I spent the past week on retreat with fellow faculty members. During part of this time, we each shared our “vocational journeys,” or the stories of how we have been led to this job and these fields of scholarship. I thought that part of my essay might have broader appeal, so I’m posting it here.

__Why do I study and teach mathematics?__ My research is in the field of *pure* mathematics, for which it may seem harder to justify an investment of time than for its adjacent field of *applied* mathematics. Applied math at least tries to tie itself directly to the needs and concerns of our immediate physical world. Pure math is happy to oblige in improving how well we understand the world, but its primary concern is math for math’s sake. (The boundary between these two types of math is highly permeable, and even pure math almost always starts with inspiration from experience.) I’d like to address the question by comparing math with two other areas represented in academia: music and science.

First, math is __like music__. The aesthetic element in mathematics is essential, not peripheral. I’m not sure, but I think that in the minds of many people mathematics is reduced to a collection of more-or-less arbitrary facts, like the fact that the area of a circle equals pi times the square of its radius. Each of these facts, however, is like the final cadence of a symphony. It may be thrilling by itself, but it’s missing the indispensable context of “where did we start?” and “how did we get here?”

This is why mathematicians insist on proving things: the proof is a whole symphony, not a single chord. Mathematicians are lauded not for stating facts, but for demonstrating their necessity, the way composers and musicians are praised for the whole course of a piece or a performance, not just its ending. When executed well, a proof has rhythm. It has themes that are developed and interwoven. It has counterpoint. It sets up expectations that are satisfied or subverted. Economy of material is valued, but not exclusively; an argument that wanders into neighboring territory, like a modulation to a neighboring key, can provide fuller appreciation of the main theme.

Proofs have a variety of forms, some as common as sonatas and minuets: direct proof, proof by contradiction, proof by induction, proof by picture, proof by exhaustion. We have computer-generated musical compositions and computer-generated mathematical proofs, and in both communities there is healthy debate about whether these artificial creations are beautiful or desirable in such quintessentially human activities. We return over and over to the same pieces and theorems that have inspired us, whether they be simple or grand, and each performer gives her or his own interpretation and inflection to the presentation.

Second, math is __like science__. Often mathematics is categorized as a science, and that’s not entirely wrong. Science is built on careful observation, winnowing data from the chaff of noise. Science seeks explanation which can be turned into prediction. It invents new tools for collecting information and improves upon those that already exist. It creates models and theories that encompass and relate as many pieces of knowledge as possible.

Where science and math differ is that science deals with the world in which we live, while the world of math is imagined. *Imagine* that there are such things as points with no volume and perfectly straight lines that connect them. *Imagine* that numbers have enough solidity that we can move them around en masse by means of undetermined variables, the *x*, the *y*, the *z*. *Imagine* that once we start counting upwards *1, 2, 3, 4, 5, a thousand, a million, a trillion, a googol,…*, we could never reach an end, not in any number of lifetimes in any number of universes. Or *imagine* that the filigree of a fractal truly exists at every scale, that we can examine it closer and closer and see the ever-increasing detail, that there is no quantum barrier to our exploration, beyond which sight and measurement cease to be meaningful.

When we imagine these things, we create the worlds in which we make our observations. The rules of these worlds are not completely arbitrary, at least not if we want to be able to know anything about them, but they are ours to choose. Each time we choose anew, we enter an undiscovered country. Once in this country, we must return to scientific methods of study. We look for patterns, try to explain them, and check that our explanations make accurate predictions. We must know when to trust the instruments we have—our minds, computer programs, results proved by other mathematicians—and when not to trust them. Like scientists, we have to winnow out the noise.

Mathematical truth persists across ages and cultures, and so it may seem timeless, but our experience of it certainly isn’t. The channels of logic through which a proof flows may be carved out once and for all in eternity or in the human mind (depending on your view of where mathematical truth lies), but like notes on a page they remain inert until they are brought to life by individual or communal study. Like the tree of life in biology or the standard model in physics, mathematical theories are crystallized around *our* experience and *our* perception of the world. As Bill Thurston wrote, “mathematics only exists in a living community of mathematicians that spreads understanding and breathes life into ideas both old and new. The real satisfaction from mathematics is in learning from others and sharing with others.”