always at a loss to figure out what was going on. And being a freshman I didn't have any sense about, why don't I look at a different book, you know, or maybe, maybe I should switch sections. Or I could ask my teaching assistant, or I could go to office hours. I didn't know to do any of that stuff.

    So anyway, this is not my favorite theorem. I was very demoralized after this experience in linear algebra. And then when I took a second semester, also an honors course, that was a rigorous calculus course with the Heine-Borel theorem, and, you know, like, all kinds of—again, no formulas, it was all about, I remember hearing this stuff about “every open cover has a finite subcover,” and I thought, “I want to take a derivative! I can't do anything here. I don't know what to do!” So anyway, after that first year, I thought, “I don't have the right stuff to be a mathematician. And so maybe I'll try physics,” which I also always loved. I say all that as preamble to this complex analysis course that I was taking in sophomore year, which, you know, I still wanted to take math, I heard complex variables might be useful for physics, I thought it would be an interesting course. I don't know. Turned out it was a really great course for me because it really looked a lot like calculus, except it was f(z) instead of f(x).

    KK: Right.

    SS: You know, but everything else was kind of what I wanted. And so I was really happy. I had a great teacher, a famous person actually named Elias Stein.

    KK: Oh.

    SS: So Stein is a well-known mathematician, but I didn't know that. To me, he was a guy who wore Hush Puppies and, you know, had always kind of a rumpled appearance, came in with his notes. And he seemed nice, and I really liked his lectures. But so one day, he starts proving this thing, Cauchy’s theorem, and he draws a big triangle on the board. And he's going to prove that the integral of an analytic function f around this triangle is zero no matter what f is. All he needs is that it's analytic, meaning that it has a derivative in the sense of a function of a complex variable. It's a little more stringent condition—actually a lot more stringent than to say a function of a real variable is differentiable, but I didn't appreciate that at the time. I mean, that's sort of the big reveal of the whole subject.

    KK: Right.

    SS: That this is an unbelievably stringent condition. You can’t imagine how much stuff follows from this innocuous-looking assumption that you could take a derivative, but okay, so I'm kind of naive. Anyway, he says he's going to prove this thing, only assuming that f is analytic on this triangle and inside it. And that's enough. And then, you know, I feel like you don't have enough information, there's nothing to do! So then he starts drawing a little triangle inside the big triangle, and then little triangles inside the little triangle. And it starts making a pattern that today I would call a fractal, though I didn't know it at the time, and he didn't say the word fractal. And actually, nobody ever says that when they're doing this proof. But it’s—right, they don’t—but it's triangles inside of triangles in a self-similar way that doesn't actually play any particular role in the proof, other than it's just this bizarre move, like, What is going on? Why is he drawing these triangles inside of triangles? And by the end, I mean, I won't go into the details of the proof, but he got the whole thing to work out, and it was so magnificent that I started clapping.

    And at that point, every kid in the room whipped their head around to look at me, and the professor looked at me, like what is wrong with you? You know, and yet, I thought, “Wow, why are you guys looking at me?” This was the most amazing theorem and the most amazing proof.” You know, so anyway, to me, it was a very significant moment emotionally because it made me feel that math was, first of all, something I could do again, something I could appreciate and love, after having really been turned off for a year and having a kind of crisis of confidence. But also, you know, aside from any of that, it's just, I think people who know would regard this proof —this is actually by a mathematician named Goursat, a French mathematician who improved on Cauchy’s original proof. Goursat’s proof of Cauchy’s theorem is just one of the great— you know, it's from “The Book” in the words of Paul Erdős, right? If God had a proof of this theorem, it would be this proof. Do you guys have any thoughts about that? I mean, I'm assuming you know what I'm talking about with this theorem and this proof.

    KK: Well, this is one of my favorite classes to teach because everything works out so well. Right? Every answer is zero because of Cauchy’s theorem, or it's 2πi because you have a pole in the middle, right?

    SS: Yeah.

    KK: And so I sort of joke with my students that this is true. But then the things you can do with this one theorem, which does—you’re right, it's very innocuous-looking, you know, you integrate an analytic function on a closed curve, and you get zero. And then you can do all these wonderful calculations and these contour integrals and, like, the real indefinite integrals and all this stuff. I just love blowing students’ minds with that, and just how clean everything is.

    EL: Yeah, I kind of—I feel like I go back and forth a little bit. I mean, like, in my Twitter bio, it does have “complex analysis fangirl.” And I think that's accurate. But sometimes, like you said, it's so many of these, you know, you're you're like teaching it or reading it and you're like, “Oh, this is complex analysis is so powerful,” but in another way, it's like our definition of derivative in the complex plane is so restrictive that like, we're just plucking the very nicest, most well-behaved things to look at and then saying, “Oh, look what we can do when we only look at the very most well-behaved things!” So yeah, I kind of go back and forth, like is it really powerful or are we just, like, limiting ourselves so much in what we think about?

    KK: And I guess the real dirty secret is that when you try to go to two complex variables, all hell breaks loose.

    SS: Ah, see, I've never done that subject, so I don't appreciate that. Is that right?

    KK: I don't, either. Yeah. But I mean, apparently, once you get into two variables, like none of this works.

    SS: Ohhh. But that's a very interesting comment you make there, Evelyn, that—you know, in retrospect, it's true. We've assumed, when we make this assumption that a function is analytic, that we are living in the best of all possible worlds, we just didn't realize we were assuming that. It seems like we're not assuming much. And yet, it turns out, it's enormously restrictive, as you say. And so then it's a question of taste in math. Do you like your math really surprising and really beautiful and everything works out the way it should? Or do you like it thorny and full of rich counterexamples and struggles and paradoxes? And I feel like that's sort of the essential difference between real analysis and complex analysis.

    EL: Yeah.

    SS: In complex analysis, everything you had dreamed to be true is true, and the proofs are relatively easy. Whereas in real analysis, sort of the opposite. Everything you thought was true is actually false. There are some nasty counterexamples, and the proofs of the theorems are really hard.

    EL: Yeah, you kind of have to MacGyver things together. “Yeah, I got this terrible epsilon and like, you know, it's got coefficients and exponents and stuff, but okay, here you go. I stuck it together.

    KK: But but that's interesting, Steve, that this is your favorite theorem because, you know, you're very famous for studying kind of difficult, thorny mathematics, right? I mean, dynamics is not easy.

    SS: Huh, I wouldn't have thought that, that's interesting that you think that. I don't think of myself as doing anything thorny.

    KK: Okay.

    SS: So that's interesting. I mean, yes, dynamical systems in the hands of some practitioners can be very subtle. I mean, those are people who have a taste for those those kinds of issues. I've never been very sophisticated and haven't really understood a lot of the subtleties. So I like my math very intuitive. I’m on the very applied end of the applied-pure spectrum, so that sometimes people will think I'm not really a mathematician at all. I look more like a physicist to them, or maybe even, God forbid, a biologist or something. So yeah, I don't really have much taste for the difficult and the subtle. I like my math very cooperative and surprising. I like—well, not surprising for mathematical reasons, but more surprising for its power to mirror things in the real world. I like math that is somehow tapping into the order in the world around us.

    EL: Yeah, so this it's interesting to me, also that you picked this because, yeah, as you say, you are a very applied mathematician. And I think of complex analysis as a very pure—I actually, I'm trying to not say “pure” math, because I think it's this weird, like, purity test or something. But you know, that like a very theoretical thing. So does it play into your field of research at all?

    SS: Well, uh, not particularly. Yeah. So that's a good question. I mean, I have to say I was a little intimidated by the title of the podcast. If you ask me what's my favorite theorem, the truth is for me, theorems are not my favorite things.

    KK: Okay.

    SS: My favorite things are examples or mathematical models. Like there’s a model in my field called the Kuramoto model after a Japanese physicist Yoshiki Kuramoto. And if you asked me what's my favorite mathematical object, I would say the Kuramoto model, which is a set of differential equations that mirrors how fireflies can get their flashes in sync, or how crickets can chirp in sync, or how other things in nature can self-organize into cooperative, collective oscillation. So that's my favorite object. I've been studying that thing for 30 years. And I suppose there are theorems attached to it, but it's the set of equations themselves and what they do that is my favorite of all. So I don't know, maybe that's my real answer.

    KK: Well, that’s fine. So yeah, it's true. We've had people who've done that in the past, they didn't have a favorite theorem, but they had a favorite thing.

    SS: But still, I mean, I am still a mathematician, part of me is, and I do have theorems that I love, and one of the things I love about Cauchy’s theorem is that in the proof, with this drawing of all the nested triangles inside the big triangle, you end up using a kind of internal cancellation. The triangles touch other triangles except on their common edge, sometimes you're going one way, and sometimes you're going in the opposite direction on that same edge. And so those contributions end up cancelling. And you end up, the only thing that doesn't cancel is what's going on around the boundary. And then that can be sort of pulled all the way into a tiny triangle in the interior, which is where you end up using the local property that is the derivative condition to get everything that you need to prove the result about the big triangle on the outside.

    But the reason I'm going into all that is that this is a principle, this internal cancellation, that is at the heart of another theorem that's been featured on your show, the fundamental theorem of calculus, which uses a telescoping sum to convert what's happening on the boundary to what's happening when you integrate over the interior. This idea of telescoping I think, is really deep. I mean, it's what we use to prove Stokes’ theorem. It's what we would use to prove all the theorems about line integrals. It comes up in topology when you're doing chains and cochains. So this is a principle that goes beyond any one part of math, this idea of telescoping. And I've been thinking I want to write an article, someday (I haven't written it yet) called “Calculus Through the Telescope” or “A Telescopic View of Calculus” or something like that, that brings out this one principle and shows its ramifications for many parts of math and analysis and topology. I think some people get it, people who really understand differential forms and topology know what I'm talking about. But no one ever really told me this, and I feel like maybe it should be mentioned, even though it is well-known to the people who know it.

    KK: Right, it's the air we breathe, right? So we don't we don't think about it.

    SS: I guess, but like, I think there are probably high school teachers, or others who are teaching calculus—like for instance, when I learned about telescoping series in my first calculus course, that's just seems like a trick to find an exact sum of a certain infinite series of numbers. You know, they show you, “Okay, you could do this one because it's a telescoping series.” And it seems like it's an isolated trick, but it's not isolated. This one idea—you can see the two- dimensional version of it in Cauchy’s theorem, and you can see the three-dimensional version of it in the divergence theorem, and so on. Anyway, so I like that. I feel like this idea has tentacles spreading in all directions.

    EL: Yeah. Well, this makes me want to go back and think about that idea more because, yeah, I wouldn't say that I would necessarily have thought to connect it to this many other things. I mean, you did preface your statement with “those who really understand differential forms,” and my dark secret is that the word “form” really scares me. It's a tough one. It's somehow, that was one of those really hard things, when I started doing more, like, hard real analysis. It's like, I feel like I always had to just kind of hold on to it and pray. And you get to the end of it. You're like, “Well, I guess I did it.” But I feel like I never really got that full deep understanding of forms.

    SS: Huh. I don't I don't claim that I have either. I'm reminded of a time I was a teaching assistant for a freshman course for the the whiz kids that—you know, every university has this where you throw outrageous stuff at these freshmen, and then they rise to the occasion because they don't know what you're asking them to do is impossible. But so I remember being in a course, like I say, as a teaching assistant, where it was called A Course in Mathematics for Students of Physics, based on a book by Shlomo Sternberg, at Harvard, and Paul Bamberg, who's a physicist there too, and a very good teacher. And that book tried to teach Maxwell's equations and other parts of physics with the machinery of differential forms and homology and cohomology theory to freshmen. But what was amazing is it sort of worked, and the students could do it. And in the course of teaching it, I came to this appreciation of integrating forms, and how it really does amount to this telescoping sum trick. And, anyway, yeah, it's true, that maybe it's not super widely appreciated. I don't know. I don't know if it is, I don't want to insult people who already know what I'm talking about. But I I do feel like there's a story to tell here.

    KK: Okay. Well, we'll be looking for that.

    EL: Yeah.

    SS: Someday.

    KK: In the New York Times, right?

    SS: Well.

    KK: So another thing we do on this podcast is we ask our guests to pair their theorem with something. And we might have sprung this on you, but you seem to have thought of a solution here. So what have you chosen to paired with Cauchy’s theorem?

    SS: Cubist painting.

    KK: Oh, excellent. Okay. Explain.

    EL: Yeah, tell us why.

    SS: Well, I'm thinking of Cubism. I don’t—look, I don't know much about art. So it might be a dumb pairing. But what I'm thinking is there's a there's a painting. I think it's by Georges Braque of a guy, or maybe it's Picasso. Someone walking down stairs. And maybe it's called a Nude Descending a Staircase, or something like that. You're nodding, do you know what I mean?

    EL: I'm a little nervous about saying, I think it is Picasso, but I'm looking it up on my phone surreptitiously.

    SS: I could try too. For some reason, I'm thinking it's George Braque, but that may be wrong. But so I'll describe the painting I have in my head and it may be totally not—

    EL: No, it’s Marcel Duchamp!

    SS: Oh, it's Marcel Duchamp?

    EL: Yeah.

    SS: And what's the name of it?

    EL: Nude descending a staircase, number two. I think.

    SS: Yeah, that's the one. Would that be considered Cubism?

    EL: Yeah.

    SS: It says according to Wikipedia, it’s widely considered a modernist classic. Okay, I don't know if it's the best example of what I'm thinking. But it's, let me just blow it up and look at it here. So, what hits me about it is it's a lot of straight lines. It's very rectilinear. And you don't see anything that really looks curved like a human form. You know, people are made of curved surfaces, our faces, our cheeks are, you know. What I like is this idea that you can build up curved objects out of lots of things made of straight lines. You know what you can do? mesh refinement on it. For instance, there's an old proof of the area of a circle where you chop it up into lots of pizza-shaped slices, right, and then you add up the areas of all those. And they can be approximated by triangles, and if you make the triangles thin enough, then those slivers can fill out more and more of the area, the method of exhaustion proof for the area of a circle. So this idea that you can approximate curved things with triangles, reminds me of this idea in Cauchy’s theorem that you first you prove it for the triangle, and then later Professor Stein proved the result for any smooth curve by approximating it with triangles, you know, a polygonal approximation to the curve, and then he could chop up the interior into lots of triangles. So I sort of think it pairs with this vision of the human form and it's sinuous descent down. You know, this person is smooth and yet they're being built out of these strange Cubist facets, or other shapes. I mean, think of other Cubist paintings you you represent smooth things with gem-like faceted structures, it sort of reminds me of Cauchy’s theorem.

    KK: Okay, good pairing. Yup.

    EL: Yeah, glad we got to the bottom of that before we made false statements about art on this math podcast.

    SS: Yeah, it may not be the best Cubist example. But what are you gonna do? You invited a mathematician.

    KK: So we also like to let our guests make pitches for things that they're doing. So you have a lot going on. You have a new podcast.

    EL: Yeah, tell us about it.

    SS: Okay. Yeah, thank you for mentioning it. I have a podcast with the confusing name Joy of X. Confusing because I also wrote a book by that name. And before that I had written an article by that name.

    KK: Yes.

    SS: So I did not choose that name for the podcast. But my producer felt like it sort of works for this podcast because it's a show where I interview scientists and mathematicians—in spirit, very similar to what we're doing here. And I talk to them about their lives and their work. And it's sort of the inner life of a scientist, but it could be a neuroscientist, it could be a person who studies astrophysics, or a mathematician. It's anything that is covered by Quanta Magazine. So Quanta Magazine, some of your listeners will know, is an online magazine that covers fundamental parts of math and science and computer science. Really, it's quite terrific. If people haven't read it, they might want to look at it online. It's free. And anyway, so Quanta wanted to start a podcast. And they asked me to host it, which was really fun because I get to explore all these parts of science. I've always liked all of the different parts of science, as well as math. And so yeah, that's the show. It's called the Joy of X where here, X takes on this generalized meaning of the unknown, not just the unknown in algebra, but anything that's unknown, and the joy of doing science and the scientific question. We'll be sure to link to that.

    EL: Yeah.

    KK: Also, I think Infinite Powers came out last year, right? 2019?

    SS: That’s true. Yes, I had a book, Infinite Powers, about calculus. And that was an attempt to try to explain to the general public what's so special about calculus, why is it such a famous part of math. I try to make the case that it really did change the world and that it underpins a lot of modern science and technology as well as being a gateway to modern math. I really do think of it as one of the greatest ideas that human beings have ever come up with. Of course, that raises the question, did we discover it or invent it? But that’s a good one.

    EL: Put that on a philosophy podcast somewhere. We don’t need that on this math podcast.

    SS: Yeah, I don't really know what to say about that. That's a good timeless question. But anyway, yes, Infinite Powers was a real challenge to write because I'm trying to tell some of the history, but I'm not a historian of math. I wanted to really teach some of the big ideas for people who either have math phobia or who took calculus but didn't see the point of it, or just thought it was a lot of, you know, doing one integral after another without really understanding why they're doing it. So it's my love song to calculus. It really is one of my favorite parts of math, and I wanted other people to see what's so lovable and important about it.

    KK: Yeah.

    SS: The book, as I say, was hard because I tried to combine history and applications and big ideas without really showing the math.

    KK: Yeah, that's hard.

    SS: And make it fun to read.

    KK: Right. It is. It's a very good book, though. I did read it.

    SS: Oh, thanks.

    KK: And I enjoyed it quite a bit.

    EL: Well, it is on my table here under a giant pile of books to read, because people need to just stop publishing.

    SS: That’s right.

    EL: There’s too much. We just need to have a year to catch up, and then we could start going again but what's what's

    KK: What’s that Japanese word, sort of the joy of having unread books? [Editor’s note: Perhaps tsundoku, “aquiring reading materials but letting them pile up in one’s home without reading them.”] There's a Japanese concept of like these books that you’ll, well, maybe even never read. But that you should have stacks and stacks of books. Because, you know, maybe you'll read them. Maybe you won't. But the potential is there.

    SS: Nice.

    KK: So I have a nightstand, on the shelf of my nightstand there's probably 20 books there right now, and I haven't read them all. I've read half of them, maybe, but I'm going to read them. Maybe.

    SS: Yeah, yeah.

    KK: Actually, you know, when you were talking about your sort of emotional feelings about Cauchy’s theorem, it reminded me of your—I don't know if it was your first book, but The Calculus of Friendship, about your relationship with your high school teacher.

    SS: Well, how nice of you to mention it.

    KK: Yeah. That was interesting to it, because it reminded me a lot of me, in the sense of, I thought I knew everything too when I was 18. Like, I thought, “Calculus is easy.” And then I get to university and math wasn't necessarily so easy. You know. And so these same sort of challenges, you know?

    SS: Well, I appreciate that, especially because that book is pretty obscure. As far as I know, not many people read it. And it's very meaningful to me because I love my old teacher, Mr. Joffrey, who is now, let’s see, he's 90 years old. And I stayed in touch with him for about 35 years after college, and we wrote math problems to each other, and solutions. And it was really a friendship based on calculus. But over the course of those 35 years, a lot happened to both of us in our lives. And yet, we didn't tend to talk about that. It was like math was a sanctuary for us, a refuge to get away from some of the ups and downs of real life. But of course, real life has a way of making itself, you know, insinuating itself whether you like it or not. And so it's it's that story. The subtitle of the book is “what a teacher and a student learned about life while corresponding about math.” And I sometimes think of it as, like, there's a Venn diagram where there's one circle is people who want to read math books with all the formulas, because I include all the formulas from our letters.

    KK: Yeah.

    SS: And then there's people who want to read books about emotional friendships between men. And if you intersect those two circles, there's a tiny sliver that apparently you're one of the people in it.

    KK: And your book might be the unique book in that in that Venn diagram too.

    SS: Maybe. I don't know. But yeah, so it was it was clear it would not be a big hit in any way. But I felt like I couldn't do any other work until I wrote that book. I really wanted to write it. It was the easiest book to write. It poured out of me, and I would sometimes cry while I was writing it. It was almost like a kind of psychoanalysis for myself, I think, because I did have a lot of guilty feelings about that relationship, which, you know, if you do read the book, anyone listening, you'll see what I felt guilty about, and I deserved to feel guilty. I needed to grow up, and you see some of that evolution in the course of the book.

    KK: Yeah. All right. Anything else you want to pitch? I mean?

    SS: Well, how about I pitch this show? I mean, I'm very delighted to be on here. Really, I think you guys are doing a great thing helping to get the word out about math, our wonderful subject. And so God bless you for doing that.

    KK: Well, this has been a lot of fun, Steve, we really appreciate you taking time out of your snow day. And so now do you have to shovel your driveway?

    SS: Oh, yeah, that may be the last act I ever commit.

    KK: Don’t you still have a teenager at home? Isn't that what they're for?

    SS: My kids, I do have—you know what, that's a good point. I have one daughter who is still in high school and has not left for college yet, so maybe I could deploy her. She's currently making oatmeal cookies with one of her friends.

    KK: Well, that's a useful, I mean that that's helping out the family too, right? I mean,

    SS: They’re both able bodied, strong young women. So I should get them out there and with me, and we could all shovel ourself out. Yeah.

    KK: Good luck with that. Thank you. Thanks for joining us.

    SS: My pleasure. Thanks for having me.

    On this episode of My Favorite Theorem, we were happy to talk with Steve Strogatz, an applied mathematician at Cornell University, about the Cauchy integral theorem. Here are some links you might find helpful.

    Strogatz’s website, which includes links to information about his books and article
    The Joy of X, the podcast he hosts for Quanta Magazine
    The Cauchy integral theorem on Wikipedia
    The Kuramoto model
    Nude Descending a Staircase no. 2 by Marcel Duchamp

    33 min
  • Episode 53 - Ruthi Hortsch
    Episode 53
    Kevin Knudson & Evelyn Lamb
    Download

    Evelyn Lamb: Hello and welcome to My Favorite Theorem, the podcast that was already quarantined. I’m one of your hosts, Evelyn Lamb. I am holed up in my house in Salt Lake City, Utah, where I'm a freelance writer. So, honestly, I have worked in my basement, you know, every day for the past five years, and that hasn't changed. This is your other host.

    Kevin Knudson: Hi, I’m Kevin Knudson, professor of mathematics at the University of Florida, which is open for business…But you can't go to campus.

    EL: Okay.

    KK: Yeah, we moved all of our classes online two weeks ago, I'm just teaching a graduate course this term, so that's sort of easier for me. I feel bad for the people who have to actually lecture and figure out how to do this all at once. My faculty have actually been great. They really stepped up. And, remarkably, I've had very few complaints from students, and I'm the chai,r so you know, they would come to me. And it's just really not—I mean, everybody has really taken the whole thing in stride. A lot of anxiety out there, though, among our students. Really, this is a really challenging time for everybody. And I just encourage my faculty to, you know, be kind to their students and to themselves. So let’s shelter in place and get through this thing, right?

    EL: Yup. Yeah, we had an earthquake a week and a half ago to just, like, shake things up, literally. So it's just like, oh, as if I pandemic sweeping through town was not enough. We'll just literally shake your house for a while.

    KK: Yeah, well, you know, we can go outside. We have a Shelter in Place Order, but it's been 90 degrees every day for the last week. And so you know, I like to go bird watching, but my favorite bird watching spot is a city park, and it's closed. So I have to just kind of sit on my back porch and see what's up. Yeah. Oh, well,

    EL: Well, yes, we're making it through it. And I hope—I mean by the time this is—we have a bit of a backlog in our past episodes, and so who even knows what's going to be happening when this is airing. [Editor’s note: We decided to publish this one out of order, so we actually recorded it pretty recently.] But whatever is happening, I know our guests will be very thrilled to be listening to Ruthi Hortsch! Hi, Ruthi. How are you today?

    Ruthi Hortsch: Hey, I'm managing.

    EL: Yeah.

    RH: It’s a weird time.

    EL: Definitely. So what do you do, and where are you?

    RH: Yeah, so I'm in New York City right now, which is kind of right now the hotbed of lots of new infections. But I've been in my apartment for the last two and a half weeks and haven't really directly been experiencing that.

    I work for an organization called Bridge to Enter Advanced Mathematics. So we're a education nonprofit. We work with low-income and historically marginalized youth. And we're trying to create a realistic pathway for them to become mathematicians, scientists, engineers, programmers.

    We start working with students when they're in middle school and we try to figure out, like, what are the things you need to get you to a place where you'll have a successful STEM career? And so we do a lot of different things, but they all are to that purpose.

    EL: Yeah, and I'm so glad that we have you on the show to talk about this. Because, yeah, I've been thinking like, we really need to get someone from BEAM on here because I think BEAM is just such a great program. My spouse, and I donate to it every year. I mean, obviously not every year, I don't even know how old it is. But you know, we've made that part of our yearly giving, and yeah, I just think it does great work. So, does that have programs in both New York and LA now?

    RH: Yes. So we started in New York City in 2011. And a few years ago, we expanded to LA. So the LA programs are still pretty new. They're building up, kind of starting with students in the first year of contact, and then adding in programming for the older students as that first class gets older. So they now have eighth graders, and that's their oldest class, and they'll continue to add in the ninth grade and the 10th grade program, et cetera, as it goes on. The other kind of exciting thing is, last year, we got a grant from the Gates Foundation. And that grant was to partner with other local programs and other cities to help them build up programs that could do some of the same things we do. So it's not the same comprehensive, really intensive support that we give our students in New York City and LA. But assuming summer camps don't get canceled this summer because of corona, there are going to be day camps in Albuquerque and Memphis that are advised by us.

    EL: Oh, that's so great. Yeah, because that's the one thing about it is that it is so localized and, of course, important places for it to be localized. But, you know, the more the, the wider, the better. So that's awesome. And what's your role there? What do you do?

    RH: Yeah, I have a hard time answering this question. So I work in programs, which is like, I work on things that are directly affecting students. I run one of our summer camps in the summer. So I run a sleepaway camp at Union College, in which students learn proof-based mathematics for the first time. The students at the sleepaway camp are all rising eighth graders, and so they get to learn number theory and combinatorics and group theory. They also do some modeling and programming and stuff.

    During the year I do some managing our other programs team, so supporting other staff. I also do all of our faculty hiring. So certainly we hire a lot of people just for the summer, and most of them are—so we hire college, university students, we hire grad students, we hire professors in various different roles. And I handle all of the, like, hiring people to teach math courses.

    EL: Wow.

    KK: That’s a lot. Are your programs sort of face to face, or are they online? Is it sort of a combination of stuff?

    RH: Yeah, so our summer we run six in-person summer camps each summer. So there's two in upstate New York that are sleepaway, one in Southern California that's sleepaway, and then one day camp in LA and two day camps in New York City. And those are all in-person, face to face. And then during the school year, we also have Saturday classes, which is a mix of life skills and enrichment. And we also do in-person advising. So we have office hours where students can come ask us anything, and then also kind of more intensive. Like, how do you apply to college? How do you get into other summer programs or other STEM opportunities? So most of our programs are face to face. Right now, we've had to cancel a bunch of our year-round stuff. So we don't have Saturday classes right now. We are doing one class for the eighth graders virtually, because we really thought it was critical. And at the moment, we're hoping the summer programs will still run, but it's really hard to say what's going to be going on in two weeks.

    KK: Yeah, well, fingers crossed.

    EL: But as wonderful as it is to talk about BEAM, what we're dying to know is what is your favorite theorem?

    RH: Yeah, so this was actually really fast for me to think of. My favorite theorem is Falting’s theorem. So Falting’s theorem is also actually known as the Mordell conjecture, because Mordell originally conjectured it in the same paper in which he proved Mordell’s theorem, I believe, or at least during the same process of research for him.

    EL: Yeah, and so for longtime listeners, was it Mathilde Lalín who, that was her favorite theorem?

    RH: Mm-hmm.

    EL: Okay, that's right. So we're kind of dovetailing right in.

    RH: Yeah. So Mordell’s theorem is about—so when you look at elliptic curves, they have a finitely-generated abelian group. And Mordell’s theorem is the theorem that proves that it actually is finitely-generated.

    KK: Right.

    RH: So when I say the finitely-generated part, it's actually only looking at the rational points on the curve. So we care about algebraic curves, kind of in general. And then we want to think about, like, how do different algebraic curves behave differently? And because I'm trained as a number theorist, I also specifically care about how many rational points are on that curve and how they behave. So this intersects also with algebraic geometry. And in some sense, this is a statement about how the arithmetic part of the curves—the rational points—interacts with the geometry of it.

    So one thing that people care about a lot in geometry is the notion of a genus. This is one of the ways to classify things. And of course, when you're looking at visual shapes, one way of thinking about the genus is how many holes does it have? So if you're just looking at a shape that’s, like, a big sphere, there's no way of poking a hole through it without actually breaking it apart. And so that has genus zero because there are zero holes. But if you're looking at a doughnut, a torus, that has one hole because there's like one place where you can poke something through. And then you can generalize from there that having more holes is higher genus. And so that's kind of a wishy-washy way of looking at things, and a very visual way. There are ways to define that formally in the algebraic sense, but in the places where both definitions make sense, the definition is the same.

    And so when you look at algebraic curves, we can ask ourselves, how do genus zero curves act differently than genus one curves, act differently than genus two curves, and does that tell us anything about the number of rational points? And so it turns out that with genus zero curves, genus zero curves are actually really just conic sections. So basically the nice lines that you study in like algebra in high school. And those have infinitely many rational points, right? So when I say rational point, you can kind of think of it as being like the points where the components have rational values.

    And genus one curves are actually exactly elliptic curves. So in that case, that's when Mordell’s theorem kicks in and the rational points are this finally generated abelian group. And sometimes they have infinitely many rational points, and sometimes they don't, and it kind of depends on what this algebraic structure, this algebraic group structure, looks like. So that's the most complicated weird point. And for genus two or higher curves, it turns out to be true that there are only finitely many rational points on a genus two or higher curve. And that's the statement of Falting’s theorem.

    EL: Okay, and so I, there's something that I, you know, you hear like genus two or higher. And I always wonder, is there a limit to how high the genus can be of these curves? Or, like, is there a maximum complexity that these curves can have?

    RH: So no. And actually, there's a statement in algebraic geometry that makes it really easy-ish— you know, “ish”— to calculate the genus, which is called Riemann-Roch. And it gives you a relationship between the degree of the equation defining it and the genus. And essentially, the genus grows quadratically with the degree. There's an asterisk on everything I'm saying. It’s mostly true.

    KK: It’s mostly true.

    EL: So if I'm remembering correctly, Mordell’s—let’s see, Mordell’s conjecture, Falting’s theorem—was really important for proving Fermat’s last theorem. Is that correct?

    RH: I don't think so, no. But all of these things are related to each other.

    EL: Okay.

    RH: A lot of the common definitions and theorems that play into all these things, they share a lot, but it's not directly, like, one thing implied the other.

    EL: Okay, yeah.

    RH: In particular, Fermat’s Last Theorem was reduced to a statement about elliptic curves, which is about genus one curves, while Falting’s theorem is really a statement about genus two or higher curves.

    EL: Okay.

    KK: So was this a love at first sight kind of theorem?

    RH: I think no. I think part of the reason that I really started appreciating it was because I had a mentor in undergrad who was really excited about it. And I didn't really understand the full implications and the context, but I was like, “Okay, this mentor I have is really about it, so I'm going to be really about it.”

    And we actually used Falting’s theorem as a black box for the REU project I was working on. So we assumed it was true and then used that to show other things. And then later on in grad school, I had a number of things that I was really interested in that Falting’s theorem was related to. One of the things that I think is really cool that's being researched right now is there’s a bunch of like, tropical geometry that is being studied. And this is, like, relating algebraic verbs to kind of more combinatorial objects. So you can actually translate these lcurves that have a more—I don't want to say analytic, but a smooth structure, and then turning them into a question about, like, counting more straight-edged structures instead.

    One of the things about Falting’s proof of Falting’s theorem is that it's not, it doesn't actually give you a bound. So it tells you that there are only finitely many points, but it doesn't give you a constructive way of saying, like, what does it actually bounded by, the number of finite points? And using tropical geometry, people have been able to make statements about bounds in certain situations, which is really cool.

    KK: Okay, I always like these tropical pictures, you know, because suddenly everything just looks almost like Voronoi diagrams in the plane, these piecewise linear things. So I guess the idea of genus probably still makes sense there in some way, once you define it properly. Right?

    RH: Yeah. And there's a correspondence between, there’s a notion of a tropical curve, which still looks like one of those Voronoi diagrams. There’s an actual correspondence, this curve in classical algebraic geometry gives you this particular diagram.

    EL: Nice. And so you say it was very easy to choose this theorem. So what's your, like, elevator sales pitch for this theorem? Keeping in mind that no one is going to be in an elevator with anyone else anytime soon. We're staying far apart, but you know.

    RH: Yeah. So, I think it’s kind of amazing that geometry can tell you something about the arithmetic of a curve. I think this is what drew me to arithmetic algebraic geometry, that there is this kind of relationship. When you think, okay, arithmetic, geometry, those are totally different fields, people study them in totally different ways, but in fact, it turns out that the geometry of a curve can tell you information about the arithmetic. And that's just bizarre, and also very powerful in that you can make a statement about how many rational solutions there are to an equation using correspondence in geometry.

    The REU project that I worked on actually is a statement that I think is really easy to understand. If you have a rational polynomial, that gives you a function from the rationals to the rationals, right?

    And so you can ask yourself: how many-to-one is that function? How many points gets sent to the same point? And if you look at only rational points, our REU project showed that it can't be more than four-to-one off a finite number points.

    So if you are willing to ignore some finite number of points, then no rational polynomial is ever more than four-to-one.

    KK: Interesting.

    RH: And that feels like a very powerful statement. And it's because we had this hammer of Falting’s theorem to just smash it in the middle.

    KK: That’s really fascinating. So no matter how high the degree it's no more than four-to-one? I wouldn’t have guessed that.

    RH: Off a finite number of points.

    KK: Yeah, sure. Generically. Yeah. Right. Interesting.

    RH: I think the real powerful thing there is that Falting’s theorem comes in.

    KK: Yes.

    RH: Oh, actually, higher degree means high complexity means high genus.

    KK: Okay, cool. So another thing we like to do is ask our guests to pair their theorem with something. So what pairs well with Falting’s theorem?

    RH: Yeah, so this is a maybe a little bit of a stretch, but I've been living in New York City for four years, and I love bagels. They’re definitely one of the best parts of living in New York City. I'm always two blocks away from a really good bagel. Traditionally, bagels are genus one, so it's actually not quite appropriate. You have to, I don't know, do the fancy cut to increase the genus—there’s a way to cut a bagel to get higher genus. But I still think since we're thinking about genuses, we're thinking about complexity of things.

    EL: Yeah. Well, like, you cut the bagel in in half, you know, to get like the cream cheese surface, and then just stick them together and you've got a genus two. Put a little cream cheese on the side. You know?

    RH: Yeah. I mean, if we're cutting holes we can cut as we want.

    EL: That’s true. So, are you more—what do you put on the bagel? What kind of bagel, also, do you prefer?

    RH: Ao I mostly like everything bagels.

    EL: Of course. Yeah. Great bagel.

    RH: There is a weird thing that goes on where some bagel shops put salt on their everything bagel and some don't. And I feel like the salt is important.

    KK: Yeah. Agree.

    EL: As long as it's not too much. Like just the right amount of salt is—

    RH: Yeah. It’s definitely important.

    KK: Well a salt bagel is a pretzel.

    EL: Yes.

    RH: And I don't actually eat cream cheese. So I do eat fish sometimes, but I generally don't eat dairy. And I so I usually get, like, tofu scallion spread. And the tofu spread that gets sold in the bagel shops here is actually really good.

    KK: Well yeah, I'm not surprised. I can't get a decent bagel in Gainesville. I mean, there's a couple of bagel shops, but they're no good.

    RH: Yeah. This is what you get for leaving New York City.

    KK: Right, right.

    EL: Yeah, it's funny, actually one of our quarantine projects we're thinking about is making bagels. I've made bagels one other time. But, yeah.

    KK: It's kind of a nuisance. You know that. That boiling step is really—I mean, it's crucial, but it just takes so much time and space.

    EL: Yeah, I mean, they were not nearly as good as a real bagel shop bagel, but fun to play with.

    KK: Yeah. So what's everyone doing to keep themselves occupied? So far I've got a batch of sauerkraut fermenting. I just started a batch of limoncello that'll be ready in a month. I made scones. Maybe that’s it. Yeah. How about you guys?

    RH: Well, I'm still trying to work 40 hours a week.

    KK: Yeah, I'm doing that too.

    RH: We're still trying to help our students respond to the crisis and helping support them both academically, but holistically also.

    KK: Yeah, it's very stressful.

    RH: And at the moment, we're still doing all of our prep work for the summer, which is a huge undertaking? But when I have free time, I've been cooking more. And I'm actually also working on writing a puzzle hunt.

    EL Ooh, cool. Well if that happens, we'll include a link to that in the show notes—if it's the kind of thing that you can do out of a particular geographical place.

    RH: Yeah, so the puzzle hunt I'm helping write is actually for Math Camp.

    EL: Okay.

    RH: So before I worked for BEAM I worked for Canada-USA Math Camp, and in theory, they're running a camp this summer, and one of the traditional events there is [the puzzle hunt]. I think the puzzle hunt often gets put up after the summer, but I’m not sure.

    EL: Oh, cool. The last thing that I, or library book that I got out from the library—it was actually supposed to be due, like, the day after the library shut down here—was 660 Curries, which is an Indian cookbook that—we don’t really cook meat at home, but it's got, I don't know, maybe a hundred-page section of legume curries and a bunch of vegetable curries, so we've been kind of working through that. We made one last night that was great. It was a mixture of moong dal and masoor dal. Yeah, we’ve been eating a lot of curry, and it just makes my early-this-year plan of, like, “Oh, I want to make more dal, so I've got to go stock up on lentils and rice,” brilliant plan, really has made it a lot easier. So yeah.

    RH: I love dal, and I don't feel like anybody around me ever likes dal as much as I do.

    KK: This is a dal-lover convention right here. It's one of my favorite things to eat. Yeah.

    EL: Oh, yeah. Well, I can recommend, if you get a chance to get 660 Curries, I don't remember if it's called mixed red and lentil dal with garlic and curry leaves, or something like that.

    KK: Yeah, I'm actually making curry tonight, but chicken curry so we'll we'll see.

    EL: Yeah, so other than that, just panicking most of the time. It’s been a big pastime for me.

    RH: I’ve had to, like, ban myself from reading the news in the evening.

    KK: Good call.

    EL: That is very smart.

    RH: I haven’t done a good job keeping to it.

    EL: Yeah, I have not done a good job with my self-control with that. So, I’m really trying to do that. I'm hoping to do some sewing projects too, maybe making some masks that I can leave out for people in the neighborhood to take. Obviously not medical grade, but maybe make people feel a little better.

    KK: So yeah, Ellen, my wife, started doing that yesterday. She made, you know, probably 15 of them yesterday real quick.

    EL: Nice.

    KK: I went to the store yesterday and you know—

    EL: Hopefully it gives people a little peace of mind and maybe decreases droplet transmission.

    KK: Let’s hope.

    EL: I’ve refrained from armchair epidemiology, which I encourage everyone to do. So yeah, I hope everyone stays safe and tries to keep keep a good spirit and help the people in your lives. I hope our listeners can do that too. And I hope they find some enjoyment in thinking about math for a little while with us.

    KK: So yeah, thanks for joining us, Ruthi. We really appreciate it.

    EL: Yeah, everyone go find BEAM online if you want to learn more about that.

    RH: Yeah. Follow us on social media.

    EL: Yeah. So what are the handles for that?

    RH: Yeah, I should have this memorized. You can find it on our website. They're all linked to on our website, beammath.org. If you're in New York or LA, we have trivia night, which is a puzzle-y, mathy trivia, usually in the fall, that you can buy tickets to. So I definitely recommend that. And otherwise, sign up for our newsletter, which you can also do on our website.

    EL: And you're on Twitter also, right?

    RH: Yes, I am. You do have to know how to spell my last name, though.

    EL: Okay.

    RH: Yeah, I'm @ruthihortsch.

    EL: All right. And that's H-O-R-T-S-C-H?

    RH: Good job!

    EL: Yeah, it’s funny, I was actually in a Zoom spelling bee last night. So yeah, I got second place.

    KK: Good for you.

    EL: Got knocked out on diaphoresis.

    KK: Diaphoresis. Wow. Yeah, that's pretty—okay, anyway. All right. Well, thanks for joining us and take care everyone.

    RH: Right. Yeah, it was nice to meet you.

    EL: Bye.

    [outro]

    On today's episode of My Favorite Theorem, we had the privilege to talk with Ruthi Hortsch, a program coordinator at Bridge to Enter Advanced Mathematics (BEAM), a math program for low-income and historically marginalized middle- and high-school students. Dr. Hortsch lives in New York City, which is currently being hit hard by covid-19. We love all our listeners and guests, and right now we are especially thinking about those in New York and other virus hot spots. You may be sick, you may be worried about loved ones, you may be suddenly parenting or caregiving in ways you hadn't expected. We wish you the best, and we hope you enjoy thinking about math for a little bit instead of the news cycle. Stay strong and healthy, friends!

    As you listen to this episode, you may find these links helpful.
    The Bridge to Enter Advanced Mathematics website, Twitter, Facebook, and Instagram pages.
    Ruthi Hortsch on Twitter
    Faltings’s theorem, Dr. Hortsch's favorite theorem
    Our episode with Matilde Lalín, whose favorite theorem was the closely-related Mordell's theorem.
    660 Curries

    Canada/USA Mathcamp

    Tropical Geometry wikipedia page

    29 min
  • Episode 52 - Ben Orlin
    Episode 52 - Ben Orlin
    Kevin Knudson & Evelyn Lamb
    Download

    Kevin Knudson: Welcome to My Favorite Theorem, a math podcast. I'm Kevin Knudson, professor of mathematics at the University of Florida. And here is your other host.

    Evelyn Lamb: Hi, I'm Evelyn Lamb. I'm a freelance math and science writer, usually based in Salt Lake City, but currently still in Providence. I'll be leaving from this semester at ICERM in about a week. So trying to eat the last oysters that remain in the state before I leave and then head back.

    KK: Okay, so you actually like oysters.

    EL: Oh, I love them. Yeah, they're fantastic.

    KK: That is one of those, it’s a very binary food, right? You either love them—and I do not like them at all.

    EL: Oh, I get that, I totally get it.

    KK: Sure.

    EL: They’re like, in some sense objectively gross, but I actually love them.

    KK: Well, I'm glad you've gotten your fill in. Probably—I imagine they're a little more difficult to get in Salt Lake City.

    EL: Yeah, you can but it’s not like you can get over here.

    KK: Might be slightly iffy. You don't know how long they've been out of the water, right?

    EL: Yeah. So there's one place that we eat oysters sometimes there, yeah, that's the only place.

    KK: Yeah, right. Okay. Well, today we are pleased to welcome Ben Orlin. Ben, why don't you introduce yourself?

    Ben Orlin: Yeah, well, thanks so much for having me, Kevin and Evelyn. Yes, I'm Ben Orlin. I’m a math teacher, and I write books about math. So my first book was called Math with Bad Drawings, and my second one is called Change Is the Only Constant.

    EL: Yeah, and you have a great blog of the same name as your first book, Math with Bad Drawings.

    BO: Yeah, thank you. And I think our blogs are, I think almost birthday, not exactly but we started them within months of each other, right? Roots of Unity and Math with Bad Drawings.

    EL: Oh, yeah.

    BO: Began in, like, spring of 2013 which was a fertile time for blogs to begin.

    EL: Yeah. Well, in a few years ago, you had some poll of readers of like, what other things they read and, and stuff and my blog was like, considered the most similar to yours, by some metric.

    BO: Yeah, I did a reader survey and asked people, right, what what other sources they read, and mostly I was looking for reading recommendations. So what else do they consider similar? Overwhelmingly it was XKCD. Not so much—just because XKCD, it’s like if you have a little light that you're holding, a little candle you're holding up, and you're like, what does this remind you of? And like a lot of people are going to say the sun because they look up, and that’s where they see visible light.

    KK: Sure.

    BO: But I think in terms of actually similar writing, I think Toots of Unity is not so different, I think.

    EL: Yeah. So I thought that was interesting because I have very few drawings on on mine. Although the ones that I do personally create are definitely bad. So I guess there’s that similarity.

    BO: That’s the key thing, committing to the low quality.

    KK: Yeah, but that's just it. I would argue they're actually not bad. So if I tried to draw like you draw, it would be worse. So I guess my book should just be Math with Worse Drawings.

    BO: Right.

    KK: You actually get a lot of emotion out of your characters, even though they're they're simple stick figures, right? There’s some skill there.

    BO: Yeah, yeah. So I tried. I tried to draw them with a very expressive faces. Yeah, they're definitely still bad drawings is my feeling. Sometimes people say like, “Oh, but they've gotten so much better since you started the blog,” which is true, but it's one of these things where they could they could get a lot better every five-year interval for the next 50 years and still, I think not look like professional drawings by the end of it.

    EL: Right. You're not approaching Rembrandt or anything.

    KK: All right, so we asked you on here, because you do have bad drawings, but you also have thoughts about mathematics and you communicate them very well through your drawings. So you must have a favorite theorem. What is it?

    BO: Yeah. So this one is drawn from my second book, actually, the second book is about calculus. And I have to confess I already kind of strayed from the assignment because it's not so much a favorite theorem as a favorite construction.

    KK: Oh, that’s cool.

    EL: You know, we get rule breakers on here. So yeah, it happens.

    BO: Yeah, I guess that's the nature of mathematicians, they like to bend the rules and imagine new premises. So pretending that this were titled My Favorite cCnstruction, I would pick Weierstrass’s function. So that you know, first introduced in 1872. And the idea is it's this function which is continuous everywhere and differentiable nowhere.

    EL: Yeah. Do you want to describe maybe what this looks like for anyone who might not have seen it yet?

    BO: Yeah, sure. So when you're picturing a graph, right, you're probably picturing—it varies. I teach secondary school. So students are usually picturing a fairly small set of possibilities, right? Like you're picturing a line, maybe you're thinking of a parabola, maybe something with a few more squiggles, maybe as many squiggles as a sine wave going up and down. But they all have a few things in common one is that almost anything that students are going to picture is continuous everywhere. So basically, it's made of one unbroken line. You can imagine drawing it with your pencil without picking the pencil up. And then the other feature that they have is that they—this one's a little subtler, but there will be almost no points that are jagged, or sort of crooked, or, you know, if I picture an absolute value graph, right, it sort of is a straight line going down to the origin from the left, and then there's a sharp corner at the origin, and then it rises away from that sharp corner. And so those kind of sharp corners, you may have one or two in a graph a student would draw, but that's sort of it. You know, like sharp corners are weird. You don't can't draw all sharp corners. It feels like between any two sharp corners on your graph, there's going to have to be some some kind of non-sharp stuff connecting it, some kind of smooth bits going between them.

    KK: Right.

    BO: And so what sort of wild about about Weierstrass’s function is that you look at it, and it just looks very jagged. It’s got a lot of sharp corners. And you start zooming in, and you see that even between the sharp corners, there are more sharp corners. And you keep zooming in and there's just sharp corners all the way down. It's what we today call it fractal. Although back then that word wasn't around. And it's just it's the entire thing. Every single point along this curve is in some sense, a sharp corner.

    EL: Yeah, it kind of looks like an absolute value everywhere.

    BO: Yeah, exactly. It has that cusp at every single point you could look at.

    KK: Right? So very pathological in nature. And, you know, I'm sure I've seen the construction of this. Is it easy to say what the construction is? Or is this going to be too technical for an audio format?

    BO: It’s actually not hard to construct. There are there whole families of functions that have the same property. But Weierstrass’s is pretty simple. He starts with basically just a cosine curve. So you sort of have cosine of πx. So picture, you know, a cosine wave that has a period of two. And then you do another one that has a much shorter period. So you can sort of pick different numbers. But let's say the next one that you add on has a period that's 21 times faster. So it's sort of going up and down much quicker. And it's shorter, though, we've shrunk the amplitude also. So it's only about a third, let's say, as tall. And so you add that onto your first function. So now we've got—we started with just a nice, gentle wave. And now we've got a wave that has lots of little waves kind of coming off of it. And then you keep repeating that process. So the next, the second one in the iteration has a period of 21 cycles for two units. The next one has 212 cycles. And it's 1/9 the height of the original.

    KK: Okay.

    BO: And then after that, you're going to do you know, 213 cycles in the same span, 214 cycles. And so it goes—I don't know if you can hear my daughter is crying in the background, because I think she she finds it sort of upsetting to imagine the function that's has this kind of weird property.

    EL: Fair.

    BO: Especially because it's such a simple construction. Right? It's just, like, little building blocks for her that we're putting together. And one of the things I like about the construction, is it at no step, do you have any non-differentiable points, actually. It's a wave with a little wave on top of it and lots of little waves on top of that, and then tons and tons of little waves on top of that, but these are all smooth, nice, curving waves. And then it's only in the limit, sort of at the at the end of that infinite bridge, that suddenly it goes from all these little waves to its differentiable nowhere.

    KK: I mean, I could see why that would be true, right?

    BO: Yeah, right. Right. It feels like it's getting worse. And you can do—Weierstrass’s function is really a whole family of functions. He came up with some conditions that you need, basically that’s the basic idea. You need to pick an odd number for the number of cycles and then a geometric series for for the amplitude.

    KK: So what's so appealing about this to you? It's just you can't draw it well, like you have to draw it badly?

    KK: Yeah, that's one thing, right. Exactly. I try to push people into my corner, force them to have to drop badly. I do like that this is something—right, graphs of functions are so concrete. And yet this one you really can't draw. I've got it in my book, I have a picture of the first few iterations. And already, you can't tell the difference between the third step and the fourth step. So I had to, I had to, you know, do a little box and an inset picture and say, actually, in this fourth step, what looks like one little wave is really made up of 21 smaller waves. So I do sort of like that, how quickly we get into something kind of unimaginable and strange. And also, you know, I'm not a historian of mathematics. And so I always wind up feeling like I'm peddling sort of fairy tales about about mathematical history more than the complicated truth that is history. But the role that this function played in going from a world where it felt like functions were kind of nice and were something we had a handle on, into opening up this world where, like, oh no, there are all these pathological things going on out there. And there are just these monsters that lurk in the world of possibility.

    KK: Yeah.

    EL: Right. And was this it—Do you know, was this maybe one of the first, or the first step towards realizing that in some measure sense, like, all functions are completely pathological? Do you know kind of where it fell there, or, like, what the purpose was of creating it in the first place?

    BO: Yeah, I think that's exactly right. I don't know the ins and outs of that story. I do know that, right, if you look in spaces of functions, that they sort of all have this property, right, among continuous functions, I think it's only a set of measure zero that doesn't have this property. So the sort of basic narrative as I understand it, leading from kind of the start of the 19th century to the end of the 19th century, is basically thinking that we can mostly assume things are good, to realizing that sometimes things are bad (like this function), culminating in the realization that actually basically everything is bad. And the good stuff is just these rare diamonds.

    EL: Yeah, I guess maybe this slight, I don't know, silver lining, is that often we can approximate with good things instead. I don't know if that's like the next step on the evolution or something.

    BO: Right. Yeah, I guess that's right. Certainly, that's a nice way to salvage some a silver lining, salvage a happy message. Because it's true, right? Even though, a simpler example, the rationals are only a set of measure zero and the reals, you know, they're everywhere, they're dense. So at least, you know, if you have some weird number, you can at least approximate it with a rational.

    EL: Yeah, I was just thinking when you were saying this, how it has a really nice analogy to the rationals. And, and even algebraic numbers and stuff like, “Okay, start naming numbers,” you'll probably name whole numbers, which are, you know, this sparse set of measure zero. It’s like, o”h, be more creative,” like, “Okay, well, I'll name some fractions and some square roots and stuff.” But you're still just naming sets of measure zero, you’re never naming some weird transcendental function that I can't figure out a way to compute it.

    BO: Yeah, it is funny, right? Because in some sense, right? We've imagined these things called numbers and these things called functions. And then you ask us to pick examples. And we pick the most unlikely, nicest hand-picked, cherry-picked examples. And so the actual stuff—we’ve imagined this category called functions, and most of what's in that category that we developed, we came up with that definition, most of what's in there is stuff that's much too weird for us to begin to picture.

    EL: Yeah.

    BO: Which says something about, I guess, our reach exceeding our grasp or something. I don't really know, but they are our definitions can really outrun our intuition.

    EL: Yeah. So where did you first encounter this function?

    BO: That’s a good question. I feel like probably as a kind of folklore bit in maybe 12th grade math. I feel like when I was probably first learning calculus, it was sort of whispered about. You know, my teacher sort of mentioned it offhand. And that was very enticing, and in some sense, that's actually where my whole second book comes from, is all these little bits of folklore, not exactly the thing you teach in class, but the little, I don't know, the thing that gets mentioned offhand. And you go “Wait, what, what was that?” “Oh, well, don't worry. You'll learn about that in your real analysis class in four years.” I don't want to learn about that in four years. Tell me about that now. I want to know about that weird function. And then I think the first proper reading I did was probably in a William Dunham’s book The Calculus Gallery, which is a nice book going through different bits of historical mathematics, beginning with the beginnings of calculus through through like the late 19th century. And he has the here's a nice discussion of the function and its construction.

    KK: So when we were preparing for this, you also mentioned there are connections to Brownian motion here. Do you want to mention those for our audience?

    BO: Yeah, I love that this turns out—so I have some quotes here from right when this function was sort of debuted, right when it was introduced to the world. You have Émile Picard, his line was, “If Newton and Leibniz had thought that continuous functions do not necessarily have a derivative, the differential calculus would never have been invented.” Which I like. If Newton and Leibniz knew what you were going to do to their legacy, they would never have done this! They would have rejected the whole premise. And then Charles Hermite? [Pronounced “her might, wonders if the pronunciation is correct]

    KK: Hermite. [Pronounced “her meet”]

    BO: That sounds better. Sounds good. Sure. Right. His line was, and I don't know what the context was, but, “I turn away with fright and horror from this lamentable evil of functions that do not have derivatives.” Which is really layering on I like the way people spoke in the 19th century. There was more, a lot more flavor to their their language.

    EL: Yeah.

    BO: And Poincaré also, he was saying 100 years ago prior to Weierstrass developing it, such a function would have been regarded as an outrage to common sense. Anyway, so I mention all those. You mentioned Brownian motion, right? The instinct when you see this function is that this is utterly pathological. This is math just completely losing touch with physical reality and giving us these weird intellectual puzzles and strange constructions that can't possibly mean anything to real human beings. And then it turns out that that's not true at all, that Brownian motion—so you look at pollen dancing around on the surface of some water, and it's jumping around in these really crazy aggressive ways. And it turns out our best models of that process, you know, of any kind of Brownian motions—you know, coal dust in the air or pollen on water—our best model to a pretty good approximation has the same property. The path is so jagged and surprising and full of jumps from moment to moment that it's nowhere differentiable, even though the particle obviously sort of has to be continuous. It can’t be discontinuous, I mean, it's jumping, like literally transporting from one place to another. So that's not really the right model. But it is non-differentiable everywhere, which means, weirdly, that it doesn't have a speed, right? Like, a derivative is a is a velocity.

    EL: So that means maybe an average speed but not a speed at any time.

    BO: Yeah, well, actually, even—I think it depends how you measure. I’d have to looked back at this, because what it means sort of between any two moments according to the model, between any two points in time, is traversing an infinite distance. So I guess it could have an average velocity, but the average speed I think winds up being infinite rates. Over a given time interval, you can just take how far it travels that time interval and divide by time, but I think the speed, if you take the absolute value of the magnitude? I think you sort of wind up with infinite speed, maybe? But really, it's just that you can’t—speed is no longer a meaningful notion. It's moving in such an erratic way. that n you can't even talk about speed.

    KK: Well, because that tends to imply a direction. I mean, you know, it’s really velocity. That always struck me as that's the real problem, is that you can't figure out what direction it's going, because it's effectively moving randomly, right?

    BO: Yeah, I think that's fair. Yeah. The only way I can build any intuition about it is to picture a single—imagine a baseball having a single non-differentiable moment. So like, you toss it up in the air. And usually what would happen is that it goes up in the air, it kind of slows down and slows down and slows down. There's that one moment when it's kind of not moving at all. And then it begins to fall. And so the non-differentiable version would be, like, you throw it up in the air, it's traveling up at 10 meters per second, and then a trillionth of a second later, it's traveling down at 10 meters per second. And what's happening at that moment? Well, it's just unimaginable. And now for Brownian motion, you've got to picture that that moment is every moment.

    KK: Right. Yeah. Weird, weird world.

    BO: Yeah.

    KK: So another thing we like to do on this podcast is ask our guests to pair their, well in your case construction, with something. What does the Weierstrass function pair with?

    BO: Yeah. So I think, I have two things in mind, both of them constructions of new things that kind of opened up new new possibilities that people could not have imagined before. So the first one, maybe I should have picked a specific dish, but I'm picturing basically just molecular gastronomy, this movement in in cooking where you take—one example I just saw recently in a book was, I think it was WD-50, a sort of famous molecular gastronomy restaurant in New York, where they had taken, the comes to you and it looks like a small, poppyseed bagel with lox. And then as it gets closer, you realize it's not a poppyseed bagel with lox, it's ice cream that looks almost identical to a poppyseed bagel with lox. So that's sort of weird enough already. And then you take a taste and you realize that actually, it tastes exactly like a poppyseed bagel with lox, because they've somehow worked in all the flavors into the ice cream.

    KK: Hmm.

    BO: Anyway, so molecular gastronomy basically is about imagining very, very weird possibilities of food that are outside our usual traditions, much in the way that Weierstrass’s function kind of steps outside the traditional structures of math.

    EL: Yeah, I like this a lot. It's a good one. Partly because I'm a little bit of a foodie. And like, when I lived in Chicago, we went to this restaurant that had this amazing, like, molecular gastronomy thing. I’m trying to remember one of the things we had was this frozen sphere of blue cheese. And it was so weird and good. Yeah, you’d get you get like puffs of air that are something, and there’s, like, a ham sandwich, but it was like the bread was only the crust somehow there's like nothing inside. Yeah, it was all these weird things. Liquefied olive that was like in inside some little gelatin thing, and so it was just like concentrated olive taste that bursts in your mouth. So good.

    BO: That sounds awesome to me the the molecular gastronomy food. I have very little experience of it firsthand.

    KK: So you mentioned a second possible pairing. What would that be?

    BO: Yeah, so the other one I had in mind is music. It's a Beatles album, Revolver.

    KK: Great album.

    BO: One of my favorite albums, and much like molecular gastronomy shows that the foods that we're eating are actually just a tiny subset of the possible foods that are out there, similarly what revolver did for for pop music and in ’65 whenever it came out.

    KK: ’66.

    BO: Okay. 66 Alright, thank you for that.

    EL: I am not well-versed in albums of The Beatles. You know, I am familiar with the music of the Beatles, don’t worry. But I don't know what's on what album. So what is this album?

    BO: So Kevin and I can probably go to track by track for you.

    KK: I’d have to think about it, but it's got Norwegian Wood on it, for example.

    BO: Oh, that's rubber sole, actually.

    KK: Oh, that’s Rubber Soul. You're right. Yeah, I lost my Beatles cred. That's right. My bad. I mean, some would argue that—so Revolver was, some people argue, was the first album. Before that, albums had just been collections of singles, even in the case of the Beatles, but Revolver holds together as a piece.

    BO: Yeah, that’s one thing. Which again, there's probably some an analogy to Weierstrass’s function there. Also, it begins with this kind of weird countdown where, I don’t remember if it's John or George, but they’re saying 1234 in the intro into Taxman.

    KK: Yeah. Into Taxman, which is probably, it's not my favorite Beatles song, but it's certainly among the top four. Right.

    BO: Yeah. So that one, already right there it’s a pop song about taxes, which is already, so lyrically, we're exploring different parts of the possibility space than musicians were before. Track two is Eleanor Rigby, which is, the only instrumentation is strings. Which again is something that you didn't really hear in pop. You know, Yesterday had brought in some strings, that was sort of innovative. Other bands have done similar things but, but the idea of a song that’s all strings, and then I’m Only Sleeping as the third track, which has this backwards guitar. They recorded the guitar and just played it backwards. And then Yellow Submarine, which is, like, this weird Raffi song that somehow snuck onto a Beatles album. Yeah, and then For No One has this beautiful French horn solo. Yes, every track is drawn from sort of a distant corner of this space of possible popular music, these kind of corners that had not been explored previously. Anyway, so my recommendation is, is think about the Weierstrass function while eating, you know, a giant sphere of blue cheese and listening to Taxman.

    EL: Great. Yeah. I strongly urge all of our listeners to go do that right now.

    BO: Yeah, if anyone does it, it'll probably be the first time that that set of activities has been done in conjunction.

    EL: Yeah. But hopefully not the last.

    BO: Hopefully not the last. That's right. Yeah. And most experiences are like that, in fact.

    KK: So we also like to let our guests plug things. You clearly have things to plug.

    BO: I do. Yeah. I'm a peddler of wares. Yes, so the prominent thing is my blog is Math with Bad Drawings, and you're welcome to come read that. I try to post funny, silly things there. And then my two books are Math with Bad Drawings, which kind of explores how math pops up in lots of different walks of life, like, you know, in thinking about lottery tickets or thinking about the Death Star is another chapter, and then Change Is the Only Constant is my second book, and it's all about calculus, and it’s sort of calculus through stories. Yeah, that one just came out earlier this year, and I'm quite proud of that one. So you should check it out.

    KK: Yeah, so I own both of them. I've only read Math with Bad Drawings. I've been too busy so far to get to Change Is the Only Constant.

    EL: And there were there been a slew of good pop—or I assume good because I haven't read most of them yet—pop math books that have come out recently, so yeah I feel like my stack is growing. It’s a fall of calculus or something.

    BO: It’s been a banner year. And exactly, calculus has been really at the forefront. Steve Strogatz’s Infinite Powers was a New York Times bestseller, and then David Bressoud [Calculus Reordered] and others who I'm blanking on right now have had one. There was another graphic, like, cartoon calculus that came out earlier this year. So yeah, apparently calculus is kind of having a moment.

    EL: Well, and I just saw one about curves.

    KK: Curves for the Mathematically Curious. It's sitting on my desk. Many of these books that you've mentioned are sitting on my desk.

    EL: So yeah, great year for reading about calculus, but I think Ben would prefer that you start that reading with Change Is the Only Constant.

    BO: It's very frothy, it's very quick and light-hearted and should be—you can use it as your appetizer to get into the the, the cheesier balls of the later books.

    KK: But it's highly non-trivial. I mean, you talk about really interesting stuff in these books. It's not some frothy thing. I mean it's lighthearted, but it's not simple.

    BO: I appreciate that. Yeah, the early draft of the book I was doing pretty much a pretty faithful march through the AP Calculus curriculum. And then that draft wasn't really working. And I realized that part of what I wasn't doing that should be doing was since I'm not teaching, you know, you had to execute calculus maneuvers. I'm not teaching how to take derivatives. I can talk about anything as long as I can explain the ideas. So we've got Weierstrass’s function in there. And there's a little bit even on Lebesgue integration, and other sort of, some stuff on differential equations crops up. So since I'm not actually teaching a calculus course and I don't need to give tests on it, I just got to tell stories.

    EL: Well, yeah, I hope people will check that out. And thanks for joining us today.

    BO: Yeah, thanks so much for having me.

    KK: Yeah. Thanks, Ben.

    [outro]

    Our guest on this episode, Ben Orlin, is a high school math teacher best-known for his blog and popular math books. He told us about Weierstrass’s construction of a function that is continuous everywhere but differentiable nowhere. Here is a short collection of links that might be interesting.

    Ben’s Blog, Math with Bad Drawings

    Math with Bad Drawings, the book

    Change is the Only Constant

    28 min
  • Episode 51 - Carina Curto
    Episode 51 - Carina Curto
    Kevin Knudson & Evelyn Lamb
    Download

    Evelyn Lamb: Hello, and welcome to My Favorite Theorem, the math theorem with no test at the end. I think I decided I liked that tagline. [Editor’s note: Nope, she really didn’t notice that slip of the tongue!]

    Kevin Knudson: Okay.

    EL: So we’re going to go with that. Yeah. I'm one of your hosts, Evelyn Lamb. I'm a freelance math and science writer in Salt Lake City, Utah. And this is the other host.

    KK: I’m Kevin Knudson, a professor of mathematics at the University of Florida. How are you doing?

    EL: I’m doing well. Yeah, not not anything too exciting going on here. My mother-in-law is coming to visit later today. So the fact that I have to record this podcast means my husband has to do the cleaning up to get ready.

    KK: Wouldn’t he do that anyway? Since it’s his mom?

    EL: Yeah, probably most of it. But now I've got a really good excuse.

    KK: Yeah, sure. Well, Ellen and I had our 27th anniversary yesterday.

    EL: Oh, congratulations.

    KK: Yeah, we had a nice night out on the town. Got a hotel room just to sit around and watch hockey, as it turns out.

    EL: Okay.

    KK: But there's a pool at the hotel. And you know, it's hot in Florida, and we don't have a pool. And this is absurd—which Ellen reminds me of every day, that we need a pool—and I just keep telling her that we can either send the kid to college or have a pool. Okay.

    EL: Yeah.

    KK: I mean, I don't know. Anyway, we're not here talking about that, we're talking about math..

    EL: Yes. And we're very excited today to have Carina Curto on the show. Hi, Carina, can you tell us a little bit about yourself?

    Carina Curto: Hi, I'm Carina, and I'm a professor of mathematics at Penn State.

    EL: Yeah, and I think I first—I don't think we've actually met. But I think the first time I saw you was at the Joint Meetings a few years ago. You gave a really interesting talk about, like, the topology of neural networks, and how your brain has these, like, basically kind of mental maps of spaces that you interact with. It was really cool. So is that the kind of research you do?

    CC: Yeah, so that was—I remember that talk, actually, at the Joint Meetings in Seattle. So that was a talk about the uses of typology for understanding neural codes. And a lot of my research has been about that. And basically, everything I do is motivated in some way by questions in neuroscience. And so that was an example of work that's been motivated by neuroscience questions about how your brain encodes geometry and topology of space.

    KK: Now, there's been a lot of a lot of TDA [topological data analysis] moving in that direction these last few years. People have been finding interesting uses of topology in neuroscience and studying the brain and imaging stuff like that, very cool stuff.

    CC: Yeah.

    EL: And did you come from more of a neuroscience background? Or have you been kind of picking that up as you go, coming from a math background?

    CC: So I originally came from a mathematical physics background.

    EL: Okay.

    CC: I was actually a physics major as an undergrad. But I did a lot of math, so I was effectively a double major. And then I wanted to be a string theorist.

    KK: Sure, yeah.

    CC: I started grad school in 2000. So this is, like, right after Brian Greene’s The Elegant Universe came out.

    EL: Right. Yeah.

    CC: You know, I was young and impressionable. And so I kind of went that route because I loved physics, and I loved math. And it was kind of an area of physics that was using a lot of deep math. And so I went to grad school to do mathematical string theory in the math department at Duke. And I worked on Calabi-Yaus and, you know, extra dimensions and this kind of stuff. And it was, the math was mainly algebraic geometry, is what right HD thesis was in So this had nothing to do with neuroscience.

    EL: Right.

    CC: Nothing. And so basically about halfway through grad school—I don't know how better to put it, then I got a little disillusioned with string theory. People laugh now when I say that because everybody is.

    KK: Sure.

    CC: But I started kind of looking for other—I always wanted to do applied things, interdisciplinary things. And so neuroscience just seemed really exciting. I kind of discovered it randomly and started learning a lot about it and became fascinated. And so then when I finished my PhD, I actually took a postdoc in a neuroscience lab that had rats and, you know, was reporting from the cortex and all this stuff, because I just wanted to to learn as much neuroscience as possible. So I spent three years working in a lab. I didn't actually do experiments. I did mostly computational work and data analysis. But it was kind of a total cultural immersion sort of experience, coming from more of a pure math and physics background.

    EL: Right. Yeah, I bet that was a really different experience

    CC: It was really different. So I kind of left math in a sense for my first postdoc, and then I came back. So I did a second postdoc at Courant at NYU, and then started getting ideas of how I could tackle some questions in neuroscience using mathematics. And so ever since then, I've basically become a mathematical neuroscientist. I guess I would call myself.

    KK: So 2/3 of this podcast is Duke alums. That's good.

    CC: Oh yeah? Are you a Duke alum?

    KK: I did my degree there too. I finished in ’96.

    CC: Oh, yeah.

    KK: Okay. Yeah.

    CC: Cool.

    EL: Nice. Well, so what is your favorite theorem?

    CC: So I have many, but the one I chose for today is the Perron-Frobenius theorem.

    KK: Nice.

    EL: All right.

    CC: And so you want to know about it, I guess?

    KK: We do. So do our listeners.

    CC: So it's actually really old. I mean, there are older theorems, but Perron proved it, I think in 1907 and Frobenius in 1912, so it carries both of their names. So it's over 100 years old. And it's a theorem and linear algebra. So it has to do with eigenvectors and eigenvalues of matrices.

    KK: Okay.

    CC: And so I'll just tell you quickly what it is. So, if you have a square matrix, so like an n×n square matrix with all positive—so there are many variations of that theorem. I'm going to tell you the simplest one—So if all the entries of your matrix are positive, then you are guaranteed that your largest eigenvalue is unique and real and is positive, so a positive real part. So eigenvalues can be complex. They can come in complex conjugate pairs, for example, but when we talk about the largest one, we mean the one that has the largest real part.

    EL: Okay.

    KK: All right.

    CC: And so one part of the theorem is that that eigenvalue is unique and real and positive. And the other part is that you can pick the corresponding eigenvector for it to be all positive as well.

    EL: Okay. And we were talking before we started taping that I'm not actually remembering for sure whether we've used the words eigenvector and eigenvalue yet on the podcast, which, I feel like we must have because we've done so many episodes, but yeah, can we maybe just say what those are for anyone who isn't familiar?

    CC: Yeah. So when you have a matrix, like a square matrix, you have these special vectors. So the matrix operates on vectors. And so a lot of people have learned how to multiply a matrix by a vector. And so when you have a vector, so say your matrix is A and your vector is x, if A times x gives you a multiple of x back—so you basically keep the same vector, but maybe scale it—then x is called an eigenvector of A. And the scaling factor, which is often denoted λ, is called the eigenvalue associated to that eigenvector.

    KK: Right. And you want x to be a nonzero vector in this situation.

    CC: Yes, you want x to be nonzero, yes, otherwise it's trivial. And so I like to think about eigenvectors geometrically because if you think of your matrix operating on vectors in some Euclidean space, for example, then what it does, what the matrix will do, is it will pick up a vector and then move it to some other vector, right? So there's an operation that takes vectors to vectors, called linear transformations, that are manifested by the matrix multiplication. And so when you have an eigenvector, the matrix keeps the eigenvector on its own line and just scales, or it can flip the sign. If the eigenvalue is negative, it can flip it to point the other direction, but it basically preserves that line, which is called the eigenspace associated. So it has a nice geometric interpretation.

    EL: Yeah. So the Perron-Frobeius theorem, then, says that if your matrix only has positive entries, then there's some eigenvector that's stretched by a positive amount.

    CC: So yeah, so it says there's some eigenvector where the entries of the vector itself are all positive, right, so it lies in the positive orthant of your space, and also that the the corresponding eigenvalue is actually the largest in terms of absolute value. And the reason this is relevant is because there are many kind of dynamic processes that you can model by iterating a matrix multiplication. So, you know, one simple example is things like Markov chains. So if you have, say, different populations of something, whether it be, say, animals in an ecosystem or something, then you can have these transition matrices that will update the population. And so, if you have a situation where if your matrix that's updating your population has—whatever the leading eigenvalue is of that matrix is going to control somehow the long-term behavior of the population. So that top eigenvalue, that one with the largest absolute value, is really controlling the long-term behavior of your dynamic process.

    EL: Right, it kind of dominates.

    CC: It is dominating, right. And you can even see that just by hand when you sort of multiply, if you take a matrix times a vector, and then do it again, and then do it again. So instead of having A times x, you have A squared times x or A cubed times x. So it's like doing multiple iterations of this dynamic process. And you can see how, then, what’s going to happen to the to the vector if it's the eigenvector. Well, if it's an eigenvector, well, what's going to happen is when you apply the matrix once, A times x, you're going to get λ times x. Now apply A again. So now you're applying A to the quantity λx, but the λ comes out, by the linearity of the of the matrix multiplication, and then you have Ax again, so you get another factor of λ, so you get λ^2 times x. And so if you keep doing this, you see that if I do A^k times x, I get λ^k times x. And so if that λ is something, you know, bigger than 1, right, my process is going to blow up on me. And if it's less than 1, it's going to converge to zero as I keep taking powers. And so anyway, the point is that that top eigenvector is really going to dominate the dynamics and the behavior. And so it's really important if it's positive, and also if it's bigger or less than 1, and the Perron-Frobenius theorem basically tells you that you have, it gives you control over what that top eigenvalue looks like and moreover, associates it to an all-positive eigenvector, which is then a reflection of maybe the distribution of population. So it's important that that be positive too because lots of things we want to model our positive, like populations of things.

    KK: Negative populations aren't good. Yeah,

    CC: Yes, exactly. And so this is one of the reasons it's so, useful is because a lot of the things we want to model are—that vector that we apply the matrix to is reflecting something like populations, right?

    KK: So already this is a very non-obvious statement, right? Because if I hand you an arbitrary matrix, I mean, even like a 2×2 rotation matrix, it doesn't have any eigenvalues, any real eigenvalues. But the entries aren't all positive, so you’re okay.

    CC: Right. Exactly.

    KK: But yeah, so a priori, it's not obvious that if I just hand you an n×n matrix with all real entries that it even has a real eigenvalue, period.

    CC: Yeah. It's not obvious at all, and let alone that it's positive, and let alone that it has an eigenvector that's all positive. That's right. And the positivity of that eigenvector is really important, too.

    EL: Yeah. So it seems like if you're doing some population model, just make sure your matrix has all positive entries. It’ll make your life a lot easier.

    CC: So there's an interesting, so do you do you know what the most famous application of the Perron-Frobenius theorem is?

    EL: I don't think I do.

    KK: I might, but go ahead.

    CC: You might, but I’ll go ahead?

    KK: Can I guess?

    CC: Sure.

    KK: Is it Google?

    CC: Yes. Good. Did you Google it ahead of time?

    KK: No, this is sort of in the dark recesses of my memory that essentially they computed this eigenvector of the web graph.

    CC: Right. Exactly. So back in the day, in the late ‘90s, when Larry Page and Sergey Brin came up with their original strategy for ranking web pages, they used this theorem. This is like, the original PageRank algorithm is based on this theorem, because they're, they have again the Markov process where they imagine some web—some animal or some person—crawling across the web. And so you have this graph of websites and edges between them. And you can model the random walk across the web as one of these Markov processes where there's some matrix that that reflects the connections between web pages that you apply over and over again to update the position of the of the web crawler. And and so now if you imagine a distribution of web crawlers, and you want to find out in the long run what pages do they end up on, or what fraction of web crawlers end up on which pages, it turns out that the Perron-Frobenius theorem gives you precisely the existence of this all-positive eigenvector, which is a positive probability that you have on every website for ending up there. And so if you look at the eigenvector itself, that you get from your web matrix, that will give you a ranking of web pages. So the biggest value will correspond to the most, you know, trafficked website. And smaller values will correspond to less popular websites, as predicted by this random walk model.

    EL: Huh.

    CC: And so it really is the basis of the original PageRank. I mean, they do fancier things now, and I'm sure they don't reveal it. But the original PageRank algorithm was really based on this. And this is the key theorem. So I think it's a it's kind of a fun thing. When I teach linear algebra, I always tell students about this.

    KK: Linear Algebra can make you billions of dollars.

    CC: Yes.

    KK: That’ll catch students’ attention.

    CC: Yes, it gets students’ attention.

    EL: Yes. So where did you first encounter the Perron-Frobenius theorem?

    CC: Probably in an undergrad linear algebra class, to be honest. But I also encountered it many more times. So I remember seeing it in more advanced math classes as a linear algebra fact that becomes useful a lot. And now that I'm a math biologist, I see it all the time because it's used in so many biological applications. And so I told you about a population biology application before, but it also comes up a lot in neural network theory that I do. So in my own research, I study these competitive neural networks. And here I have matrices of interactions that are actually all negative. But I can still apply the theorem. I can just flip the sign.

    EL: Oh, right.

    CC: And apply the theorem, and I still get this, you know, dominant eigenvalue and eigenvector. But in that case, the eigenvalue is actually negative, and I still have this all-positive eigenvector that I can choose. And that's actually important for proving certain results about the behavior of the neural networks that I study. So it's a theorem I actually use in my research.

    EL: Yeah. So would you say that your appreciation of it has grown since you first saw it?

    CC: Oh for sure. Because now I see it everywhere.

    EL: Right.

    CC: It was one of those fun facts, and now it’s in, you know, so many math things that I encounter. It's like, oh, they're using the Perron-Frobenius theorem. And it makes me happy.

    EL: Yeah, well, when I first read the statement of the theorem, it's not like it bowled me over, like, “Oh, this is clearly going to be so useful everywhere.” So probably, as you see how many places it shows up, your appreciation grows.

    CC: Yeah, I mean, that's one of the things that I think is really interesting about the theorem, because, I mean, many things in math are like this. But you know, surely when Perron and Frobenius proved it over 100 years ago, they never imagined what kinds of applications it would have. You know, they didn't imagine Google ranking web pages, or the neural network theory, or anything like this. And so it's one of these things where it's like, it's so basic. Maybe it could look initially like a boring fact of linear algebra, right? If you're just a student in a class and you're like, “Okay, there's going to be some eigenvector, eigenvalue, and it's positive, whatever.” And you can imagine just sort of brushing it off as another boring fact about matrices that you have to memorize for the test, right? And yet, it's surprisingly useful. I mean, it has applications in so many fields of applied math and in pure math, and so it's just one of those things that gives you respect for even seemingly simple and not obviously, it doesn't bowl you over, right, you can see the statement and you're not like, “Wow, that's so powerful!” But it ends up that it's actually the key thing you need in so many applications. And so, you know, it's earned its place over time. It's aged nicely.

    EL: And do you have a favorite proof of this theorem?

    CC: I mean, I like the elementary proofs. I mean, there are lots of proofs. So I think there's an interesting proof by Birkhoff. There are some proofs that involve the Brouwer fixed point theorem, which is something maybe somebody has chosen already.

    EL: Yes, actually. Two people have chosen it!

    CC: Two people have chosen the Brouwer fixed point theorem. Yeah, I would imagine that's a popular choice. So, yeah, there are some proofs that rely on that, which I think is kind of cool. So those are more modern proofs of it. That's the other thing I like about it, is that it has kind of old-school elementary proofs that an undergrad in a linear algebra class could understand. And then it also has these more modern proofs. And so it's kind of an interesting theorem in terms of the variety of proofs that it admits.

    KK: So one of the things we like to do on this podcast is we like to invite our guests to pair their theorem with something. So I'm curious, I have to know what pairs well with the Perron-Frobenius theorem?

    CC: I was so stressed out about this pairing thing!

    KK: This is not unusual. Everybody says this. Yeah.

    CC: What is this?

    KK: It’s the fun part of the show!

    CC: I know, I know. And so don't know if this is a good pairing, but I came up with this. So I went to play tennis yesterday. And I was playing doubles with some friends of mine. And I told them, I was like, I have to come up with a pairing for my favorite theorem. So we chatted about it for a while. And as I was playing, I decided that I will pair it with my favorite tennis shot.

    EL: Okay.

    CC: So, my favorite shot in tennis is a backhand down the line.

    KK: Yes.

    CC: Yeah?

    KK: I never could master that!

    CC: Yeah. The backhand down the line is one of the basic ground strokes. But it's maybe the hardest one for amateur players to master. I mean, the pros all do it well. But, you know, for amateurs, it's kind of hard. So usually people hit their backhand cross court. But if you can hit that backhand down the line, especially when someone's at the net, like in doubles, and you pass them, it's just very satisfying, kind of like, win the point. And for my tennis game, when my backhand down the line is on, that's when I'm playing really well.

    EL: Nice.

    CC: And I like the linearity of it.

    EL: Right, it does seem like, you know, you're pushing it down.

    CC: Like I'm pushing that eigenvector.

    KK: It’s very positive, everything's positive about it.

    CC: Everything’s positive. The vector with the tennis ball, just exploding down the line. It's sort of maybe it's a stretch, but that's kind of what I decided.

    EL: A…stretch? Like with an eigenvalue and eigenvector?

    CC: Right, exactly. I needed to find a pairing that was a stretch.

    EL: I think this is a really great pairing. And you know, something I love about the pairing thing that we do—other than the fact that I came up with it, so of course, I'm absurdly proud of it—is that I think, for me at least it's built all these bizarre connections with math and other things. It's like, now when I see the mean value theorem, I'm like, “Oh, I could eat a mango.” Or like, all these weird things. So now when I see people playing tennis, I'll be like, “Oh, the Perron-Frobenius theorem.”

    CC: Of course.

    EL: So are you a pretty serious tennis player?

    CC: I mean, not anymore. I played in college for a little bit. So when I was a junior, I was pretty serious.

    EL: Nice. Yeah, I’m not really a tennis person I've never played or really followed it. But I guess there's like some tennis going on right now that's important?

    CC: The French Open?

    EL: That’s the one!

    KK: Nadal really stuck it to Federer this morning. I played obsessively in high school, and I was never really any good, and then I kind of gave it up for a long time, and I picked up again in my 30s and did league tennis when I lived in Mississippi. And my team at our level—we were just sort of very intermediate players, you know—we won the state championship two years in a row.

    CC: Wow.

    KK: And then and then I gave it up again when I moved to Florida. My shoulder can't take it anymore. I was one of these guys with a big booming serve and a pretty good forehand and then nothing else, right?

    CC: Yeah.

    KK: So you know, if you work my backhand enough you're going to destroy me.

    EL: Nice. Oh, yeah, that's a that's a lot of fun. And I hope other our tennis appreciator listeners will now have have an extra reason to enjoy this theorem too. So yeah, we also like to give our guests a chance, like if they have a website or book or anything they want to mention—you know, if people want to find them online and chat about tennis or linear algebra— is there anything you want to mention?

    CC: I mean, I don't have a book or anything that I can plug, but I guess I wanted to just plug linear algebra as a subject.

    KK: Sure.

    CC: I feel like linear algebra is one of the grand achievements of humanity in some ways. And it should really shine in the public consciousness at the same level as calculus, I think.

    EL: Yeah.

    KK: Maybe even more.

    CC: Yeah, maybe even more. And now, everybody knows about calculus. Every little kid knows about calculus. Everyone is like, “Oh, when when are you going to get to calculus?” You know, calculus, calculus. And linear algebra—it also has kind of a weird name, right, so it sounds very elementary somehow, linear and algebra—but it's such a powerful subject. And it's very basic, like calculus, and it's used widely and so I just want to plug linear algebra.

    EL: Right. I sometimes feel like there are basically—so math can boil down to like, doing integration by parts really well or doing linear algebra really. Like, I joked with somebody, like, I didn't end up doing a PhD in a field that used a lot of linear algebra, but I sort of got my PhD in applied integration by parts, it's just like, “Oh, yeah. Figure out an estimate based on doing this.” And I think linear algebra, especially now with how important social media and the internet are, it is really an important field that, I agree, more people should know about. It is one of the classes that when I took it in college, it's one of the reasons I—at that time, I was trying to get enough credits to finish my math minor. And I was like, “Oh, yeah, actually, this is pretty cool. Maybe I should learn a little more of this math stuff.” So, yeah, great class.

    CC: And you know, it's everywhere. And you know, there are all these people, almost more people have heard of algebraic topology than linear algebra, outside, you know, because it's this fancy topology or whatever. But when it comes down to it, it's all linear algebra tricks. With some vision of how to package them together, of course, I’m not trying to diminish the field, but somehow linear algebra doesn't get it’s—it’s the workhorse behind so much cool math and yeah, doesn't get its due.

    EL: Yes, definitely agree.

    KK: Yeah. All right. Well, here's to linear algebra.

    EL: Thanks a lot for joining us.

    CC: Thank you.

    KK: It was fun.

    [outro]

    Our guest on this episode, Carina Curto, is a mathematician at Penn State University who specializes in applications in biology and neuroscience. She talked about the Perron-Frobenius theorem. Here are some links you may find useful as you listen to this episode.

    Curto’s website
    A short video of Curto talking about how her background in math and physics is useful in neuroscience and a longer interview in Quanta Magazine
    An article version of Curto’s talk a few years ago about topology and the neural code
    Curto ended the episode with a plug for linear algebra as a whole. If you’re looking for an engaging video introduction to the subject, check out this playlist from 3blue1brown.

    30 min
  • Episode 50 - aBa
    Episode 50 - aBa
    Kevin Knudson & Evelyn Lamb
    Download

    Evelyn Lamb: Hello, and welcome to My Favorite Theorem, a math podcast. I'm one of your hosts, Evelyn Lamb. I'm a freelance math and science writer, usually in Salt Lake City, Utah, currently in Providence, Rhode Island. And this is your other host.

    Kevin Knudson: Hi. I’m Kevin Knudson, professor of mathematics, almost always at the University of Florida these days. How's it going?

    EL: All right. We had hours of torrential rain last night, which is something that just doesn't happen a whole lot in Utah but happens a little more often in Providence. So I got to go to sleep listening to that, which always feels so cozy, to be inside when it's pouring outside.

    KK: Yeah, well, it's actually finally pleasant in Florida. Really very nice today and the sun's out, although it's gotten chilly—people can't see me doing the air quotes—it’s gotten “chilly.” So the bugs are trying to come into the house. So the other night we were sitting there watching something on Netflix and my wife feels this little tickle on her leg and it was one of those big flying, you know, Florida roaches that we have here.

    EL: Ooh

    KK: And our dog just stood there wagging at her like, “This is fun.” You know?

    EL: A new friend!

    KK: “Why did you scream?”

    EL: Yeah, well, we’re happy today to invite aBa to the show. ABa, would you like to introduce yourself?

    aBa Mbirika: Oh, hello. I’m aBa. I'm here in Wisconsin at the University of Wisconsin Eau Claire. And I have been here teaching now for six years. I tell them where I'm from?

    EL: Yeah.

    KK: Sure.

    aM: Okay. I am from, I was born and raised in New York City. I prefer never to go back there. And then I moved to San Francisco, lived there for a while. Prefer never to go back there. And then I went up to Sonoma County to do some college and then moved to Iowa, and Iowa is really what I call home. I'm not a city guy anymore. Like Iowa is definitely my home.

    EL: Okay.

    KK: So Southwestern Wisconsin is also okay?

    aM: Yeah, it's very relaxing. I feel like I'm in a very small town. I just ride my bicycle. I still don't know how to drive, like all my friends from New York and San Francisco. But I don't need a car here. There's nowhere to go.

    EL: Yeah.

    aM: But can I address why you just called me aBa, as I asked you to?

    EL: Yeah.

    aM: Yeah, because maybe I'll just put this on the record. I mean, I don't use my last name. I think the last time I actually said some version of my last name was grad school, maybe? The year 2008 or something, like 10 years ago was the last time anyone's ever heard it said. And part of the issue is that it's It's pronounced different depending on who's saying it in my family. And actually it's spelled different depending on who’s in the family. Sometimes they have different letters. Sometimes there's no R. Sometimes it’s—so in any case, if I start to say one pronunciation, I know Americans are going to go to town and say this is the pronunciation. And that's not the case. I can't ask my dad. He's passed now, but he didn’t have a favorite. He said it five different ways my whole life, depending on context. So he doesn't have a preference, and I'm not going to impose one. So I'm just aBa, and I'm okay with that.

    EL: Yeah, well, and as far as I know, you're currently the only mathematician named aBa. Or at least spelled the way yours is spelled.

    aM: Oh yeah, in the arXiv. Yeah, on Mathscinet that it’s. Yeah, I'm the only one there. Recently someone invited me to a wedding and they were like, what's your address? And I said, “aBa and my address is definitely enough.”

    EL: Yeah, so what theorem would you like to tell us about?

    aM: Oh, okay, well I was listening actually to a couple of you shows recently, and Holly didn’t have a favorite theorem, Holly Krieger. I'm exactly the same way. I don't even have a theorem of, like, the week. She was lucky to have that. I have a theorem of the moment. I would like to talk about something I discovered when I was in college, that’s kind of the reason. but can I briefly say some of my like, top hits just because?

    EL: Oh yeah.

    KK: We love top 10 lists. Yeah, please.

    aM: Okay. So I'm in combinatorics, loosely defined, but I have no reason—I don't know why people throw me in that bubble. But that's the bubble that that I've been thrown in. But my thesis—actually, I don’t ever remember the title, so I have to read it off a piece of paper—Analysis of symmetric function ideals towards a combinatorial description of the cohomology ring of Hessenberg varieties.

    KK: Okay.

    aM: Okay, all those words are necessary there. But my advisor said, “You're in combinatorics.” Essentially, my problem was, we were studying an object and algebraic geometry, this thing called a Hessenberg variety. To study this thing we used topology. We looked at the cohomology ring of this, but that was very difficult. So we looked at this graded ring from the lens of commutative algebra. And I studied the algebra the string by looking at symmetric functions, ideals of symmetric functions, and hence that's where my advisor said, “You're in combinatorics.” So it was the main tool used to study a problem an algebraic geometry that we looked at topology. Whatever, so I don't know what I am. But any case for top 10 hits, not top 10, but diagram chasing. Love it. Love it.

    EL: Wow, I really don't share that love, but I’m glad somebody does love it.

    aM: Oh, it's just so fun for students.

    KK: So the snake lemma, right?

    aM: The snake lemma, yes. It's a little bit maybe above the level of our algebra two class that I teach here for undergrads, but of course I snuck it in anyways. And the short five lemma. Those are like, would be my favorites if the moment was, like, months ago. In number theory I have too many faves, but I’m going to limit it to Euler-Fermat’s theorem that if a and n are coprime, then a to the power of the Euler totient function of n is congruent to 1 mod n. But that leads to Gauss’s epically cool awesome theorem on the existence of primitive roots. Now, this is my current craze.

    EL: Okay.

    aM: And this is just looking at the group of units in Z mod nZ, or more simply the multiplicative group of units of integers modulo n. When is this group cyclic? And Gauss said it's only cyclic when n is the numbers 2, or 4, or an odd prime to a k power, or twice an odd prime to some k-th power. And basically, those are very few. I mean, those are very little numbers in the broad spectrum of the infinity of the natural numbers. So this is very cool. In fact, I'm doing a non-class right now with a professor who retired maybe 10 years ago from our university, and I emailed him and said, “Want to have fun on my like my research day off?” And we’re studying primitive roots because I don't know anything about it. Like, my favorite things are things I know nothing about and I want to learn a lot about.

    EL: Yeah, I don't think I've heard that theorem before. So yeah, I'll have to look that up later.

    aM: Yes. And then the last one is from analysis, and I did hear Adrianna Salerno talked about it and in fact, I think also someone before her on your podcast, but Cantor’s theorem on uncountability of the real numbers.

    EL: Yeah, that's that's a real classic.

    aM: I just taught that two days ago in analysis, and like, it's like waiting for their heads to explode. And I think, I don't know, my students’ heads weren't all exploding. But I was like, “This is so exciting! Why are you not feeling the excitement?” So yeah, yeah, it was only my second time teaching analysis. So maybe I have to work on my sell.

    EL: Yeah, you'll get them next time.

    aM: Yeah. It's so cool! I even mentioned it to my class that’s non-math majors, just looking at sets, basic set theory. And this is my non-math class. These students hate math. They're scared of math. And I say, “You know, the infinity you know, it's kind of small. I mean, you're not going to be tested on this ever. But can I please take five minutes to like, share something wonderful?” So I gave them the baby version of Cantor’s theorem. Yeah, but that's it. I just want to throw those out there before I was forced to give you my favorite theorem.

    EL: Yes. So now…

    KK: We are going to force you, aBa. What is your favorite theorem?

    EL: We had the prelude, so now this is the main event.

    aM: Okay, main event time. Okay, you were all young once, and you remember—oh, we’re all young, all the time, sorry—but divisibility by 9. I guess when we're in high school—maybe even before that—we know that the number 108 is divisible by 9 because 1+0+8 is equal to 9. And that's divisible by 9. And 81 is divisible by 9 because 8+1 is 9, and 9 is divisible by 9. But not just that, the number 1818 is divisible by 9 because 1+8+1+8 is 18. And that's divisible by 9. So when we add up the digits of a number, and if that sum is divisible by 9, then the number itself is divisible by 9. And students know this. I mean, everyone kind of knows that this is true. I guess I was a sophomore in college. That was maybe a good 4 to 6 years after I started college because, well, that was hard. It's a different podcast altogether, but I made some choices to meet friends who made it really hard for me to go to school consistently in San Francisco—part of the reason why I'm kind of okay not going back there much anymore. Friends got into trouble too much.

    But I took a number theory course and learned a proof for that. And the proof just blew my mind because it was very simple. And I wasn't a full-blown math major yet. I think I was in physics— I had eight majors, different majors through the time—I wasn't a math person yet. And I was on a bus going from—Oh, this is in Sonoma County. I went to Sonoma State University as my fourth or fifth college that I was trying to have a stable environment in. And this one worked. I graduated from there in 2004. It definitely worked. So I was on a bus to visit some of my bad friends in San Francisco—who I love, by the way, I'm just saying of the bad habits—and I was thinking about this theorem of divisibility by 9 and saying, what about divisibility by 7? No one talks about that. Like, we had learned divisibility by 11. Like the alternating sum of the digits, if that's divisible by 11, then the number is divisible by 11. But what about 7? You know, is that doable? Or why is it not talked about?

    EL: Yeah.

    aM: So it was an hour and a half bus ride. And I figured it out. And it was extremely, like, the same exact proof as the divisibility by 9, but boiled down to one tiny little change. But it's not so much that I love this theorem. I actually haven't even told it to you yet. But that I did the proof, that it changed my life. I really—that’s the only thing I can go back to and say why am I an associate professor at a university in Wisconsin right now. It was the life-changing event. So let me tell you the theorem.

    EL: Yeah.

    aM: It’s hardly a theorem, and this is why I don't know if it even belongs on this show.

    EL: Oh, it totally does!

    aM: Okay, so I don't even think I had calc 2 yet when I discovered this little theorem. All right, so here we go. So look at the decimal representation of some natural number. Call it n.

    EL: I’ve got my pencil out. I'm writing this down.

    aM: Oh, okay. Oh, great. Okay, I'm reading off a piece of paper that I wrote down.

    EL: Yeah, you said something about it to us earlier. And I was like, “I'm going to need to have this written down.” It’s funny that I do a podcast because I really like looking at things that are written down. That helps me a lot. But let's podcast this thing.

    aM: Okay, so say we have a number with k+1 digits. And so I'm saying k+1 because I want to enumerate the digits as follows: the units digit I'm going to call a0, the tens digit I’ll call a1 the hundreds place digit a2 etc, etc, down to the k+1st digit, which we’ll call ak. So read right to left, like in Hebrew, a0, a1 a2 … (or \cdots, you LaTeX people) ak-1 then the last far left digit ak.

    EL: Yeah.

    aM: So that is a decimal representation of a number. I mean, we're just, you know, like number 1008. That would be a0 is the number 8, a1 is the number 0, a2 is number 0, a3 is the number 1. So we just read right to left. So we can represent this number, and everybody knows this when you're in junior math, I guess in elementary school, that we can write the number—now I'm using a pen—123 as 3 times 1 plus—how many tens do we have? Well, we have two tens. So 2 times 10. How many hundreds do we have? Well, we have one of those. So 1 times 100. So just talking about, yeah, this is mathematics of the place value system in base 10. No surprise here. But a nicer way to write it as a fat sum, where i, the index goes from 0 to k of ai times 10i.

    EL: Yeah.

    aM: That’s how we in our little family of math nerds, how we compactly write that. So when we think about when does this number divisible by 7? It suffices to think about when what is the remainder when each of these summands is—when we divide each of these summands by 7, and then add up all those remainders and then take that modulo 7. So the key and crux of this argument is that what is 10 congruent to mod 7? Well, 10 leaves the remainder of 3 when you divide by 7. In the great language of concurrences—Thank you, Gauss—10≡3 mod 7. So now we can look at this, all of these tens we have. We have a0 ×100+ a1 ×101 + a2 ×102, etc, etc. When we divide this by 7, this number really is now a0 ×30 — because I can replace my 100 with 30 —plus a1 ×31 instead of—because 101 is the same as 31 in modulo 7 land—plus a2 ×32, etc. etc…. to the last one, ak ×3k. Okay, here I am on the bus thinking, “This is only cool if I know all my powers of 3.”

    EL: Yeah. Which are not really that much easier than figuring it out in the first place.

    aM: Okay, but I'm young mathematically and I'm just really super excited. So one little example, I guess this is not, I can't remember what I did on the bus, but 1008 is is a number that's divisible by seven. And let's just perform this check, using this check on this number. So is 1008 really divisible by 7? What we can do is according to this, I take the far right digit, the units digit, and that's 8 ×30, so that's just the number 8, 8×1, plus 0×31. Well, that's just 0, thankfully. Then the next, the hundreds place, that’s 0×32. So that's just another 0. And then lastly, the thousands place, 1×33 and that's 27. Add up now my numbers 8+0+0+27. And that's 35. And that's easy to know that the divisibility of. 7 divides 35 and thus 7 divides 1008. And, yeah, I don't know, I’m traveling back in time, and this is not a marvelous thing. But everybody, unfortunately, who I saw in San Francisco that day, and the next day, learned this. I just had to teach all my friends because I was like, “Well, this is not what I'm doing for college. This is something I figured out on the bus. This math stuff is great.”

    EL: Yeah, just the fact that you got to own that.

    aM: Yeah. And that also it wasn't in the book, and actually it wasn't in subsequently any book I've ever looked in ever since. But it's still just cute. I mean, it's available. And what it did, I guess it just touched me in a way, where I guess I didn't know about research, I didn't know about a PhD program. My end goal was to get a job, continue at the photocopy place that was near the college, where I worked. I really told my boss that, and I really believed that I was going to do that. And our school never really sent people to graduate programs. I was one of the first. And I don't know, it just changed me. And there were a lot of troubles in my life before then. And this is something that I owned. And that's my favorite theorem on that bus that day.

    KK: It’s kind of an origin story, right?

    aM: Yes, because people ask me, how did you get interested in math? And I always say the classic thing. Forget this story, but I'm also not speaking to math people. My usual thing is the rave scene. I mean, that was what I was involved in in San Francisco, and then, I don't know if you know what that is, but electronic dance music parties that happen in beaches and fields and farms and houses.

    EL: What, you don’t think we go to a lot of raves?

    aM: I don’t know if raves still happen!

    EL: You have accurately stereotyped me.

    aM: Okay. Now, I have to admit my parents were worried about that. And they said, “Ecstasy! Clubs!” and I was like, “No, Mom. That's a different rave. My people are not indoors. We’re outdoors, and we're not paying for stuff, and there's no bar, and there's no drinking. We're just dancing and it's daytime. It was a different thing. But that's really why I got involved in this math thing. In some sense, I wanted to know how all of that music worked, and that music was very mathematical.

    EL: Oh.

    aM: But then I kind of lost interest in studying the math of that because I just got involved in combinatorics and all the beautiful, theoretical math that fills my spirit and soul. But the origin story is a little bit rave, but mostly that bus.

    EL: Yeah. A lot of good things happen on buses.

    aM: You guys know about the art gallery theorem? Guarding a museum.

    EL: Yeah. Yeah.

    aM: What’s the minimum number of guards? Okay, I took the seat of someone—my postdoc was at Bowdoin college, and sadly the person who passed away shortly before I got the job was a combinatorialist named Steve Fisk (I hope I’ve got the name right). In any case, he's in the Proofs from the Book, for coming up with a proof for that art gallery theorem. You know, the famous Proofs from the Book, the idea that all the beautiful proofs are in some book? But yeah, guess where he came up with that, he told the chair of the math department when I started there: on a bus! And he was somewhere in Eastern Europe on a bus, and that's where he came up with it. And it's just like, yeah, things can happen on a bus, you know?

    EL: Yeah. Now I want our listeners to, like, write in with the best math they've ever done on a bus or something. A list of bus math.

    aM: You also have to include trains, I think, too.

    EL: Yeah. Really long buses.

    aM: All public transportation.

    EL: Yeah. So something that we like to do on this podcast is ask our guests to pair their theorem with something. So what have you chosen to pair with your favorite theorem?

    aM: Oh my gosh, I was supposed to think about that. Yes. Okay. Oh, 7.

    EL: I feel like you have so many interests in life. You must you must have something you can think of.

    aM: Oh, no, it's not a problem. I do currently a lot of mathematics. I'm in my office, sadly, a lot of hours of the day, but sometimes I leave my office and go to the pub down the road. And I call it a pub because it's really empty and brightly lit and not populated by students. It's kind of like a grown up bar. But I do a lot of recreational math there, especially on primitive roots recently. So I think I would pair my 7 theorem with seven sips of Michelob golden draft light. It's just a boring domestic beer. And then I would go across the street to the pizza place that's across from my tavern, and I would eat seven bites of a pizza with pepperoni, sausage, green pepper, and onion.

    EL: Nice.

    aM: I have a small appetite. So seven people would say yes, he can probably do seven bites before he’s full and needs to take a break.

    EL: Or you could you could share it with seven friends.

    aM: Yes. Oh, I'm often taking students down there and buying pizza for small sections of research students or groups of seven. Yes.

    EL: Nice. So I know you wanted to share some other things with us on this podcast. So do you want to talk about those? Or that? I don't know exactly what form you would like to do this in.

    aM: Oh, I wrote a poem. Yeah, I just want to share a poem that I wrote that maybe your listeners might find cute.

    EL: Yeah. And I'd like to say I think the first time—I don't think we actually met in person that time, but the first time I saw you—was at the poetry reading at a Joint Math Meeting many years ago.

    aM: Oh my gosh! I did this poem, probably.

    EL: You might have. I’ll see I remember you. Many people might have seen you because you do stand out in a crowd. You know, you dress in a lot of bright colors and you have very distinctive glasses and hair and everything. So you were very memorable at the time. Yes, right now it's pink, red, and yeah, maybe just different shades of pink.

    aM: Yes.

    EL: But yeah, I remember seeing you do a poem at this this joint math poetry thing and then kept seeing you at various things and then we met, you know, a few years ago when I was at Eau Claire, I guess, we actually met in person then. But yeah, go ahead, please share your poem with us.

    aM: Okay, this is part of the origin story again. This was just shortly after this seven thing from the bus. I was introduced to a proofs class, and they were teaching bijective functions. And I really didn't get the book. It was written by one of my teachers, and I was like, you know, I wrote a poem about it. And I think I understand my poem a little bit more than what you wrote in your book. And like, they actually sing this song now. So they recite it, so say the teachers at Sonoma State, each year to students who are taking this same course. But here it is, I think it's sometimes called a rap because I kind of dance around the room when I sing it. So it's called the Bijection Function Poem. And here you go. Are you ready?

    EL: Yes.

    KK: Let’s hear it.

    aM: All right.

    And it clearly follows that the function is bijective
    Let’s take a closer look and make this more objective
    It bears a certain quality – that which we call injective
    A lovin’ love affair, Indeed, a one-to-one perspective.
    Injection is the stuff that bonds one range to one domain
    For Mr. X in the domain, only Miss Y can take his name
    But if some other domain fool should try to get Miss Y’s affection,
    The Horizontal Line Police are here to check for 1 to 1 Injection.

    (Okay, that’s a little racy.)

    Observe though, that injection does not alone grant one bijection
    A function of this kind must bear Injection AND Surjection
    Surjection!? What is that? Another math word gone surreal
    It’s just a simple concept we call “Onto”. Here‟s the deal:
    If for EVERY lady ‘y’ who walks the codomain of f
    There exists at least one ‘x’ in the Domain who fancies her as his sweet best.
    So hear the song that Onto sings – a simple mathful melody:
    “There ain’t a Y in Codomain not imaged by some X, you see!”
    So there you have it 2 conditions that define a quality.
    If it’s injective and surjective, then it’s bijective, by golly!

    (So this is the last verse. And there's some homework problems in my last verse, actually.)

    Now if you’re paying close attention to my math-poetic verse
    I reckon that you’ve noticed implications of Inverse
    Inverse functions blow the same tune – They biject oh so happily
    By sheer existence, inverse functions mimic Onto qualities (homework problem 1)
    And per uniqueness of solution, another inverse golden rule (homework problem 2)
    By gosh, that’s one-to-one & Onto straight up out the Biject School!
    Word!

    aM: Yeah, I never tire that one. I love teaching a proofs class.

    EL: Yeah. And you said you use it in your class every time you teach it?

    aM: Every time I have to say bijection. I mean, the song works, though. My only drawback in recent times is my wording long ago for “Mr. X in the domain” and “Miss Y can take his name” and the whole binary that this thing is doing. So I do have versions, I have a homosexual version, I have a this version—this is the hetero version—then I have the yet-to-be-written binary-free version, which I don't know how to make that because I was thinking for “Person X in the domain, only Person Y can take his name,” but you know person doesn't work. It's too long syllabically so I'm working on that one.

    EL: Yeah.

    aM: I’m working on that one.

EL: Well, yeah, modernize it for for the times we live in now.

    aM: Yes. I kind of dread reading and reciting this is purely hetero version, you know? And also there's not necessarily only one Miss Y that can take Mr. X’s name. I mean, you know, there's whole different relation groups these days.

    EL: Yeah.

    aM: But I'm talking about the injection and surjection.

    EL: Yeah, the polyamorous functions are a whole different thing.

    KK: Those are just relations, they’re not functions. It’s a whole thing.

    aM: Oh, yes, relations aren't necessarily functions, but certain ones that be called that right?

    EL: Yeah. Well, thank you so much for joining us. Is there anything else you would like to share? I mean, we often give our guests ways to find—give our listeners ways to find our guests online. So if there's anything, you know, a website, or anything you’d like to share.

    aM: Can you just link my web page or should I tell you it? [Webpage link here] Actually googling “aBa UWEC math.” That's all it takes. UWEC aBa math. Whenever students can’t find our course notes, I just say like, “I don't know, Google it. There's no way you cannot find our course notes if you remember the name of your school, what you're studying and my name.” Yeah.

    EL: We’ll put a link to that also in the show notes for people.

    aM: Yeah, one B, aBa, for the listeners.

    EL: Yes, that's right. We didn't actually—I said it was the only one spelled that way but we didn't spell it. It's aBa, and you capitalize the middle, the middle and not the first letter, right?

    aM: No, yes, that's fine. It looks more symmetric that way.

    EL: Yeah. You could even reverse one of them.

    aM: I usually write the B backwards. Like the band, but I can't do that usually, though. I don't want to be overkill to the people that I work around. But yes, at the bottom of my webpage, I have the links to videos of me singing various songs to students, complex analysis raps, PhD level down to undergraduate level, just different raps that I wrote for funs.

    And I wanted to plug one thing at JMM. I mean, not that it's hard to find it in the program, but I'm an MAA invited speaker this time, and I'm actually scared pooless a little bit to be speaking in one of those large rooms. I don't know how I got invited. But I said yes.

    KK: Of course you said yes!

    aM: Well, I'm excited to share two research projects that I've been doing with students. Because I like doing research just for the sheer joy of it. And I think the topic of my talk is “A research project birthed out of curiosity and joy” or something like that, because one of the projects I'm sharing wasn't even a paid research project. I just had a student that got really excited to study something I noticed in Pascal's triangle, and these tridiagonal real symmetric matrices. I mean, it was finals week, and I was like, “You want to have fun?” And we spent the next year and a half having fun, and now she's pursuing graduate school, and it's great. It's great, research for fun. But one thing I'm talking about that I'm really excited about is the Fibonacci sequence. And I know that's kind of overplayed at times, but I find it beautiful. And we're looking at the sequence modulo 10. So we're just looking at the last, the units digits.

    EL: Yeah, last digits.

    aM: And whenever you take the sequence mod anything, it's going to repeat. And that's an easy proof to do. And actually Lagrange knew that long, long ago. But recently, in 1960, a paper came out studying these Fibonacci sequences modulo some natural number, and proved the periodicity bit and proved—there’s tons of papers in the Fibonacci Quarterly related to this thing. But what I'm looking at in particular is a connection to astrology—which actually might clear the room, but I'm hoping not—but the sequence has a length of periods 60. So if you lay that in a circle, it repeats and every 15th value in the Fibonacci number ends in 0. That's something you can see with the sequence, but it’s a lot easier to see when you're just looking at it mod 10. and that's something probably people didn't know now. Every 15th Fibonacci number ends in 0.

    KK: No, I didn't know that.

    aM: And if it ends in 0, it's a 15th Fibonacci number. And so, it’s an if and only if. And every 5th Fibonacci number is a multiple of five. So in astrology, we have the cardinal signs: Aries, Cancer, Libra and Capricorn. And you and you lay those on the zeros. Those are the zeros. And then the fixed and mutable signs, like Taurus, Gemini, etc, etc. As you move after the birth of the astrological seasons, those ones lay on the fives, and then you can look at aspects between them. Actually, I'm not going to say much astrology, by the way, in this talk. So people who are listening, please still come. It's only math! But I'm going to be looking at sub-sequences, but it got inspired by some videos online that I saw by a certain astrologer. And I—there was no mathematics in the videos and I was like, “Whoa, I can fill these gaps.” And it's just beautiful. Certain sub-sequences in the Fibonacci sequence mod 10 give the Lucas sequences mod 10. The Lucas sequence, and I don't know if your listeners or you guys know what the Lucas sequence is, but it's the Fibonacci sequence, but the starting values are 2 and then 1.

    KK: Right.

    aM: Instead of zero and one.

    EL: Yeah.

    aM: And Edward Lucas is the person, actually, who named the Fibonacci sequence the Fibonacci sequence! So this is a big player. And I am really excited to introduce people to these beautiful sub-sequences that exist in this Fibonacci sequence mod 10. It's like, just so sublime, so wonderful.

    EL: I guess I never thought about last digits of Fibonacci numbers before, but yeah, I hope to see that, and we'll put some information about that in the show notes too. Yeah, have a good rest of your day.

    aM: All right, you too, both of you. Thank you so much for this invitation. I’m happy to be invited.

    EL: Yeah, we really enjoyed it. v
    KK: Thanks, aBa.

    aM: All right. Bye-bye.

    On this episode of My Favorite Theorem, we talked with aBa Mbirika, a mathematician at the University of Wisconsin Eau Claire. He told us about several favorite theorems of the moment before zeroing in on one of his first mathematical discoveries: a way to determine whether a number is divisible by 7.

    Here are some links you may find interesting after listening to the episode.

    aBa’s website at UWEC
    Snake lemma
    Short five lemma
    Euler-Fermat’s theorem
    Gauss’s primitive roots
    Adriana Salerno’s episode of the podcast
    Steve Fisk’s “book proof” of the art gallery theorem
    Information on aBa’s MAA invited address at the upcoming Joint Mathematics Meetings

    36 min
  • Episode 49 - Edmund Harriss
    Episode 49 - Edmund Harriss
    Kevin Knudson & Evelyn Lamb
    Download

    Kevin Knudson: Welcome to My Favorite Theorem, math podcast and so much more. I'm Kevin Knudson, professor of mathematics at the University of Florida, and I am joined today by your other host.

    Evelyn Lamb: Hi, I'm Evelyn Lamb. I'm a freelance math and science writer, usually based in Salt Lake City, but today coming from the Institute for Computational and Experimental Research in Mathematics at Brown University in Providence, Rhode Island, where I am in the studio with our guest, Edmund Harriss.

    KK: Yeah. this is great. I’m excited for this, this new format where we're, there's only two feeds to keep up with instead of three.

    EL: Yeah, he even had a headphone splitter available at a moment's notice.

    KK: Oh, wow.

    EL: So yeah, this is—we’re really professional today.

    KK: That’s right.

    EL: So yeah, Edmund, will you tell us a little bit about yourself?

    Edmund Harriss: I was going to say I'm the consummate unprofessional. But I'm a mathematician at the University of Arkansas. And as Evelyn was saying, I'm currently at ICERM for the semester working on illustrating mathematics, which is an amazing program that's sort of—both a delightful group of people and a lot of very interesting work trying to get these ideas from mathematics out of our heads, and into things that people can put their hands on, people can see, whether they be research mathematicians or other audiences.

    EL: Yeah. I figured before we actually got to your theorem, maybe you could say a little bit about what the exact—or some of the mathematical illustration that you yourself do.

    EH: So, yeah, well, one of the big pieces of illustration I've done will come up with a theorem,

    EL: Great.

    EH: But I consider myself a mathematician and artist. And a part of the artistic aspect, the medium—well, both the medium but more than that, the content, is mathematics. And so thinking about mathematical ideas as something that can be communicated within artwork. And one of the main tools I've used for that is CNC machines. So these are basically robots that control a router, and they can move around, and you can tell it the path to move on and carve anything you like. So even controlling the machine is an incredibly geometric operation with lots of exciting mathematics to it. When I first came across—so one of the sorts of machine you can have is called a five-axis machine. That's where you control both the position, but also the direction that you're cutting in. So you could change the angle as its as its cutting. And so that really brings in a huge amount of mathematics. And so when I first saw one of these machines, I did the typical mathematician thing, and sort of said, “Well, I understand some aspects of how this works really well. How hard can the stuff I don't understand be?” It took me several years to work out just how hard some of the other problems were. So I've written software that can control these machines and turn—in fact, even turn a hand-drawn path into a something the machine can cut. And so to bring it back to the question, which was about illustrating mathematics: One of the nice things about that idea is it takes a sort of hand-drawn path—which is something that's familiar to everyone, especially people in architecture or art, who are often wanting to use these machines, but not sure how—and the mathematics comes from the notion that we take that hand-drawn path, and we make a representation of that on the computer. And so you've got a really interesting function, they're going from the hand drawn path through to the the computer representation, you can then potentially manipulate it on the computer before then passing it again back to the machine. And so now the output of the machine is something in the real world. The initial hand-drawn path was in the real world, and we sort of saw this process of mathematics in the middle.

    Amongst other things, I think this is a really sort of interesting view on a mathematical model. you have something in the real world, you pull it into an abstract realm, and then you take that back into the world and see what it can tell you. In this case, it's particularly nice because you get a sense of really what's happening. You can control things, both in the abstract and in the world. And I think, you know, to me that really speaks to the power of thinking and abstraction of mathematics. Of course, also controlling these machines allows you to make mathematical models and objects. And so a lot of my my work is sort of creating mathematical models through that, but I think the process is a more interesting, in many ways, mathematical idea, illustration of mathematics, that the objects that come out

    KK: Okay, pop quiz. What's the configuration space of this machine? Do you know what it is?

    EH: Well, it depends on which machine.

    KK: The one you were describing, where you can where you can have the angles changing. That must affect the topology of the configuration space.

    EH: So it’s R3 crossed with a torus.

    KK: Okay.

    EH: And so even though you're changing the angle of the bit, you really need to think about a torus. It's really also a subset of a torus because you can't reach all angles.

    KK: Sure, right.

    EH: But it is a torus and not a sphere.

    KK: Yeah. Okay.

    EH: So if you think about how to get from one position of the machine to another, you really want to—if you think about moving on a sphere, it's going to give you a very odd movement for the machine, whereas moving along a torus gives the natural movement.

    KK: Sure, right. All right. So, what's your favorite theorem?

    EH: So my favorite theorem is the Gauss-Bonnet.

    KK: All the way with Gauss-Bonnet!

    EL: Yes. Great theorem. Yeah.

    EH: And I think in many ways, because it speaks to what I was saying earlier about the question: as we move to abstraction, that starts to tell us things about the real world. And so the Gauss-Bonnet theorem comes at this sort of period where mathematics is becoming a lot more abstract. And it's thinking about how space works, how we can work with things. You're not just thinking about mathematics as abstracted from the world, but as sort of abstraction in its own right. On the artist side, a bit later you have discussion of concrete art, which is the idea that abstract art starts with reality and then strips things away until you get some sort of form, whereas concrete art starts from nothing and tries to build form up. And I think there's a huge, nice intersection with mathematics. And in the 19th century, you've got that distinction where people were starting to think about objects in their own right. And as that happens, suddenly this great insight, which is something that can really be used practically—you can think about the gospel a theorem, and it's something that tells you about the world. So I guess I should now say what it is.

    EL: Yeah, that would be great. Actually, I guess it must have been almost two years ago at this point, we had another guest who did choose the Gauss-Bonnet theorem, but in case someone has not religiously listened to every single episode—

    KK: Right, this was some time ago.

    EL: Yeah, we should definitely say it again.

    EH: So the gospel out there links the sort of behavior of a surface to what happens when you walk around paths on that surface. So the simplest example is this: I start off, I’m on a sphere, and I start at the North Pole and I walk to the equator. At the equator, I turn 90 degrees, I walk a quarter of the way around the Earth, I turn 90 degrees again, and I walk back to the North Pole. And if I turn a final 90 degrees, I’m now back where I started facing in the same direction that I started. But if I look at how much I turned, I didn't go through 360 degrees. So normally if we go around a loop on a nice flat sheet, if you come back to a started pointing in the same direction, you've turned through 360 degrees. So in this path that I took on sphere, I turned through 270 degrees, I turned through too little. And that tells me something about the surface that I'm walking on. So even if I knew nothing about the surface other than this particular loop, I would then know that the surface inside must be mostly positively curved, like a sphere.

    And similarly if I did the same trick, but instead of doing it on the sphere, I took a piece of lettuce and started walking around the edge of a piece of lettuce, in fact, I’d find that when I got back to where I started, I’d turned a couple of hundred times round, instead of just once, or less than once, as in the case the sphere. And so in that case, you've got too much turning. And that tells you that the surface inside is made up of a lot of saddles. It's a very negatively curved surface. And one of the motivations of creating this theorem for Gauss, I believe—I always find it dangerous to talk about history of mathematics in public because you never know what the apocryphal stories are—one of the questions Gauss was interested in was not whether or not the earth was a sphere. Well, actually, whether or not the earth was a sphere. So not whether or not it was round, or topologically a ball, but whether it was geometrically really a perfect sphere. And now we can go up into space and have a look back at the earth, and so we can sort of do a three-dimensional version of that, regard the earth as a three dimensional sphere, but Gauss was stuck on the surface of the earth. So he really had this sort of two dimensional picture. And what you can do is create different triangles and ask, for those triangles, what’s the average amount of curvature? So I look at that turning, I look at the total area, the size of the triangle, and ask does that average amount of curvature change as I draw triangles in different places around the earth? And at least to Gauss’s measurements—again, in the potentially apocryphal story I heard—the earth appeared to be a perfect sphere up to the level of measurement, they were able to do then. I think now, we know that the earth is an oblate spheroid, in other words, going between the poles is a slightly shorter distance than across the equator.

    KK: Right.

    EH: I believe that it was only a couple of years ago that we managed to make spheres that were more perfect than the Earth. So it was sort of, yeah, the Earth is one of the most perfect spheres that anyone has experience of, but it's not quite a perfect sphere when your measurements are fine enough.

    KK: So what's the actual statement of Gauss-Bonnet?

    EH: So, the statement is that the holonomy, which is a fancy word for the amount of turning you do as you go around a path on the surface, is equal to—now I’m forgetting the precise details—so that turning is closely related to the integral of the Gaussian curvature as you go over the whole surface.

    KK: Right.

    EH: So it's relating going around that boundary—which is a single integral because you're just moving around a path—to the double integral, which is the going over every point in the surface. And the Gaussian curvature is the notion of whether you're like a sphere, whether you're flat, or whether you're like a saddle at each individual point.

    KK: And the Euler characteristic pops up in here somewhere if I remember right.

    EH: Yeah. So the version I was giving was assuming that you’re bounding a disk in the surface, and you can do a more powerful version that allows you to do a loop around something that contains a donut.

    EL: Yeah, and it relates the topology of a surface, which seems like this very abstract thing, to geometry, which always seems more tangible.

    EH: Yeah. Yeah, the notion that the total amount of curvature doesn't change as you shift things topologically.

    EL: Right.

    EH: Even though you can push it about locally.

    KK: Yeah. So if you're if you're pushing it in somewhere, it has to be pooching out somewhere else. Right? That's essentially what's going on, I guess. Right?

    EH: Yeah. You know, another thing that's really nice about the the Gauss-Bonnet theorem, it links back to the Euler characteristic and that early topological work, and sort of pulls the topology in this lovely way back into geometric questions, as Evelyn said. And then the Euler characteristic has echoes back to Descartes. So you're seeing this sort of long development of the mathematics that's coming out. It’s not something that came from nowhere. It was slowly developed by insight after insight, of lots of different thinking on the nature of surfaces and polyhedra and objects like that.

    EL: Yeah. And so where did you first encounter this theorem?

    EH: So this is rather a confession, because—when I was a undergraduate, I absolutely hated my differential equations course. And I swore that I would never do any mathematics involved in differential equations. And I had a very wise PhD advisor who said, “Okay, I'm not going to argue with you on this, but I predict that at some point, you will give me a phone call and say you were wrong. And I don't know when that will be. But that's my prediction.”

    KK: Okay.

    EH: It did take several years. And so yes, many years later, I'd learned a lot of geometry, and I wanted to get better control over the geometry. So I sort of got into doing differential geometry not through the normal route—which is you sort of push on through calculus—but through first understanding the geometry and then wanting to really control—specifically thinking about surfaces that were neither the geometry of the sphere, the plane, or the hyperbolic plane. Those are three geometries that you can look at without these tools. But when you want to have surfaces that have saddles somewhere and positive curvature—I mean, this relates back to the CNC because you're needing to understand paths on surfaces there in order to take our tool and produce surfaces.

    And so I realized that the answers to all my questions lay within differential equations, and actually differential equations were geometric, so I was foolish to dislike them. And I did call up my advisor and say, “Your prediction has come true. I'm calling you to say I was wrong.”

    EL: Yeah.

    EH: So basically, I came to it from looking at geometry and trying to understand paths on surfaces and realizing from from there that there was this lovely toolkit that I had neglected. And one of the real gems of this toolkit was this theorem. And I think it's a real shame that it's not something that's talked about more. I’ve said this is a bit like the Sistine Chapel of mathematics. You know, most people have heard of the Sistine chapel.

    KK: Sure.

    EH: Quite a lot of people can tell you something that's actually in it.

    EL: Right.

    EH: And slowly, only a few people have really seen it. And certainly a very few people have studied it and really looked and can tell you all the details. But in mathematics, we tend to keep everything hidden until people are ready to hear the details. And so I think this is a theorem that you can really play with and see in the world. I mean, it's not a—there are some models and things you can build that are not great for podcasts, but it's something you can really see in the world. You can put it put items related to this theorem into the hands of people who are, you know, eight or nine years old, and they can understand it and do something with it and and see how what happens because all you have to do is give people strips of paper and ask them to start connecting them together, just controlling how the angles work at the corners.

    And depending on whether those angles add up to less than 360 degrees—well not the angles at the corner—depending on whether the turning gives you less than 360, exactly 360, or more than 360, you're going to get different shapes. And then you can start putting those shapes together, and you build out different surfaces. And so you can then explore and discover a lot of stuff in a sort of naive way You certainly don't need to understand what an integral is in order to have some experience of what the Gauss-Bonnet theorem is telling you. And so this is sort of it's that aspect, that this is something that was always there in the world. The sort of experiments, the sort of geometry you can look at, through differential geometry and things like the Gauss-Bonnet, that was available to the whole history of mathematics, but we needed to make a break from just geometry as a representation of the world to then sort of step back and look at this result that is a very practical, hands-on one.

    You know, if you really want to control things, then you do need to have solid multivariate calculus. So generally, the three-semester course of calculus is often meant to finish with Gauss-Bonnet, and it's the thing that's dropped by most people at the end of the semester, because you don't quite have time for it. And there's not going to be a question on the test. But it's one of those things that you could sort of put out there and have a greater awareness of in mathematics. Just as: this is an interesting, beautiful result. I would say, you know, it's one of humanity's greatest achievements to my mind. You don't have to really be able to understand it perfectly in order to appreciate it. You certainly—as I proved you—can appreciate it without being able to state it exactly.

    EL: Yeah, well, you've sold me—although, as we've learned to this podcast, I'm extremely open—susceptible to suggestion.

    KK: That’s true. Evelyn's favorite theorem has changed multiple times now. That's right.

    EL: Yeah. And I think you brought it back to Gauss-Bonnet. Because when when we had Jeanne Clelland earlier, who said Gauss-Bonnet, I was like, “Well, yeah, I guess the uniformization theorem is trash now”—my previous favorite theorem, but now—it had been pulled over to Cantor again, but you’ve brought it back.

    KK: Excellent. All right, so that's another thing we do on this podcast is ask our guest to pair their theorem with something. So Edmund, what pairs well with Gauss-Bonnet?

    EH: Well, I have to go with a walnut and pear salad.

    KK: Okay.

    EL: All right.

    KK: I’m intrigued.

    EH: Well, I think I've already mentioned lettuce.

    EL: Yes.

    EH: Lettuce is an incredibly interesting curved surface. Yeah. And then you've got pears, which gives you—

    KK: Spheres.

    EH: A nice positively curved thing. But they're not just boring spheres.

    EL: Yeah.

    EH: They have some nice interesting changes of curvature. And then walnuts are also something with very interesting changing curvature. They have very sharply positively curved pieces where they're sort of coming in but then they've got all these sort of wrinkly saddley parts. In fact, one of the applications of the Gauss-Bonnet theorem in nature is how do you create a surface that sort of fits onto itself and fills a lot of space—or doesn't fill that much space but gives you a very high surface area to volume ratio. So walnut is an example—or brains or coral—you see the same forms coming up. And the way many of those things grow is by basically giving more turning as you grow to your boundary.

    KK: Right.

    EH: And that naturally sort of forces this negatively-curved thing. So I think the salad really shows you different ways in which this surface can—the theorem can affect the behaviors of the surfaces.

    EL: Yeah, well, what I want now is something completely flat to put in the salad. Do you have any suggestions?

    KK: Usually you put goat cheese in such a thing, but that doesn't really work.

    EL: That’s—well, parmesan. You could shave paremesan.

    EH: Yeah, shavings of parmesan. Or maybe some thin-cut salami.

    EL: Okay.

    EH: And so even though those things would bend over—I mean, we’re now on to a different theorem of Gauss, and I don’t meant to corrupt Evelyn away—but you know, when you thinly cut the salami, it can it can bend but it doesn't actually change its curvature.

    KK: Right.

    EH: Your loops on that salami are going to have the same behavior that they had before. And I guess I should also say that I did create a toy that makes that paper model that I talked about easier to use. You don't have to use tape. You can hook together pieces. And so the toy is called Curvahedra.

    KK: I was going to say, you should promote your toy. Yeah.

    EH: I’m terrible at self-promotion, yes.

    EL: We will help you. Yes, this is a very fun toy. I actually got to play with it for first time a few weeks ago when you did a little short thing and I think when I had seen pictures of it before I thought it was not going to be as sturdy as it is. But this is—yeah, it's called Curvahedra—look it up. It’s these quite sturdy—you know, you don't need to worry about ripping the pieces as you put them together—but you can create these things that look really intricate, and you can create positive curvature, or flat things, or negative curvature in all these different conformations. It's a very fun thing to play with.

    EH: And it is a sort of physical version of exactly the Gauss-Bonnet theorem. As you hook together pieces, you're controlling what happens on a loop. And then as you put more of those loops together, you can get a variety of different surfaces, from hyperbolic planes to spheres to—of course, kids have made animals and creatures with it. So you get this sort of control. In fact, it's one of those things that, you put it into the hands of kids, and they do things that you didn't think were really possible with it because their ability to play with these ideas and be free is always so inspiring. So that's what I said, this is a theorem that you can—people can understand as something in the real world. And then you can tell the story of how this understanding of the world is linked directly back to abstract, esoteric mathematics, of the most advanced sort.

    KK: Right. One of my favorite things about Curvahedra, though, is the video that you put online somewhere—I think was on Twitter—of it popping out of your suitcase, like you compressed it down into your suitcase to travel home one time?

    EH: Yes, I have a model that's about to a two-foot cube. And so you can’t travel with that easily, but it can compress very small. And that same object has been in my suitcase and other things several times, and it's now sitting in my office here.

    KK: That’s great fun. And also you've made similar models out of metal, correct?

    EH: Yes. So the basic system—not the big one you can crush down to put into suitcases.

    KK: No, certainly not.

    EH: I’ve made a couple of the spheres. And we're currently working on a proposal to go outside the Honors College at the University of Arkansas. That grew out of a course—it was a design that was created from Curvahedra and other inspirations—by a course I taught with Carl Smith, who is a landscape architect in our landscape architecture school. And so there's going to be—hopefully at some point there's going to be a 12-foot tall Curvahedra-style model outside the Honors College at University of Arkansas.

    KK: Very nice.

    EL: Nice.

    KK: Yeah, this has been great fun. Anything else we want to talk about?

    EL: Yeah, well, do you want to say a website or Twitter account or anything where people can find you online?

    EH: So I’m actually @Gelada on Twitter, and there is @Curvahedra, and my blog, which is very rarely updated, but has some nice stuff, is called Maxwell Demon.

    EL: Yeah, and can you spell your Twitter?

    EH: Yes, so Gelada is spelled G-E-L-A-D-A. They are baboons in Ethiopia, or it’s a cold beer in Brazil. I discovered that latter one after being on Twitter, and I regularly get @-ed by people in Brazil, who were not wanting to talk to me at all, but they're asking each other out for beers.

    EL: Ah.

    EH: And yeah, so then there's also curvahedra.com, where you can get that toy.

    EL: Cool. Thanks for joining us.

    KK: Yeah, thanks Edmund.

    EH: Thank you.

    [outro]

    On today’s episode, we were pleased to talk with Edmund Harris, a mathematician and mathematical artist at the University of Arkansas, who is our second guest to sing the praises of the Gauss-Bonnet theorem. Below are some links you might find useful as you listen to the episode.



    Edmund’s Twitter account, @Gelada


    His blog, Maxwell’s Demon


    The website and Twitter account for Curvahedra, the toys he makes that help you explore the Gauss-Bonnet theorem and just have a lot of good fun with geometry


    Our episode with Jeanne Clelland, who also chose the Gauss-Bonnet theorem


    Edmund and Evelyn both attended the Illustrating Mathematics program at the Institute of Computational and Applied Mathematics (ICERM). The program website, which includes videos of some interesting talks at the intersection of math and art, is here.

    32 min
  • Episode 48 - Sophie Carr
    Episode 48 - Sophie Carr
    Kevin Knudson & Evelyn Lamb
    Download

    Kevin Knudson: Welcome to My Favorite Theorem, a math podcast and so much more. I'm one of your hosts, Kevin Knudson. I'm a professor of mathematics at the University of Florida. And here is your other host.

    Evelyn Lamb: Hi, I'm Evelyn Lamb. I'm a freelance writer, usually based in Salt Lake City, but currently coming to you from Providence, Rhode Island.

    KK: Hooray! Yeah, you're at ICERM.

    EL: Yes. The Institute for computational and experimental research in mathematics, an acronym that I am now good at remembering.

    KK: I’m glad you told me. I was trying to remember what it stood for this morning because I'm going next week. We'll be in the same place for, like, only the second time ever.

    EL: Yeah.

    KK: And the universe didn't implode the first time. So I think we're safe.

    EL: Yeah.

    KK: So the ICERM thing is visualizing mathematics, I mean, we're sort of doing like—next week is about geometry and topology, which since both of us are nominally that, that's just the right place for us to be.

    EL: Yeah, it's it's going to be a fun semester. I'm also very excited because I recently turned in—it feels weird to call it a manuscript, but it is being published by a place that publishes books. It is the final draft of a page-a-day calendar about math. And I hope that by the time we air this, I will be able to have a link where people can purchase this and give it to give it to themselves or to their favorite mathematician.

    KK: Yeah.

    EL: So that's just, every day you can have a little morsel of math to start your morning.

    KK: I’m looking forward to that. That’s really exciting. Yeah, that's that's great. All right, so we're continuing a tradition in this episode.

    EL: Yes.

    KK: So Christian Lawson-Perfect organizes this thing through the Aperiodical called the Great Internet Math-Off [Editor’s note: Whoops, it’s called the Big Internet Math-Off!] of which you were a participant in the first one but not this one, not the second go-around. And we had the first winner on. The winner gets named the World's Most Interesting Mathematician (among those people who Christian could round up and who were free in July). And so we wanted to keep this trend going of getting the most interesting mathematicians in the world on this podcast. And we are pleased to welcome this year's winner, Sophie Carr. Sophie, you want to introduce yourself, please?

    SC: Oh, hello, thank you very much. Yeah, I'm Sophie Carr. I studied Bayesian networks at university, and now I own and run a data analytics company.

    EL: Yeah, and you’re the most interesting mathematician!

    SC: I am! For this year, I am the most interesting mathematician in the world. It's entirely Nira’s fault that I entered because he suggested, and put me forward.

    KK: That’s right. Nira Chamberlain was last year's winner. And so when we interviewed him he was sitting in his attic wearing a winter coat. It was wintertime and it seemed very cold where he was. You look very comfortable. It looks like you have a very lovely home in the background.

    SC: Yes, I mean, I am in two jumpers. Autumn has definitely arrived. Summer has gone, and it's a little chilly at the moment.

    KK: I can only dare to dream. Yeah.

    EL: Yeah, Florida and UK have slightly different seasons.

    KK: Just a little bit. So you own a consulting company? That’s correct?

    SC: Yeah, I do. I set it up 10 years ago now. There’s me and two other people who work with me. We just have an awful lot of fun finding patterns in numbers. I still find it amazing that we're still going. It's just the best fun ever. We get to go and work on all sorts of different problems with all sorts of different people. It's fantastic.

    KK: Yeah, that's great. I mean, I'm glad companies are starting to come around to the idea that mathematicians might actually have something to tell them. Right?

    SC: Yes. It really is. When you explain to them, you're not going to do magic and it's not a black box, and you can tell them how it works and how it can really make a difference, they are coming around to that.

    KK: That’s fantastic. All right, so we're here to talk about theorems.

    EL: Yeah. What is your favorite theorem?

    SC: My favorite theorem in the whole world is Bayes’ Theorem.

    EL: Yay, I'm so glad that someone will be talking about this! Because I know that this is a great theorem and—confession: I just, I don't appreciate it that much.

    KK: You know, same.

    EL: I need to be told why it's great.

    KK: Yeah, I taught probability one time and I said, “Okay, here's Bayes’ theorem.” I kind of went all right. Fine, but of course the question is what's the prior, Mr. Bates? So tell us. Tell us, please.

    EL: Yeah, preach!

    KK: Preach for Reverend Bayes.

    SC: You know, I don't think there's any any preaching needed. Because I always say this. I mean, there are two bits of statistics, there’s the frequentist and the Bayesian. And I always liken it to rugby union, and rugby league, which are two types of rugby in England. It's different codes, but it's the same thing. So to me, Bayes’ theorem, it's just the way that we naturally think. And it's beautifully simple, and all it does is let you take everything that you know and every piece of information that you have, and use that to update the overall outcome. And you're right, that the really big arguments come about from what the prior is. What is the background information that we have, and can we have actually genuinely have a true prior? And some people say no, because you might not have any information. But that's the great bit! Because then you can go and find out what the prior is. You have to be absolutely open about what you're putting in there. I think the really big debate comes around whether people are happy with uncertainty. Are they happy for you to not give an exact answer? If you go and you say, well, this is the prior, this is what we think the information is as well. And we combine these all, combine these priors, and this is the answer. Let's have a debate. Let's start talking about what we can have. Because at its simplest, you've got two things you’re timesing together. Just two numbers. Something that runs your mobile phone. I mean, that’s quite nifty.

    KK: So can we can we remind our listeners what Bayes’ theorem actually says?

    SC: Okay, so Bayes’ theorem takes two things. It takes the initial, or the prior distribution. Okay, and that's the bit where the argument is. And that might be just, what's the chance of something happening? What do you think the probability is of something happening? And you combine that with something called likelihood ratio. And it's real simple. The likelihood ratio is just a ratio of the probability of the information, or the evidence you have, assuming one hypothesis,divided by the probability of that information assuming another hypothesis. So you just have to have those two values. [And I say you just have to keep it.

    And then all you have to do is times them together! That really is it, and when you start to say to people, it's just two numbers—Now, you can turn that into three numbers if you want. You can turn the likelihood ratio bit into its two separate parts. And you can show Bayes’ theorem very, very simply with decision trees, and that was part of the reason I used decision trees in the Math-Off, was just to show the power of something that is really quite simple, that can drive so, so far. And that's what I love about Bayes’ theorem. I always describe it as something that is stunningly elegant, but unbelievably powerful. And I always liken it to Audrey Hepburn. I think if it were to be a person, it would be Audrey Hepburn. Quite small! I'd say it's, it's this amazing little thing that has two simple numbers. But goodness me, getting those numbers, well, I mean, you can just have so much fun! I think you can.

    And maybe it's just me that likes finding the patterns in the numbers and finding those distributions. Coming up with the priors. So come on, Kevin, you said, you sat there and your class said, “Well, what's the prior?”

    KK: Yeah.

    SC: What do you say? How would you tell people to go about finding a prior? Are they going to use their subjective opinion? Are they going to try and find it from data?

    KK: Well, that that is the question, isn't it? Right? So, I mean, often, the problem with probability sometimes is that—at least, like, in political forecasting, right—people tend to round up probabilities to 1 or lop them off to zero. Right? So for example, when, you know, when Trump won the election in 2016, everybody thought it was a huge shock. But you know, 538 had it as, you know, Hillary Clinton was a two-to-one favorite. But two-to-one favorites lose all the time, right?

    SC: Yeah.

    KK: And and so the question then is, yeah, people like to think about one-off events. And then the question is, how do you estimate the probability of a one time event? And you have to make some guess, right, at the prior. And that’s—I think that's where people get suspicious of Bayes’ theorem, or Bayesian statistics, because how you make this estimate? So how do you make estimates in your daily work as a consultant?

    SC: Okay, so we do it in a variety of different ways. So if we're really lucky, there’s some historical data we can go looking at.

    KK: Sure.

    SC: And often just mining that historical data gives you a good starting point. I always get slightly suspicious of flat distributions. Because if we really, really don't know anything other than that, I think maybe a bit of research before where you find the prior is always a good thing. My favorite priors are when we go and talk to people and start to get out of them their subjective opinion. Because I like statistics, I genuinely love statistics, because of the debate that goes on around it. And I think one of the things that people forget about math is that it's such a living subject. And there are so many brilliant debates—and you can call some of them arguments— people are prepared to go and say, “Look, this is my opinion and this is what I think the shape is.” And then we can do the analysis. Inevitably somebody will stand up and go, “Well, that bit is wrong.” Okay, so tell me why!

    EL: Yeah.

    SC: What evidence have you got for us to change the shape, or why do you think it should be skewed, or Poisson, or whatever we're using? And sometimes, if we haven't got time to do that we can start to put in flat distributions. We can say, “Well, we think it's about normal.” Or “We think on average, it'll be shoved a little bit to the right or a little bit to the left.” That's the three main ways we go about doing it. And I think the ability to be absolutely open and up front about what you know and what you don’t know helps you find that prior. And I don't really understand why people would be scared of running away from that. Why you would not want to say what the uncertainty is or what you're not sure about. But that might go a long way when people think that math is certain.

    EL: Yeah.

    SC: That when you say the answer is 12, well it’s 12. And not, “Well, it’s 12 because we kind of do it like this, and actually if something changes, that number might change.” And I think getting comfortable with uncertainty and being uncomfortable, is really the crux for developing those priors.

    EL: Yeah. Well, I guess for me, it's hard to reason about statistics in a non frequentist way. Meaning—you know, I'm comfortable with non frequentist statistics to a certain degree. But just like what, as you were, saying, like, what does a 30% chance mean if it's not that we could do this 10 times that have it happen three times. But you can't have a presidential election—the same election—10 times, or you can't run Monday’s weather 10 times, or something like that. But it's just hard for me to interpret what does it mean if there isn't a frequentist interpretation?

    SC: Yeah. One of the things we found that works really well is if you start showing patterns—and that's why I always talk about patterns, that we find patterns. It's when you're doing Bayesian stats with priors if you start to show the changes as curves, and I don't mean the distribution, but I mean, just as that rising and falling of numbers, people start to understand what's driving the priors, what assumptions are changing those priors. And then you start to see the impact of that, how the final answer changes. That can be incredibly powerful. Often people don't want that set answer. They want to know what the range is, they want to understand how that changes. And showing that impact as a shape—because I think most people are visual. When you show somebody a surface or, you know, a graph, or whatever it is, that's something you can really get a grip with. And actually I come from a Bayesian belief network. So I kind of found out about Bayes’ theorem by chance. I never set off to learn Bayes’ theorem. I set off to design [unintelligible]. That’s what I grew up wanting to do. But I ended up working on Bayesian networks. That’s the short version of what happened.

    EL: So, how—was this a “love at first sight” theorem? Or what was your initial encounter with this theorem? And how did you feel about it? Since this is all about subjective feelings anyway!

    SC: Well, my PhD was part-time. I spent eight years collecting subjective opinions. So I started a PhD in Bayesian networks, and there was this brilliant representation of a great big probability table. And this is a while ago now. And I’ve moved on a lot into [unintelligible]. But I've got this Bayesian network and supervisor said, “Here we go,” and I went, “Ah, it’s just lots of ovals connected with arrows”

    And I went, “There must be something more to this.” And he went, “There’s this thing called Bayes’ theorem that underpins it and look at how it flows. It’s how the information affects it.” And I went, “Okay!” And so, as with all PhDs, you have this pile of reading, which is apparently going to be really, really good for you.

    So I got my pile of reading. I went, “Okay.” And genuinely I just thought, “Yeah, it's just kind of how we all work, isn't it?” And I really had not liked statistics at university at all because I’d only really done frequentist statistics. And it’s not like I dislike frequentist statistics. I just didn’t fall in love with it. But when there was something I could see—and I genuinely think it’s because it's visual. I see the shapes move, I could see the numbers flow, I could see the information flow. I thought, “Oh, this is cool stuff. I understand this. I can get my head around this.” And I could start to see how to put things in and how they changed. And I think also I've got at times a very short attention span. So running millions of replicates never really did it for me.

    EL: Yeah.

    SC: So I had a bit of an issue with frequentist, where we just have to run lots and lots and lots lots of replicates.

    EL: Right.

    SC: Can we not assume it's kind of like this shape and see what happens? Then change that shape. Look, that’s great. That's much better for me.

    EL: Yeah. So it was kind of a conversion experience there.

    SC: I think, for people my age, probably. Because I don’t think Bayesian statistics, years ago, was taught that commonly. it's only really in the past sort of maybe decade that I think it's become really mainstream and been taught in the way it is now. Certainly with its its wide applications. That's what I think people just go, something that they've never heard of is now all in the AI world and it’s in your mobile phone, and it's in your medicine, and it's in your spam filters. And when it suddenly becomes really popular, people start to see what it can do. That's when it's taught more. And then you get all these other debates.

    KK: So the other fun thing we like to do on this podcast is ask our guests to pair their theorem with something. So what pairs well with Bayes’ theorem?

    SC: So this caused a lot of debate in our household.

    KK: It always does.

    SC: Yeah. And I am going to pair Bayes’ theorem with my favorite food, which is risotto, because risotto only takes three things. It only needs rice and onions and a good stock.

    KK: Yes.

    SC: And Bayes’ theorem is classically thought with three numbers. And it’s really powerful and gorgeous. And risotto only takes three ingredients, and it’s really gorgeous.

    KK: And also, the outcome is uncertain sometimes, right?

    SC: Oh, frequently uncertain. And if you change those prior proportions, you will get a very different outcome.

    KK: That’s right. You might get soup, or it might might burn.

    SC: So, I am going to say that Bayes' theorem is like a risotto.

    EL: And you mentioned Audrey Hepburn earlier so maybe it’s even more like sharing a risotto with Audrey Hepburn.

    SC: That would be brilliant. How cool would that be?

    EL: I know!

    SC: I will have my Bayes’ theorem discussion with Audrey Hepburn over risotto. That would be a pretty good day.

    EL: Yeah, you could probably get a cardboard cutout. Just, like, invite her to dinner.

    SC: Yeah, I'll do that. I'll try and set up a photo, superimpose them.

    EL: Yeah.

    KK: But Audrey Hepburn should be breakfast somewhere right?

    EL: But you can eat risotto for breakfast.

    SC: Yeah, you can eat risotto any time of the day.

    KK: Sure.

    SC: There’s never a bad time for risotto.

    KK: No, there isn't. Yeah. My wife actually doesn't like risotto very much, so I never make it.

    EL: So is that one of your restaurant foods? So we have this whole like foods that you you tend to order at a restaurant because your partner doesn't like them. And so it's like something that you can—like I don't really like mushrooms, so my partner often will order a mushroom thing at a restaurant.

    KK: Yeah, so for me, I don't go out for Italian food because I can make it at home.

    EL: Okay.

    KK: So I just have a generic I don't I don't eat Italian out. There’s kind of no point, I think.

    SC: So you’re right that risotto is my restaurant food because my husband doesn't like it.

    KK: Oh.

    EL: Aw.

    SC: It's my most favorite thing in the world, so yeah, every time we go out, the kids go, “Mom, just don't get the menu. There’s no point. We know what you’re getting.

    EL: Yeah. So you said this caused a debate. Did he have a different opinion about what your pairing should be?

    SC: Well, there were discussions about whether it was my favorite drink with [a bag of crisps?], and what things could be combined together. And I said, “No, it just has to be risotto.”

    KK: Okay. Excellent.

    EL: Yeah, we do make that at home. And actually the funny thing is I don't really like mushrooms, but I do like the mushroom risotto that we make.

    SC: Oh.

    EL: Yeah.

    SC: So you've not got a flat prior. You've actually got a little bit of a skew on there.

    EL: Yeah, I guess. I’m trying to figure out how to quantify this. Yeah, like my prior distribution for mushroom preference is going to depend on whether it is cooked with arborio rice or not.

    SC: See, there we go and you don’t have to worry about numbers you just draw a shape.

    EL: Yeah, nice.

    KK: Cool. So we also like to give our guests a chance to plug anything they want to plug. Do you have things out there in the world that you want people to know about?

    SC: So the only thing I think that's worth mentioning is I do some Royal Institution maths master classes, where we go out and we take our favorite bit of math, and we go and take it to students who are between the ages of about 14 to 17. And that's really what I'm doing coming up in the near future, and they are a brilliant way for lots of people to engage with maths.

    EL: Oh, nice.

    KK: That’s very cool.

    SC: Yeah. They are really good fun.

    KK: Have you been doing that for very long?

    SC: I’ve been doing them for about two years now. And the first one I ever did was on Bayes’ theorem. And I've never been so terrified, because I don’t teach. And then you have this group of students, and they come up with just the best and most fantastic questions. Every time you do it, you go, “I hadn’t thought of that.”

    KK: Yeah.

    SC: “And I don't know how to answer that question straight away.” So it's brilliant, and I love doing them. So that's kind of what we've got coming up. And you know, work is just going to be keeping me nicely busy.

    EL: Nice.

    SC: Yeah.

    KK: Well, this has been great fun. Thank you for joining us, and congratulations on being the world's most interesting mathematician for this year.

    EL: Yes. Yeah, thanks a lot.

    SC: Thank you. I’ve been so excited to do this. I've been listening to your podcast for quite a long time, and I couldn't believe it when you emailed.

    okay, thank you very much.

    Okay. Thanks.

    On this episode, we had the pleasure of talking with Sophie Carr, a statistics consultant and winner of Christian Lawson-Perfect’s Big Internet Math-Off last summer. Here are some links you may enjoy as you listen to this episode.

    As we mentioned at the top of the show, Evelyn’s math page-a-day calendar is available for purchase in the AMS bookstore!
    Sophie Carr’s twitter account
    The Big Internet Math-Off at the Aperiodical
    Royal Institution Masterclasses
    Sophie Carr is this year’s World’s Most Interesting Mathematician. We also had last year’s World’s Most Interesting Mathematician, Nira Chamberlain, on the show in January. Find his episode here.

    24 min
  • Episode 47 - Judy Walker
    Episode 47 - Judy Walker
    Kevin Knudson & Evelyn Lamb
    Download

    Kevin Knudson: Welcome to My Favorite Theorem, a podcast about mathematics and all kinds of crazy stuff, and I have no idea what it's going to be today. It is a tale of two very different weather formats today. So I am Kevin Knudson, professor of mathematics at the University of Florida. Here's your other host.

    Evelyn Lamb: Hi, I'm Evelyn Lamb. I'm a math and science writer in Salt Lake City, Utah, where I am using the heater on May 28.

    KK: Yes, and it's 100 degrees in Gainesville today, and I'm miserable. So this is bad news. Anyway, so today, we are pleased to welcome Judy Walker. Judy, why don't you introduce yourself?

    Judy Walker: Hello. Thank you for having me. I'm Judy Walker. I'm a professor of mathematics at the University of Nebraska.

    KK: And what else? You're like—

    JW: And I am Associate Vice Chancellor for faculty and academic affairs, so that’s, like, Vice Provost for faculty.

    KK: That sounds—

    EL: Yeah, that does sound very official!

    JW: It does sound very official, doesn't it?

    KK: That’s right. Like you're weighing T & P decisions in your hands. It's like, you're like Caesar, right? With the thumbs up and the—

    JW: I have no official power whatsoever.

    KK: Right.

    JW: So yes.

    KK: But, well, your power is to make sure procedures get followed, right?

    JW: Yes. And I have a lot of I have a lot of influence on other things.

    KK: Yeah. Right. Yeah. That sounds like a very challenging job.

    JW: And for what it's worth, I will add that it is cloudy and windy today. But I think we're supposed to be, like, 67 degrees. So right in the middle.

    KK: All right. Great.

    EL: Okay, perfect.

    KK: So if we could see the map of the US, there'd be these nice isoclines. And here we are. Right. So we're, my mine is very hot. Mine's red. So we're good. Anyway, we came to talk about math. You’re excited to talk about math for once, right?

    JW: Exactly. I guess I'm kind of going to be talking about engineering, too. So—

    EL: Great.

    KK: That’s cool. We like it all here. So what's your favorite theorem?

    JW: So my favorite theorem is the Tsfasman-Vladut-Zink theorem.

    KK: Okay, that's a lot of words.

    JW: It is—well, it’s a lot of names. It's three names. And it's a theorem that is in error-correcting codes, algebraic coding theory. And it's my favorite theorem, because it solves a problem, or maybe not solves a problem, but shows that something's possible that people didn't think necessarily was possible. And the way that it shows that it's possible is by using some pretty high-powered techniques from algebraic geometry, which had not previously been brought into the field at all.

    EL: So what is the basic setting? Like what kind of codes can you correct with this theorem?

    JW: Right. So the codes are what does the correcting. We don't correct the codes, we use the codes to correct. So I used to tell my — actually, my advisor told me and then I've told all my PhD students — that you have to have a sentence that you start everything with. And so my sentence is: whenever information is transmitted across a channel, errors are bound to occur. So that is the setting for coding theory. You've got information that you're transmitting. Maybe it's pictures from a satellite, or maybe it's just storing things on a computer, or whatever, but you're storing this information. Or you're transmitting this information, and then on the other end, or when you retrieve it, there's going to be some mistakes. And so it's the goal of coding theory to add redundancy in such a way that you can find those mistakes and fix them. Okay?

    And we don't actually consider it an error if you fix the mistake. So an error is when so many mistakes happened in the transmission or in the storage and retrieval, that what you think was sent was not what was actually sent, if that makes sense.

    KK: Sure. Okay.

    JW: So that's the basic setting for coding theory, and coding theory kind of started in 1948 with Shannon's theorem.

    KK: Right.

    JW: So Shannon's theorem says that reliable communication is possible. So what it says really, is that whatever your channel is, whether it's transmitting satellite pictures, or storing data, or whatever—whatever your channel is, there is a kind of maximum efficiency that's possible on the channel. And so what Shannon’s theorem says is that for any efficiency up to that maximum, and for any epsilon greater than zero, you can find a code that is that efficient has less than epsilon probability of error, meaning the probability that what you sent is not what you think was sent at the end. Okay?

    So that's Shannon's theorem. Right? So that's a great theorem.

    EL: Yeah.

    JW: It’s not my favorite theorem. It’s not my favorite theorem because it actually kind of bothers me.

    KK: Why does it bother you?

    JW: Yeah, so the reason that bothers me are — there are two reasons it bothers me. One is that it doesn't tell us how to find these codes. It says good codes exist, but it doesn't tell us how to find them, which is kind of useless if you're actually trying to transmit data in a reliable way. But it's actually even worse than that. It's a probabilistic proof. And so it doesn't just say that good codes exists, it says they're everywhere, but you can't find them. Right? So it's like it's taunting us. So I just—. So yeah. So that's Shannon's theorem. And that's why it's not my favorite theorem. But why it's a really great theorem is that it started this whole field. So the whole field of coding theory has been — or of channel coding, at least, which is what we've been talking about is to find those codes, and not just find them, but find them along with efficient decoding algorithms for them. And so that's Shannon's challenge is to find the good codes with efficient decoding algorithms for those good codes. That's 1948, that that started. Right? Okay.

    So just as a digression, let me say that most mathematicians and engineers will agree that at this point in time — so a little more than 70 years after Shannon's theorem, that Shannon's challenge has been met, so that we can find these good codes. They're not going to agree on how it's been met. But they'll all agree that it has been met. So on the one hand, in the late ‘90s — mid-to-late 90s — engineers found turbo codes, and they rediscovered low-density parity check codes. And these are codes that in simulations come very, very close to meeting Shannon's challenge. The theory around these codes is still being developed. So the understanding of why they meet Shannon challenge is still try to be solved. But the engineers will say that it's solved, that Shannon's challenge is met because they've got these simulations, and they're so confident about it, that these codes are actually being used in practice now.

    EL: So I have a naive question, which is, like, does the existence of us talking over the internet on on this call, sort of demonstrate that this has been met? Like we we are hearing each other — I mean, not with perfect fidelity, but we're able to transmit messages. Is that? Or is that just not even in the same realm?

    JW: No, that's exactly exactly what we're talking about, exactly what we're talking about. And not only that, but I don't know if you've noticed, but every once in a while, Kevin gets a little glitchy, and he doesn't move for a while. That's the code catching up and fixing the errors.

    KK: Yeah, that's the irony is this this call has been very glitchy for me.

    JW: Right.

    KK: Which is why we each record our own channel.

    EL: Yeah.

    JW: Exactly. So in fact, low-density parity-check codes and turbo codes are being used now in mobile phones, in satellite communications, in digital TV, and in Wi-Fi. So that's exactly what we're using.

    EL: Okay.

    JW: But the mathematicians will say, “Well, it's not really—we’re not really done. Because we don't know why. We don't really understand these things. We don't have all the theoretical underpinnings of what's going on.” A lot of work has been done a lot, and a lot of that is there. But it's still a work in progress. About 10 years ago, kind of on the flip side, polar codes were discovered. And polar codes are the first family of codes to provably achieve capacity. So they actually provably meet Shannon's challenge. But at this moment, they are unusable. There's just still a lot of work to understand how we can actually use polar codes. So the mathematicians say, “We've met the challenge, because we've got polar codes, but we can't use them.” And the engineers say, “We've met the challenge because we've got turbo codes and LDPC codes, but we don't know why.” Right? And that's an oversimplification, but that's kind of the current state. And so different people are working on different things now. And of course, there are other kinds of coding that that aren’t — that isn't really channel coding. There are still all kinds of unsolved problems. So if anybody tells you that coding theory is dead, tell them they're wrong.

    EL: Okay!

    JW: It’s still very much alive. Okay, so we talked about Shannon's theorem from 1948. And we talked about the current status of coding theory. And my favorite theorem, this Tsfasman-Vladut-Zink, is from 1982. So in the middle.

    EL: Halfway in between.

    JW: Yes, yes. Just like my weather being halfway in between. Yes. So around this time, in the early ‘80s, and and preceding that, the way that mathematicians were approaching Shannon's challenge was through the study of linear codes. So linear codes are just subspaces, and we might as well think of—in a lot of applications, the data is zeros and ones. But let's go to Fq instead of F2, so q is any prime power.

    KK: Okay, so we're doing algebraic geometry now, right?

    JW: We’re not yet. Right now, we’re just talking about finite fields.

    KK: Okay.

    JW: We will soon be be doing algebraic geometry, but not yet. Is that okay?

    EL: You’re just trying to transmit some finite set of characters.

    JW: Yes, some finite string of characters. Order matters, right? So it's a string. And so the way that we think about it, we can think about it as a systematic code. So the first k characters are information, and then we're adding on n−k redundancy characters that are computed based on the first k.

    KK: Okay.

    JW: So if we're in a linear setting, then this collection of code words that include the information and the redundancy, that collection of code words is a subspace, say it's a k-dimensional subspace, of Fqn. So that's a linear code. And we can think about that ratio, k/n, as a measure of how efficient the code is.

    KK: Right.

    JW: Because it's the number of information bits divided by the total number of bits, or symbols, or characters. So, let's call that ratio, R for rate, right? k/n, we’ll call it R. And then how many errors can the code correct? Well, if you look at that Hamming distance—so that's the number of characters and number of positions in which to code words differ—then the bigger that distance, the more errors you can make and still be closest to the code word that was sent. So then that's not really an error. Right? So maybe we say the number of mistakes goes up.

    EL: Yeah. So again, let's normalize that minimum distance of the code by dividing by the length of the code. So we have a ratio, let's call that ∂. So that's our relative minimum distance for the code. So one way to phrase this is if we want a certain error-correcting capability, so a certain ∂, how efficient can the code be? How big can R be? Okay, so there are a lot of bounds relating R and ∂, our information rate and our error-correcting capability, or our relative minimum distance. So one that I want I tell you about is that Gilbert-Varshamov bound.

    So the Gilbert-Varshamov bound is from 1952. And it says that there's a sequence of codes, or a family of codes if you want, of increasing length, increasing dimension, increasing minimum distance, so that the rate converges to R and the minimum distance to converges to ∂. And R is at least 1−Hq(∂), where Hq is this entropy function. So you may have heard of the binary entropy function, there's a q-ary entropy function, that's what Hq(∂) is. So one such sequence is the so-called classical Goppa codes, and I want to say that that's from, 1956, so just a little bit later. And those codes were the best-known codes from this point of view for about 30 years. Okay, so let me just say that again. So the Gilbert-Varshamov bounds says that there's a sequence of codes with R at least 1−Hq(∂). The Goppa codes satisfy r=1−Hq(∂). And for 30 years, we couldn't find any codes with R greater than.

    EL: That were better than that.

    JW: Right. That were greater than this 1−Hq(∂).

    KK: Okay.

    JW: So people at this point were starting to think that maybe the Gilbert-Varshamov bound wasn't a bound as much as it was the true value of how good can R be given ∂, how efficient can codes be given given their relative minimum distance. So this is where this Tsfasman-Vladut-Zink theorem comes in. So in 1978—and Kevin, now we can talk about algebraic geometry. I know you’ve been waiting for that.

    KK: All right, awesome.

    JW: Yes. Right. So in 1978, Goppa defined algebraic geometry codes. So the way that definition works: remember, a code is just a subspace of Fqn, right? So how are we going to get a set of space of Fqn? Well, what we're going to do is we're going to take a curve defined over Fq that has a lot of rational points, Fq-rational points, right? So we're going to take one of those points and take a multiple of it and call that our divisor on the curve. And then we're going to take the rest of them. And we're going to take the rational functions in this space L(D). D is our divisor, right? So these are the functions that only have poles at this chosen point of multiplicity, at most the degree that we've chosen.

    KK: Okay.

    JW: And we're going to evaluate all those functions at all the rest of those points. So remember, those functions form a vector space, and evaluation is a linear map. So what we get out is a vector space. So that's our code. And if we make some assumptions, so if we assume that that degree of that divisor, so that multiplicity that we've chosen, is at least twice the genus minus 2, twice the genus of the curve minus 2, then Riemann-Roch kicks in, and we can compute the dimension of L(D). But if we also assume that that degree is less than the number of points that we're evaluating at, then the map is injective. And so we have exactly what the dimension of the code is. The dimension of the code is the degree of the divisor, so that multiplicity that we chose, plus 1 minus the genus. And the minimum distance, it turns out, is at least n minus the degree of the divisor. So lots of symbols, lots of everything.

    EL Yeah, trying to hold this all in my mind, without you writing it on the board for me!

    JW: I know, I’m sorry. But when you put it all together, and you normalize out by dividing by the length, what you get is that if you have a family of curves with increasing genus, and an increasing number of rational points, then we can end up with a family of codes, so that in the limit, R, our information rate, is at least 1−∂—that’s that relative minimum distance—minus the limit of the genus divided by the number of rational points. Okay. So g [the genus] and n are both growing. And so what's that limit? So that's that was Goppa’s contribution. I mean, not his only contribution. But that's the contribution of Goppa I want to talk about, just that definition of algebraic geometry code. So it's a pretty cool definition. It’s a pretty cool construction. It’s kind of brand new in the sense that nobody was using algebraic geometry in this very engineering-motivated piece of mathematics.

    EL: Right.

    JW: So here is algebraic geometry, here is a way of defining codes, and the question is, are they any good? And it really depends on what—how fast can the number of points grow, given how fast the genus is growing? So what Drinfeld and Vladut proved—so this is not the TVZ theorem, not my favorite theorem, but one more theorem to get there—Drinfeld and Vladut proved that if you take, if you define Nq(g) to be the maximum number of Fq-rational points on any curve over Fq of genus g, then as you let g go to go to infinity, and for a fixed q, the limit superior, the lim sup, of the ratio g/Nq(g), is at most 1/√(q−1). Okay, fine. Why do we care? Well, the reason we care is that the Tsfasman-Vladut-Zink theorem, which is again my favorite theorem, it says—so actually, my favorite theorem is a corollary of the Tsfasman-Vladut-Zink theorem. So the Tsfasman-Vladut-Zink theorem says that if q is a square prime power, then there's a sequence of curves over Fq of increasing genus that meets the Drinfeld-Vladut bound.

    EL: Okay.

    JW: Okay, so the Drinfeld-Vladut bound said you can be at most this good. And Tsfasman-Vladut-Zink says, hey, you can do that.

    EL: Yeah, it's sharp.

    JW: So if we put it all together, then the Gilbert-Varshamov bound gave us this curve, right? So it was a concave-up curve that intersects the vertical axis, which is the R-axis, at 1 and the horizontal axis, which is the ∂-axis, at 1−1/q. So it's this concave-up thing that's just kind of curving out. Then the Tsfasman-Vladut-Zink line—the theorem gives you a line that looks like R=1−∂−1/√(q−1). Right? So it's just a line of slope −1, right, with y-intercept 1−1/√(q−1). So the question is, does that line intersect that curve? And it turns out that if you have q, a square prime power q at least 49, then the line intersects the curve in two points.

    EL: Okay.

    JW: So what that is really doing for us is it's telling us that in that interval between those two points, we have an improvement on the Gilbert-Varshamov bound. We have better codes than we thought were possible for 30 years.

    EL: Wow!

    JW: Yes. So that's my, that's my favorite theorem.

    KK: I learned a lot.

    EL: And where did you first encounter this theorem?

    JW: In graduate school? Okay, in graduate school, which was not in 1982. It was substantially after that, but it was said to me by my advisor, “I think there's a connection between algebraic geometry and coding theory, go learn about that.”

    KK: Oh.

    JW: And I said, “Okay.”

    KK: And so two years later.

    JW: Right. Right, right. Actually, two years later, I graduated.

    KK: Okay. All right. So you’re much faster than I am.

    JW: Well, there was four years before that of doing other things.

    EL: So was it kind of love at first sight theorem?

    JW: Very much so. Because I mean, it's just so beautiful, right? Because here's this problem that nobody knew how to solve, or maybe everybody thought was solved. Because nobody had any techniques that could get any better than the Gilbert-Varshamov bound. And then here's this idea, just way out of left field saying, hey, let's use algebraic geometry to find some codes. And then, hey, let's look at curves with many points. And hey, that ends up giving us better codes than we thought were possible. It's really, really pretty. Right? It's why mathematicians are better than electrical engineers.

    EL: Ooh, shots fired!

    JW: Gauntlet thrown. I know.

    EL: But it does make you wonder how many other things in math will eventually find something like this, like, will will find for these problems—you know, factoring integers or things like this— that we think are difficult, will someone swoop in with some completely new thing and throw it on its head?

    JW: Yes. Exactly. I mean, I don't know anything about it. Maybe you do. But the idea that algebraic topology, right, is useful in big data.

    KK: Yeah, sure. That's what I've been working on lately. Yeah. Right.

    JW: I love that.

    KK: Yeah. Sure.

    JW: I love that. I don't know anything about it. But I love it.

    KK: Well, the mantra is data has shape. Right? So let me just, you know, smack the statisticians here. So they want to put everything on a straight line, right? But a circle isn't a straight line. So what if your data’s a circle? So topology is very good at finding circles.

    JW: Nice.

    KK: Well, that's the mantra, at least. So yeah. All these unexpected connections really do come up. I mean, it's really—that’s part of why we keep doing what we're doing, right? I mean, we love it. But we never know what's out there. It's, you know, to boldly go where no one has gone before. Right?

    JW: Exactly. And Evelyn, it's funny that you should bring up factoring integers, because you know that the form of cryptography that we use today to make it safe to use our credit cards on the internet, that’s very much at risk when quantum computers are developed.

    EL: Right.

    JW: And so, it turns out that algebraic geometry codes are not being used in practice, because LDPC codes and turbo codes are much more easily implementable. However, one of the very few known so far unbreakable methods for post-quantum cryptography is based on algebraic geometry codes.

    KK: Excellent.

    EL: Nice.

    JW: So even if we can factor integers,

    KK: I can still buy dog food at Amazon. Right?

    JW: You can still shop at Amazon because of algebraic geometry codes.

    EL: Yeah, the important things.

    KK: That’s right.

    EL: Well, so another thing we like to do on this podcast is invite our guests to pair their theorem with something, the way we would pair food with fine wines. So what have you chosen for this theorem?

    JW: So that was very hard. Yeah. I mean, it's just kind of the most bizarre request.

    EL: Yeah.

    JW: So I mean, I guess the way that I think about this Tsfasman-Vladut-Zink theorem, I was looking for something that was just, you know, unexpected and exciting and beautiful. But I couldn't come up with anything. And so instead, what I'm going with is lemon zest.

    KK: Okay.

    EL: Okay.

    JW: Which I guess can be unexpected and exciting in a dessert, but also because of the way that you just kind of scrape it off that curve of the lemon. And that's what the Tsfasman-Vladut-Zink theorem is doing, is it’s scraping off a little bit of that Gilbert-Varshamov curve.

    KK: This is an excellent visual. I've got it. I zest lemons all the time. I understand now. This is it.

    EL: Yeah.

    JW: There you go.

    KK: So all right. Well, we also like to give our guests a chance to plug anything. You wrote a book once. Is that still right? I have it on my shelf.

    JW: Yeah. I did write a book once. So that book actually was—Yeah, so I wasn't going to plug anything, but I will plug the book a little bit, but more I'm going to plug a suite of programs. So the book is called, I think, Codes and Curves.

    KK: That sounds right.

    JW: You would think I would know that.

    KK: I’d have to find it. But it is on my shelf.

    JW: Yes. It's on mine too, surprisingly, which is right behind me, actually, if you have the video on.

    So that book really just a grew out of lecture notes from lectures I gave at the program for women and mathematics at the Institute for Advanced Study. Okay, so I will take my opportunity to plug something to plug that program, to plug EDGE, to plug the Carleton program, and to plug the Smith post-bac program, and to plug the Nebraska conference for undergraduate women in mathematics. So what do all these programs have in common they have in common? They have in common two things that are closely related. One is that they are all programs for women in mathematics. And the other is that they were all the subject of study of a recent NSF grant that I had with Ami Radunskaya and Deanna Haunsperger and Ruth Haas that studied what are the most important or effective aspects of these programs and how can we scale them?

    EL: Oh, nice.

    JW: Yes. And some of the results of that study, along with a lot of other information, are on our website. That is women do math.org?

    EL: I will be visting it as soon as we get off this phone call.

    JW: Right. Awesome. I hope it's functioning

    KK: And because Judy won't promote herself, I will say, you know, she's been a significant leader in promoting programs for women in mathematics through the University of Nebraska’s math department there. There's a picture of her shaking Bill Clinton's hand somewhere.

    JW: Well, that's also on my shelf. Okay. Yeah, I think it's online somewhere, too.

    KK: Right. Their program won a national excellence award from the President. Really excellent stuff there at the University of Nebraska. Really a model nationally.

    EL: Yeah, I’m familiar with that as one of the best graduate math programs for women.

    JW: Thank you.

    EL: Yeah. Great job!

    EL: Yeah, well, we'll have links to all of those programs on the website. So if you didn't catch one, and you're listening, you can to the website for the podcast and find all those. Yeah. Well, thank you so much for joining us, Judy.

    JW: Thank you for the opportunity.

    KK: Yeah, this has been great fun. Thanks.

    JW: All right. Thank you.

    On this episode, we were happy to talk with Judy Walker, who studies coding theory at the University of Nebraska. She told us about her favorite theorem, the Tsfasman-Vladut-Zink theorem. Here are some links to more information about topics we mentioned in the episode.


    Goppa (algebraic geometry) code

    Hamming distance

    Gilbert-Varshamov bound

    Judy Walker’s book Codes and Curves

    The Program for Women and Mathematics at the Institute for Advanced Study

    EDGE 

    The Carleton Summer Mathematics Program for women undergraduates

    The Smith College post-baccalaureate program for women in math

    The Nebraska Conference for Undergraduate Women in Mathematics (Evelyn will be speaking at the conference in 2020)

    WomenDoMath.org

    32 min
  • Episode 46 - Adriana Salerno
    Episode 46 - Adriana Salerno
    Kevin Knudson & Evelyn Lamb
    Download

    Evelyn Lamb: Hello, and welcome to My Favorite Theorem, a math podcasts where there's no quiz at the end. I’m coming up with a new tagline for it.

    Kevin Knudson: Good.

    EL: I just thought I'd throw that in. Yeah, so I'm one of your hosts, Evelyn Lamb. I'm a freelance math and science writer from Salt Lake City—or in Salt Lake City, Utah, not originally from here. And here's your other host.

    KK: I’m Kevin Knudson, professor of mathematics at the University of Florida in Gainesville, but not from Gainesville. This is part of being a mathematician, right? No one lives where they're from.

    EL: Yeah, I guess probably a lot of professions could say this, too.

    KK: Yeah, I don’t know. It’s also a sort of a generational thing, right? I think people used to just tend to, you know, live where they grew up, but now not so much. But anyway.

    EL: Yeah.

    KK: Oh, well, it's okay. I like it here.

    EL: Yeah. I mean, it's great here right now it's spring, and I've been doing a ton of gardening, which always seems like such a chore and then I'm out smelling the dirt and looking at earthworms and stuff, and it's very nice.

    KK: I’m bird watching like crazy these days. Yesterday, we went out and we saw the bobolinks were migrating through. They're not native here, they just come through for, like, a week, and then they're gone.

    EL: The what?

    KK: Bobolinks, B-O-B-O-L-I-N-K. They kind of fool you, they look a little bit like an oriole, but the orange is on the wrong side. It's on the back of the neck instead of underneath.

    EL: Okay, I'll have to look up a picture of that later.

    KK: And then this morning for the first time ever, we had a rose-breasted grosbeak at our feeder. Never seen one before and they're not native around here, they just migrate through. So this is

    EL: Very nice. Yes.

    KK: This is what I'm doing in my late middle age. This is what I do. I just took up bird watching, you know?

    EL: Yeah. Well, I can see the appeal.

    KK: Yeah, it's great.

    EL: Yes. But we are excited today to be talking with Adriana Salerno. Do you want to introduce yourself?

    Adriana Salerno: Hi. Yeah, I'm Adriana Salerno. Now I am an associate professor of math at Bates College in Maine. And I am also not from Maine. I live in Maine. I'm originally from Caracas, Venezuela, so quite a ways away.

    EL: Yeah.

    AS: Again, you don't choose where you live, but maybe you get to choose where you work. So that's nice.

    EL: Yeah. And you're you're not only a professor there, but you're also the department chair right now, right?

    AS: Oh, yeah. Yeah, I'm trying to forget. No, I’m kidding.

    EL: Sorry!

    KK: You know, speaking of, before we we started recording here, I spent my afternoon writing annual faculty evaluations. I’m in the first year as chair. I have 58 of them to write.

    AS: Oh, I don't have to do those, which I'm very happy about. But we are hiring a staff position, and I'm in charge of that. And that's been a lot.

    EL: And we actually met because both of us have done this mass media fellowship for people interested in math or science and writing. And so you've done a lot of writing not for mathematicians as well, throughout your career path.

    AS: Yeah, yeah. I mean, I did the mass media fellowship in 2007. And since then, I've been trying to write more and more about mathematics for a general audience. These days, I mostly spend time writing for blogs for the AMS. And right now I'm editing and writing for inclusion/exclusion. I wish I had more time to write than I do. It's one of those things that I really like to do, and I don't think I do enough of, but these opportunities are great because I get to use those—or scratch that itch, I guess, by talking to you all.

    EL: Yes.

    KK: Well, so speaking of, we assume you have a favorite theorem that you want to tell us about. What is it?

    AS: Well, so it's always hard to decide, right? But I guess I was inspired by a conversation I had with Evelyn at the Joint Math Meetings. So I've decided my favorite theorem is Cantor's diagonalization argument that the real numbers do not have the same cardinality as the natural numbers.

    EL: Yes, and I’m so excited about this! Ever since we talked at the Joint Meetings, I’ve been very excited about getting you to talk about this.

    AS: Good. Good.

    EL: Because really, it’s such a great theorem.

    AS: Yeah. Well, I was thinking about it today. And I'm like, how am I going to explain this? But I have chosen that, and I'm sticking with it. Yeah.

    EL: Yes.

    KK: Good.

    AS: So yeah, it’s—one of the coolest things about it is sort of it’s this first experience that you have, as a math student—at least it was for me—where you realize that there are different sizes of infinity. And so another way of saying that is that this theorem shows, without a doubt, I believe—although some students still doubt me after we go over it—it shows that you can have different sizes of infinity. And so the first step, even, is to say, “How do you decide if two things have the same size of infinity?” Right? And so it's a very, very lovely sort of succession of ideas. And so the first thing is, how do you decide that two things are the same size? Well, if they're finite, you count them, and you see that you have the same number of things. But even when things are finite—and say, you're a little kid, and you don't know how to count—another way of saying there's the same number of things is if you can match them up in pairs, right? So you know, if you want to say I have the same number of crayons as I have apples, you can match a crayon to an apple and see that you don't have anything left over, right?

    EL: Yeah.

    AS: And so it's just a very natural idea. And so when you think about infinite sets—or not even infinite sets—but you can think of this idea of size by saying two things are the same size if I can match every element in one set to every element in another set, just one by one. And so I really like, I'm borrowing from Kelsey Houston-Edwards’ PBS show, but what I really like that she said that you have two sets, and every element has a buddy, right? And so then I love that language, and so I'm borrowing from from her. But then that works for finite sets, but you can extend it to an infinite set.You can say, for example, that two infinite sets are the same size if I can find a matching between every element in the first set and every element in the second set. It’s very hard to picture in your head, I think, but we're going to try to do this. So for example, you can say that the natural numbers, the counting numbers, 1, 2, 3, 4, etc, have the same size as the even numbers, because you can make a matching where you say, “Match the number 1 with the number 2 on the other side. And then the number 2, with the number 4 on the other side.” And you have all the counting numbers, and for every counting number, you have two times that number as the even buddy.

    EL: Yeah. And I think this is, it's a simple example that you started with, but it even hints at the weirdness of infinity.

    AS: Yeah.

    EL: You’ve got this matching, but the even numbers are also a subset of the natural numbers. Ooh, things are going to get a little weird here.

    KK: Clearly, there aren’t as many even numbers, right?

    AS: Yeah.

    KK: This is where you fight with your students all the time.

    AS: That’s exactly—so when you're teaching this, the first thing you do is talk about things that have the same cardinality. And then everybody, it can take a while, you know, like, infinity is so weird that you can actually do these matching. So Hilbert’s infinite hotel is a really great way of doing this sort of more conceptually. So you have infinitely many rooms. And so for example, suppose that rooms numbers from 1, 2, 3, to infinity, mean, and so on. Yes, you have to be careful because infinity is not a number. You have to be careful with that. But say that all the rooms are occupied. And so then, you know, say someone shows up in the middle of the night, and they say, “I need a room.” And so what you do if you're the hotel manager is you tell everyone to move one room over. And so everyone moves one room over and you put this person in, and room number one. And so that's another way of seeing that. So the one-to-one pairing, or the matching here is every person has a room. And so the number of rooms and the number of people are the same—the word is cardinality because you don't want to say number because you can't count that.

    KK: Right.

    AS: And so you you say cardinality instead. But it's really weird, right? Because the first time you think about this, you say, “Well, you know, there's infinity, and there's infinity plus one.” That's like the kind of thing that you would say as a kid, right? And they're the same! When you have the natural numbers and the natural numbers and one extra thing, or like with zero, for example—unless you're in the camp that says zero is an actual number—but we're not going to get to that discussion right now.

    KK: I’m camp zero is a natural number.

    AS: Okay. I feel like I know maybe half people who say zero is a natural number and the other half say it's not. And I don't think anyone has good arguments other than, ah, it must be true! And so then the cool thing is, once you start doing that, then you start seeing, for example—and these are, these are kind of tricky examples, it can get tricky. Like, you can say that the integers like the positive whole numbers, negative whole numbers and zero, that also has the same cardinality as the natural numbers. Because you can just start with zero—I mean, basically, when you want to say that something has the same cardinality as the natural numbers, what you're really trying to do is to find a buddy, so you're trying to pair someone with one or two, or three. But really, what you can do is just list them in order, right? Like you can have like the first one, the second one, the third one, the fourth one, and you know that that's a good matching. It's like the hotel. You can put everyone in a room. And then you know they're the same number. Everyone has a room. So with the integers, for example, the whole numbers, positive, negative and zero, then you can say, “Okay, put zero first, then one, then negative one, and two, then negative two then three, then negative three,” and then they're the same size, right? And so once you start thinking about this—I remember this pretty clearly from from college—once you start thinking about this, then you're like, “Well, obviously, because infinity is infinity.” That’s the next step. So the first step is like, well no, infinity plus one and infinity are different. But then you get convinced that there is a way of matching things that where you can get things that seem pretty different, or a subset of a set, and they have the same cardinality. And then you go the other direction, which is “Well, of course, anything infinite is going to be the same size as anything else that's infinite.” And so then it turns out that even the rationals are the same size as the natural numbers. And that's way more complicated than we have time for. But if you add real numbers, meaning irrationals as well, then you have a whole different situation.

    KK: You do indeed.

    AS: It’s mind blowing, right? And so if you just think about the real numbers between zero and one, so just get go real simple. I mean, small, relatively. So you're just looking at decimal expansions. And so if those numbers had the same cardinality as the natural numbers, then you should be able to have a first one and a second one and a third one, and a fourth one. Or you can pair one number was the number 1, one number with the number 2, etc. And that list should be complete, and in the words of Kelsey Houston-Edwards, everyone should have a buddy. And so then, here's the cool thing, this is a proof that these two sizes of infinity are not the same, and it's a proof by contradiction, which is, again, your favorite proof when you are learning how to prove things. I mean, when I was learning proofs, I wanted to do everything by contradiction. So proving something by contradiction means you want to assume, “Well, what if we can list all the all the real numbers?” There’s a first one, a second one, a third one, etc. So Cantor’s amazing insight was that you can always find a number that was not on that list. Every time you make this list: a first one, a second one, a third one… there is some missing element.

    And so you line up all your decimals. So you have the first number in decimal. And so you have like, you know, 0.12345… or something like that. And then you have the next one. And the next one. And like, I mean, this is really hard to do verbally, but we're going to do it. And so you sort of line them up, and you have infinite decimals. So you have point, a whole bunch of decimals, point, a whole bunch of decimals. And so you can make a missing number by taking that first number in the first decimal place a just changing that number. Okay, so if it was a 1, you write down a 2. And so you know, because we’ve known how to compare decimals since we were little kids, that what you need to compare is decimal place by decimal place. So these are different because they're different in this one spot, right? And then you go to the second number, and the second decimal point. And then you say, “Well, whatever number I see there, I'm going to make the second decimal point of my new number different.” So if you had a 3, you change it to a 4, whatever it is, as long as it's not the original number. And and this is why it's called the diagonalization argument, or the diagonal argument, because you have lined all those numbers up, and you can go through the diagonal, and for each one of those decimal points, at each decimal place, you just change the value. And what you're going to get is a number, another real number, infinitely many decimals, and it's going to be different from every number on your list, just by virtue of how you made it. And so then, what that shows is that the answer to the “what if” is: you can’t. The “what if” is, if you have a list of all real numbers, it's not complete. So there is never going to be a way that you can make that list complete. And this is the part where every time I tell my students, at some point, they're like, “Wait, there are different sizes of infinity? What?” Then—and that’s sort of lovely, because it's just this this mind-blowing moment where you've convinced yourself, by the way, that you were to infinity is infinity, and then you realize that there's something bigger than the cardinality of the natural numbers. And and then it's really fun when you tell them, “Well, is there something in between?” They’re like, “Of course! There must be!” And then you're like, “Wait, no one knows.”

    KK: Maybe not.

    EL: Yeah.

    AS: So yeah, I just love that argument. And I love how simple it is. And at the same time, it's, simple, but it's very, very deep, right? You really have to understand how these numbers match up with each other. And it requires a big leap of imagination to just think of doing this and realizing that you could make a number that was not on this infinite list by just doing that simple trick.

    EL: Yeah.

    AS: And so I just think it's a really, really beautiful theorem. And then I also have a really personal connection to this theorem. But it's one of my favorite things to teach. And I'm going to be teaching at this term, and I’m really looking forward to seeing how that how that lands. Sometimes it lands really well. Sometimes people are like, “Eh, you’re just making stuff up.” Yeah.

    EL: Yeah.

    KK: Well, then you can really blow their minds then when you show them the Cantor set, right?

    AS: Yeah, yeah.

    KK: And say, “Well, look, I mean, here's this subset of the reels that has the same cardinality, but it's nothing.”

    AS: Exactly. Yeah, there's nothing there. Yeah.

    EL: Yeah. I remember, then, when I first saw this argument, really carefully talking myself through, “Like, okay, but what if I just added that number I just made to the end of the list? Why wouldn't that work?” And trying to go through, like, “Why can't I—Oh, and then there must be other numbers that don’t fit on the list either.” It's not like we got within 1 of being the right cardinality.

    AS: Right.

    EL: For these infinite number. So yeah, it's a really cool idea. But you said you had some personal connections to this. So do you want to talk more about those?

    AS Sure. So I am from Venezuela, and I went to college there. And I liked college, it was fine. I knew—Well, one thing that you do have to decide when you're a student in high school is, you don't really apply to college, you apply to a major within the college. And so then I knew I wanted to do math. And I signed up for math at a specific university. And so then the first year was very similar to what you would do in the States, which is sort of this general year where everybody's thinking calculus, or everybody's taking—you have some subset of things that everybody takes. And then your second year, you start really going into the math major. And so this was my first real analysis class. This was my first serious proof-y class in my university. And we learned Cantor’s diagonalization argument, which was pretty early. But I loved this argument. I felt so mind-blown. You know, I was like, “This is why I want to do math,” you know, I was just so excited. And I knew I understood everything. And so I took the exam, and I got horrible grade. And in particular, I got zero points on the “prove that the real and the natural numbers don't have the same cardinality.” And so I went to the professor, and I saw my exam, and I was really confused. And I went to the professor, and I said, “I really don't understand what's wrong with this problem. Could you help me understand?” Because I thought I understood this. And then—you know, that's a typical thing. I probably said it in a more obnoxious way than I remember now. But I felt like I was being pretty reasonable. I was not the kind of kid that would go up to my professors too often to ask for points. I really was like, “I don't know what I did wrong.” And especially because I felt like I really got it.

    EL: Right.

    AS: And so then he just looked at me and said, “If you don't understand what's wrong with this problem, you should not be a math major.” And that was it. That was the end of that conversation. Well, I still don't know what's wrong with this problem, and now you just told me I need to do something else. Just go do something else at a different school. Right? And I mean, I don't know that that was particularly sexist. But I do know that I was the only woman in that class, and I know that I felt it a lot. I think he probably would have said that—I really do think that he in particular would have said that to any student. I don’t think it was just me being female that affected that at all. But I do think that if I had been less stubborn about my math identity, I might have taken him up on that. But I was just like, “No, I'm going to show you!” And eventually I got an A in his class. He taught real analysis every semester, so I had to take the class with him every time and at some point, I cracked his code. And he at some point respected me, and thought I deserved to be there. But he was just very old-fashioned. You know, I don't think it's even sexism. It's just very, very, like, this is how we do things. And then I went—eventually, I did talk to someone. I think it was a teaching assistant. And I was like, “I don't know what's wrong with this problem.” And he looked at it. And he said, “Well, here's the problem. When you were listing—so you needed to list all these generic numbers and their decimal expansion. And I did, “Okay, the first number is point A1, A2, A3, etc. The second number is point B1, B2, B3, etc. The third one is point C1, C2, C3, etc, dot dot dot, right? And he said, “You have listed 26 numbers. And that's not going to be an infinite list.” Right?

    KK: That’s cheap.

    AS: And I was just like, “Okay, but I got the idea, right?” I was like, “Okay, it's true.” He’s like, “The way you wrote it is incorrect.” And I'm like, sure.

    EL: Sort of.

    KK: I’ve written that same thing on a chalkboard.

    AS: You know, this shows you—like, fine, you can be more careful, you can be more precise, but from this, you shouldn't be a math major? That’s pretty intense.

    EL: Yeah.

    AS: And I knew the mechanics, I knew what was supposed to be happening, I knew how to make the missing number, right? Like you just need A1: you change it to some other number, B2: you change it to some other number, C3: you change it to some other number. And so, I just thought—I mean, that was a moment where I was just literally told I should not be in math because I made a silly mistake. And it was a moment where I realized that—now looking back, I realize my math identity was pretty strong, because I just said, “Well, ask someone else to see what was wrong, and I'm not going to ask this guy anymore because it's clear what he thinks.”

    EL: Yeah.

    AS: And sort of the stubbornness of, “Well, I’ll show him that I do deserve to be here.” But I think of all the students who might have taken classes with him, who would have heard that and then been like, “Yeah, maybe I need to do something else.” I mean, it just makes me really sad to hear, especially now that I'm a professor, and teaching these kinds of things. It just makes me sad to see which people were just scared away by someone like that, you know?

    EL: Yeah.

    AS: So that was a big moment for me. Yeah.

    EL: Yeah. Quite a disproportionate response to, what’s basically a bookkeeping difficulty.

    AS: Yeah.

    EL: So, you know, we like to get our mathematicians to pair their theorems, with something on this show. And what have you chosen as your pairing for Cantor's diagonalization argument?

    AS: Well, now that you suggested, music and other things, I'm maybe changing my mind.

    EL: You could pair more than one thing.

    AS: I was trying to find something that was just like—I need to sort of express the sort of mind-blowing nature of this, right? And so I was like, a tequila shot! You know, really just strong. And like, “Whoa, what just happened?” And so that was one thing that I thought about. And then—I don't know, just mind-blowing experiences, like, when I saw the Himalayas from an airplane, or when—you know, there are some moments where you're just like, “I can't believe this exists.” I can't believe this is a thing that I get to experience. So I guess, you know, there's been—most of these have been with traveling, where you just see something that you're just like, “I can't believe that I get to experience this.” And so I think Cantor's diagonalization argument is something like that, like seeing this amazing landscape where you're just like, “How does this even exist?”

    EL: Yeah, I like that. I mean, I've had that experience looking out of airplane windows too. One time we were just flying by the coast of Greenland. And these fjords there. Of course, an airplane window is tiny and it's not exactly high-definition picture quality out of the thick plastic there, but it just took my breath away.

    AS: Yeah.

    EL: Yeah, I like that. And we can even invite our listeners to think of their own mind-blowing favorite experiences that they've they've had. Hopefully legal experiences in their jurisdiction.

    KK: Well, oh wait, it's not 4/20 anymore. Oh, well. So we also like to invite our guests to plug anything they want to plug. So you write for the AMS, the inclusion/exclusion blog, are there other places where we might find your mathematical writing for the general public?

    AS: Well, that's my main plug and outlet right now. But I I do write for the MAA Focus magazine sometimes. That's sort of my main, and sometimes the AWM newsletter. So you might find some of my writing there. And the blog. I mean, again, now that I'm chair and doing a lot of other things, I'm not writing as much, but I definitely like to—I’ve gotten really into maybe this is a weird plug, but I've gotten really into storytelling.

    EL: Oh yeah, you’ve been on Story Collider?

    AS: Yeah, I was on one Story Collider. I've done some of the local stuff. But you can find me on the internet telling stories about being a mathematician. Some of them about some pretty fantastic experiences, and some not so great experiences.

    EL: Yeah. Okay. Yeah. Well, we'll link to your Twitter, and that can help people find you too.

    AS: Oh, yeah. Cool.

    EL: Thanks a lot for joining us.

    AS: Yeah. Thanks for having me and listen to me ramble about infinity.

    EL: Oh, I just love this theorem so much.

    KK: Yeah, we could talk about infinity all day. Thanks, Adriana.

    AS: Yeah. Thank you so much.

    We were excited to have Bates College mathematician Adriana Salerno on the show. She is also the chair of the department at Bates and a former Mass Media Fellow (just like Evelyn). Here are some links you might enjoy along with this episode.


    Salerno's website

    Salerno on Twitter
    AAAS Mass Media Fellowship for graduate students in math and science who are interested in writing about math and science for non-experts
    Hilbert’s Infinite Hotel
    Evelyn’s blog post about the Cantor set
    Salerno’s StoryCollider episode
    The inclusion/exclusion blog, an AMS blog about diversity, inclusion, race, gender, biases, and all that fun stuff

    31 min

About My Favorite Theorem

From the publisher's feed

Join us as we spend each episode talking with a mathematical professional about their favorite result. And since the best things in life come in pairs, find out what our guest thinks pairs best with…

More shows like My Favorite Theorem

Radiolab by WNYC Studios

Radiolab

43,766 Listeners

Freakonomics Radio by Freakonomics Radio + Stitcher

Freakonomics Radio

31,959 Listeners

Hidden Brain by Hidden Brain, Shankar Vedantam

Hidden Brain

44,076 Listeners

Wait Wait... Don't Tell Me! by NPR

Wait Wait... Don't Tell Me!

39,164 Listeners

More or Less by BBC Radio 4

More or Less

878 Listeners

Pivot by New York Magazine

Pivot

9,783 Listeners

Universe Today Podcast by Fraser Cain

Universe Today Podcast

597 Listeners

The Quanta Podcast by Quanta Magazine

The Quanta Podcast

543 Listeners

The Daily by The New York Times

The Daily

112,262 Listeners

Sean Carroll's Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas by Sean Carroll

Sean Carroll's Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas

4,288 Listeners

Practical AI by Daniel Whitenack and Chris Benson

Practical AI

204 Listeners

The Numberphile Podcast by Brady Haran

The Numberphile Podcast

455 Listeners

The Ancients by History Hit

The Ancients

3,332 Listeners

The Ezra Klein Show by New York Times Opinion

The Ezra Klein Show

16,290 Listeners

The Joy of Why by Steven Strogatz, Janna Levin and Quanta Magazine

The Joy of Why

515 Listeners