Under my data

A regrettable lecture, and learning to see

“How many times can a man turn his head and pretend that he just doesn’t see?” - Bob Dylan

A spiral and an elliptical galaxy from Galaxy Zoo (as displayed on Astrobites)

Galaxies are pretty big, complicated things. For much of my life I’ve been trying to fit these million trillion mile objects into my little 6 inch brain, trying to do what I can to look at them, and understand.

Usually “understanding,” in this context, involves trying to turn each galaxy into a number, or several numbers: Andromeda is 46 kiloparsecs across, Triangulum has a brightness of 5.7 magnitudes. I, along with other astronomers, use these numbers to make comparisons: Triangulum, for example, is only 18 kiloparsecs across, less than half Andromeda’s size.

With enough numbers we can make a graph, each galaxy just a single dot. We look for patterns: these galaxies group together, those galaxies follow a trend. We construct systems of classification: elliptical galaxies look like blobs, numerically have something called a “Sersic n parameter” near 4. Spiral galaxies look like whirlpools, numerically have n near 1.

We use the patterns to identify outliers – galaxies that just don’t fit in – and often look at the outliers to see what is “wrong.” We quickly settle into ideas of normal and weird – rules for how the universe is. An accomplishment, I suppose. We dust off our hands like we’ve learned something, rest in our neat wrangling of chaos into order. We know so much. Case closed.

—

Early in graduate school I was part of a team of teaching assistants for the 300+ student Astronomy 1101 course. Most of my duties involved grading, writing homework problems, and leading discussion sections. But I vividly remember, early in my second year, when I got asked to fill in for the professor and teach all 300 students a class on galaxies.

Nervous, excited, I wanted to do a good job, so I made slides that really broke it down. Galaxies aren’t that hard, I told them. There are basically two types. Ellipticals, which appear reddish, and spirals, which appear blueish. I had easy “clicker” questions, where the students could submit their answers live. Do ellipticals live in clusters - galaxy cities, or isolation - the galactic countryside? Cities - yes! Over 90% of the class would get it right. See, galaxies are easy: red and blue, black and white: a spoonful of sugar with very little medicine to take down.

How I regret that lecture now.

I mean, sure, pedagogically it left a bit to be desired. Good clicker questions, I’d learn later, should be a challenge: learning happens when misconceptions are challenged, in the balance between oversimplified, and overwhelmed. But that’s not why I regret it.

When you get to work trying to classify a galaxy you soon realize it’s not as easy as it appears. Take the simple measurement of the “brightness” of a galaxy: the Sloan Digital Sky Survey - one of the astronomical standards for the better part of the last two decades - contains over 50 different measurements of brightness for each galaxy. Do we mean the peak brightness? The brightness within some radius? At what wavelengths, or colors? Do you include that blob that might be in front or behind? Et cetera, et cetera.

In fact, most of my thesis work, six years of my life, looked at a class of galaxies called ultra diffuse galaxies, where standard measurements didn’t cut it. The light in ultra diffuse galaxies (UDGs for short) is far more spread out than typical galaxies, so while UDGs have lots of light in total – just as much as typical galaxies – they are difficult to detect because there is only a little light in each pixel. So they had previously escaped detection: to see the brightness of these galaxies required more effort, or at least, a different way of looking.

And finding these galaxies has changed our understanding: they spin differently, they form stars differently, they have different amounts of dark matter. Thanks to UDGs, we now see the universe in a whole new light.

Which, I think, gets at the difficulty with my rookie teacher self. The world of galaxies doesn’t require nuance, my lecture implied. With just a couple of categories you know enough of the ways of the cosmos. But the better I’ve gotten to know galaxies, the more my stark categories break down.

—

At one point in my study of UDGs I came up on an uncomfortable realization. My graphs and computations could only get me so far: I needed to visually examine about 3000 potential UDGs to determine their validity. A painfully repetitive task, I stalled, I thought whether there was any way to avoid it. But it needed to be done.

I still remember the off white shade of the desk I was using at the time, as for two long days I sat there and actually looked at the galaxies behind my data. I remember the seemingly interminable grind, like when you have a bad cold, and though you know this too will pass, and relatively quickly, in the moment it seems like eternity itself has settled on your sinuses and the discomfort will never subside.

But there’s something else I remember. The more I looked at these galaxies, the more something in me seemed to open. I felt my pulse quicken, synapses began to fire. Previously comatose places of my brain snapping to attention as galaxy after galaxy would surprise me. What are those blue blobs? Why is this one so off center? What is that weird fuzzy stuff over there? It’s a moment in my research career when I remember feeling most alive, new ideas flooding in, a wonder at a whole beautiful complicated universe of off center warps and blue blobs and weird fuzz.

I wish now I’d told those students that – more than just two categories – galaxies come in all shapes, forms and sizes. That Galaxy Zoo, a huge project with thousands of volunteers trying to better understand galaxy classification, found so many different morphologies they were able to find galaxies the shape of every letter of the english alphabet, capital and lower case!

I wish I told them how astronomers too have a hard time with galaxies that don’t fit their models. How intimidating and difficult it can be to actually look, again and again.

But mostly I wish I’d challenged them, as I’ve been challenged again and again, to consider how while we need models and stereotypes to understand the world, these are not the world. That to see the galaxy under each data point is more complicated, but also so worthwhile.

—

There’s a bit of a cliche fact that we like to tell students in Astronomy 1101. We like to point out that the elements in their bodies were formed in the guts of stars: that each of us there in that lecture hall – board or sweating or preoccupied with what their crush said last night – is composed of galaxy stuff. The same carbon and hydrogen that pulses through galactic veins composes you and me.

Something else I wonder: humans too are complicated things, collections of 30 trillion cells that also, at the end of the day, don’t really fit in my 6-inch brain. Right now I feel like I live in a world where so much effort is expended in making me see numbers, not people – 300 deported, 50,000 killed, 2 million incarcerated. I don’t know if I can see past it. But if galaxies are any indication, it’s worth the effort, lest I miss so much light I’d otherwise fail to see.