How was massive astronomy done in the 1800’s? Who were the Harvard Computers? How did Annie Jump Cannon figure out how to classify stars? I discuss these questions and more in today’s Ask a Spaceman!

This episode is sponsored by BetterHelp. Give online therapy a try at betterhelp.com/spaceman and get on your way to being your best self. Visit BetterHelp to get 10% off your first month!

Support the show: http://www.patreon.com/pmsutter
All episodes: http://www.AskASpaceman.com

Follow on Twitter: http://www.twitter.com/PaulMattSutter
Like on Facebook: http://www.facebook.com/PaulMattSutter
Watch on YouTube: http://www.youtube.com/PaulMSutter

Read a book: http://www.pmsutter/book

Keep those questions about space, science, astronomy, astrophysics, physics, and cosmology coming to #AskASpaceman for COMPLETE KNOWLEDGE OF TIME AND SPACE!

Big thanks to my top Patreon supporters this month: Justin G, Chris L, Barbara K, Duncan M, Corey D, Justin Z, Naila, Scott M, Rob H, Justin, Louis M, John W, Alexis, Gilbert M, Joshua, John S, Thomas D, Simon G, Erin J, Jessica K, Valerie H, David B, Frank T, Tim R, Tom Van S, Mark R, Alan B, Craig B, Richard K, Dave L, Stephen M, Maureen R, Stace J, Neil P, COTFM, Stephen S, Ken L, Alberto M, Matt C, Joe R, David P, Ulfert B, Sean M, Edward K, Tracy F, Sarah K, Steven S, Ryan L, Ella F, Richard S, Sam R, Thomas K, James C, Jorg D, R Larche, Syamkumar M, John S, Fred S, Homer V, Mark D, Colin B, Bruce A, Steven M, Brent B, Bill E, Tim Z, Thomas W, Linda C, David W, Aissa F, Marc H, Avery P, Scott M, Thomas H, Farshad A, Matthias S, Kenneth D, Maureen R, Michael W, Scott W, David W, Neuterdude, Cha0sKami, Robert C, Robert B, Gary K, Stephen J, dhr18, Anna V, Matthew G, Paul & Giulia S, Ron D, Steven M, Louis M, Michael C, Alyssa K, Lode D, Roger, Bob C, Patti H, Red B, Benjamin M, BlueDragon, Stephen A, Ian S, James R, Skip M, Robert O, Adam I, Lynn D, Jeffrey C, Allen E, Paul G, Michael S, Jordan, Colin H, Jessica M, Thomas H, Reinaldo A, and Amy Z!

Thanks to Cathy Rinella for editing.

Hosted by Paul M. Sutter.

 

EPISODE TRANSCRIPT (AUTO-GENERATED)

There was a time, not as long ago as you might think, that the word computer wasn't for an electronic device, but it was for a job description. Physics, and really all of science, but physics in particular, is a mathematical exploration of nature. We use mathematics to describe, to understand, to predict the many wonderful patterns we see in the universe, and how we do that deserves an entirely separate episode. Feel free to ask. And mathematics often requires a serious amount of number crunching.

Today, that number crunching happens on our computers and our calculators. Everything from computing a simple formula to solving complex equations, it all happens on the machine. But we scientists in the 21st century aren't the only ones who have needed to compute simple formulas and solve complex equations. So what are you supposed to do? If you're a pre 20th century scientist that needs to get some math done, but you don't have a magical little box that does it quickly, efficiently, and correctly for you, what do you do?

Well, you hire someone to do it. You go out on the job market, you say, hey, I need someone to do all my math for me so that I can do the the higher level thinking or the data collection or the analysis and the infras, but I need someone to crunch the numbers. I need someone to go through all the data, find the commonalities. I need someone to apply this equation. I need someone to calculate this or that.

A famous example of this computer was Johannes Kepler. He was the computer. He was hired as the computer for Tycho Brahe. Tycho Brahe was operating as perhaps the greatest astronomer to ever live, and he was collecting a lot of data. He was trying to make lots of predictions.

He was trying to get a lot of math done and he hired Kepler to do it. And then Tycho died and then Kepler took all the data. I'm not saying there's a conspiracy there, but I'll let you draw your own conclusions. After the development of what we recognize now as mathematical science and mathematical astronomy with the Scientific Revolution, the problems only got bigger because we became more sophisticated. It wasn't just naked eye astronomy, it was a telescope based astronomy.

We we started to gather more data and so there's more and more need for computation and so there's more and more need for computers. You couldn't just have 1 or 2 assistants running around doing the calculations for you, you needed teams of people. There's another famous example, George Irie, who was the head of the Greenwich Observatory in the 1800s, had a room full of kids doing math all the time in something called the octagon room, which strikes me as a great name for a sequel to the squid games or something, but no, it was a room full of kids doing math all the time. And it was this wonderfully elaborate system where each kid was just assigned a certain task, and they just had to repeat that task over and over and over again. And the kids would pass notes or messages to each other, so like, the kid in front of you would solve one part of the problem, he would pass it back to you, and then you would solve the next part, and then pass it to the kid on your right, and then if the kid on your left gave you a note, then you applied a different algorithm or solve that, and then you passed it behind, and so is this like it looked like a computer algorithm, but it was performed by human beings, mostly kids, just doing rote mathematics over and over and over and over again.

Just the grunt work of doing astronomy and trying to find mathematical descriptions of nature that involves a lot of mathematics, involves a lot of number descriptions of nature that involves a lot of mathematics, involves a lot of number crunching, here were the computers to do it. By the late 1800s, this job of human computer, or back then they just called it computer, was taken up almost entirely by women. Why? Well, there are a few potential reasons for it. 1, kids can be very, kid like.

If you've ever noticed kids, they're not so great at sitting in a desk for hours at a time, performing the same task repeatedly, especially a boring task repeatedly. So, you know, a boring task repeatedly. So, you know, that starts to get a little old pretty quickly. And, as science grew, as our data collection efforts grew, the calculations and theories we were doing were becoming more complex and it was beyond the scope of most kids' mathematical abilities. It wasn't just multiplying numbers or dividing numbers.

It wasn't just doing simple things. It was doing very complex nuanced things, and I'll give you some examples of that soon, that were simply beyond the scope of most kids. Hey, little Johnny, I need you to run-in, best fit algorithm and you've already lost them. 2, there was a large social movement happening in the time in the US and Europe to allow women access to higher education or or education at all. This was highly controversial.

Some people were dead set against it, some people were dead set for it, but no matter what, there were becoming institutions, college level institutions aimed at women. At the time, almost all educational institutions were for men only, and so there were these parallel institutions that admitted women. So you had women going into college, getting an advanced education, being interested in science because they're human, so of course some of them are interested in science, and they ended up obtaining scientific degrees and expertise with all the same qualifications as their male peers, but you can't just let women do science. That would be crazy. But there they were, standing around being smart and being educated and asking for a job, so you had to give them something to do, voila, let's give them the job of computer.

We'll let the men do the science, we'll let the women do the grunt work. In the world of astronomy, the most famous of these computers were the Harvard Computers, a team of well educated women hired at the Harvard Observatory conveniently enough, starting in 18/77 by its director, Edward Pickering. These women were not allowed to operate the observatory's telescopes because that was a man's job, of course, but those telescopes, which were now scattered around the globe on various observing missions, were producing an enormous amount of data. What's leading into the late 1800 here is the development of photography and the photographic plate, and this was as much of a revolution to astronomy as Galileo's invention of the astronomical telescope in the first place. For two reasons.

1, a photograph can see much farther in much dimmer objects than the unaided human eye can because a photograph can add, you can sit a plate at the bottom of a telescope and it can just keep adding the light over the course of an entire night onto that single plate. You can accumulate an image, you can't do that, it doesn't matter how much you stare at the night sky with your eyeball, you're only gonna see the same set of stars. I mean, after a while you'll get night adjusted, all that, but after that, there's, you can't just like keep staring and eventually accumulate enough light, that's not how our eyes work, but that is how photographs work. So we were able to collect a lot more stars, and other various random objects, because we could see a lot further. We could see the small ones, we could see the dim ones, we could see the distant ones.

And, we were able with photography to preserve them. So, it wasn't any longer a matter of an astronomer sitting next to the telescope looking through the eyepiece and then sketching it, then that's the only physical record of what that astronomer saw that night. Instead, we have a photograph and you can just go out and observe for like 3 months in some primo location with nice beautiful dark skies, and then collect crate after crate after crate of photographic plate, and then send it back home. The amount of data that astronomers were able to begin collecting in the late 1800 was simply too much. It was too much for a researcher or a small team to handle.

You needed computers. You needed human beings to sit there and process the data, look at every photographic plate, tag the stars, identify them, categorize them, put them in a little label, as was the fashion in the late 1800s, the goal of these massive observing programs was a, to find all the stars above some brightness threshold and b, fit every single star into a nice neat categorization system because obviously that's how nature works and this shouldn't be hard at all. It's like taxonomy, but for stuff in space, and we need computers for that. We need people to look at these stars and say, okay, category A, category B, category C. Next.

Boom. Done. Boom. Done. Boom.

Done. Just do it over and over. The exact kind of thing that we would now develop a computer algorithm to do. And it shouldn't be so hard. You get a bunch of stars, you can categorize things, nature is nice and easy, God designed an orderly universe, etcetera, etcetera, so we this should just pop out any day now.

Surprise, this isn't how nature works at all. And classifying and categorizing stars is incredibly complicated. So the solution here was more data. Okay. So if we have 10,000 stars, and nothing makes sense.

If we have, let's go out and get a 100,000 stars, and then with more stars we can come up with the categorization system that is just gonna make sense. Why are they categorization system that is just gonna make sense. Why are they so invested 1800s in this categorization scheme for different kinds of stars in the sky? It's because they wanted to know how stars work. They wanted to understand.

You look out with your eye on the night sky and you see different kinds of stars. Some are brighter, some are dimmer, some are bluer, some are redder, some are whiter. What's going on? And as we were collecting 1,000, tens of thousands, 100 of thousands of stars, we were seeing all kinds of stars. Some of them stay fixed, some of them varied, some of them varied in very strange ways.

Some were dim and red, some were bright and red, some were medium and whitish, some were super bright and bluish like, what is going on? How are they born? What are the different kinds? Why are they the different kinds? How do stars live?

How do they die? What are all the kinds of properties of stars and how are they connected together? How do they evolve? The hope is that you can take this massive data set and use it to understand something about nature. If you just like took pictures of people and collected more and more and more pictures that eventually you would understand how humans develop.

That was the hope. One of the biggest challenges they were facing besides, you know, the enormous amount of data that these programs were collecting is that they actually had very little information about stars to go on, especially early on. They have, like okay. You're gonna take a picture of a star. What do you got?

You have their position on the sky. You have their color? Okay. You have their distance. You have their distance if they're close enough.

You can get a parallax distance to them. Okay. And, oh, that's right. We have their spectrum. You know, we have the breakdown of the different wavelengths of light being emitted by that star, so a fancier version of their color.

And that's it. That's all we got. You're doing these surveys of 100 of 1000 of stars and you're collecting a very tiny amount of information about each star. And from there, you're trying to unlock how stars work. It's like if Charles Darwin only knew the color and the size of the animals he studied, and he is using that to try to write origin of the species.

Well, points awarded to astronomers for at least being ambitious in the late 1800s. Obviously, all this work needed computers in a big way, and it's exactly this kind of task that modern electronic computers excel at, churning through vast quantities of data, looking at certain key features, computing straightforward arithmetic, and spitting out a result. Okay. This is the program. We're gonna classify the stars and we're gonna use that to understand how stars work.

Question from the back. The goal is to classify stars and somehow use that classification to understand stars, but, but, how do you classify stars? We have the data, we have gobs of it, not a lot about Estar, we know where it is and its color, its spectrum. But how do you classify stars if you don't know how stars work? We have the data to use to classify them but we don't have the scheme to classify them.

How do- like, this is insanity. It'd be one thing if we knew how stars worked and from there developed a classification scheme, but in classic astronomical fashion. Astronomers are going to study things and assign categories to things before we understand that, and hopefully by categorizing them, we can get some sense of how they work. It's a big leap, but we're gonna give it a shot. And it turns out someone did it, and the person to do it was one of the Harvard Computers, Annie Jump Cannon.

She was born in 18/63 to a well-to-do family in Dover, Delaware. Trust me, it's not poor people signing up to be Harvard Computers. She was encouraged by her mother to study astronomy, and science, and all that good stuff, which as you might imagine was rare for that time. She attended Wellesley College in Massachusetts, which was one of the few higher education options for women at the time. She graduated valedictorian, obviously very smart.

She also casually went to Europe to develop, pun intended, her photography skills. After the usual sort of unsteady job hunting that plagues academics to this day, in 18/96, she landed a job as one of Pickering's Harvard Computers. Her job was straightforward. Given a photograph of a star in its spectrum, catalog its properties, and assign a classification to it. That's it.

And, just do that, like, a lot, and accurately, and quickly. Now, at the time, she wasn't working in a vacuum. Astronomers had already been trying to develop a classification scheme for some time. And, to be fair, almost every single astronomer had their own classification scheme going because nobody had any idea what was going on. But there was one that was gaining popularity, it was developed in the 18 sixties seventies by an Italian Catholic priest and astronomer by the name of Angelo Secchi.

Now remember, astronomers don't have a lot to go on to classify stars, so Secchi settled on how strong the hydrogen spectral line appeared. So, like, you you're gonna see a star. There's some hydrogen there. It might be glowing and emitting radiation at very specific wavelengths. It might be blocking very specific wavelengths.

It might be absorbing light. But whatever that line is gonna be there, and it might be weak, it might be strong, it might be really thin, it might be really broad, depending on the situation. Again, we have no idea how or why the hydrogen is doing what it's doing, but you can look at different kinds of stars, and they have different appearances of hydrogen features. And so that's what Seki did. Class 1 was a white and blue star with broad and heavy hydrogen lines.

Class 2 was yellow stars with, less strong hydrogen lines, but still still kind of there. And then class 3 was orange and red stars with kind of a lot going on in the spectra, not just hydrogen but, you know, very complicated. This class, this classification scheme implies some sort of ordering, some sort of perhaps hopefully, evolutionary connection between them, like maybe stars start out white or blue, and then evolve to yellow, and then change to orange and red, or maybe the other way around. That is the hope, that if you classify these stars in the right way, you will discover some links between them. We knew stars had to be formed somehow, and we knew that stars have to live somehow, and that as they live, they probably evolve, and then they eventually die.

And so if we can define the right ordering, we might be able to make some sense of this. Like if you if you looked at people and measured height, and you knew nothing about how people worked, and you divided them into 3 classes, there's gonna be class 1, which are human beings less than half a meter tall, and then class 2 is gonna be human beings between half a meter and 1 and a half meters tall, and then class 3 is all the humans taller than 1 and a half meters. Okay. It's a start, like it's something you can measure and it's a way you can divide this population up and then you hope this tells you something about how humans work. Oh, maybe humans start out small, and then as they live, they get bigger.

Of course there are gonna be outliers, of course it's not always gonna connect, but like, that's the hope that if you find the right classification scheme, this tells you something important. Now if you divided humans by hair color, class 1 is all the brunettes, class 2 is all the blondes, class 3 redheads, and then class B is all the bald people. You hope that you generated some classification scheme that tells you something interesting about humans, it turns out it doesn't really tell you much interesting about humans, except that they have different colors of hair. But this is what astronomers are trying to do. They're trying to find a classification scheme that tells them something interesting.

In 18/90, Pickering and his Harvard Computers published a catalog of 10,000 stars, and with that number of stars, they found ones that didn't fit this simple classification scheme. White stars with few hydrogen lines, yellow stars with strong lines, orange stars with not a lot going on, etcetera, etcetera. There's just stuff that didn't fit. As the data increased in abundance, we had to develop a more sophisticated and nuanced classification scheme to account for them all. So the Harvard folks subdivided Secchi's original groups.

Instead of class 1, class 2, class 3, we're gonna break these down into subgroups and instead of calling them 123, we're gonna use an alphabetical scheme. So now we have a stars, m stars and so on. Again, hopefully implying some sort of evolution or ordering between them going from a at one end to o on the other end. We we're not exactly sure what those ends are. Are these ends of temperature, of life stages, of size, who knows, but it's something.

But as the years went on, the number of stars in the Harvard catalog continued to grow, and that system was itself starting to break down. There were some stars that didn't fit in any category, and worse, the categories didn't seem to line up next to each other. With more stars, more data, the classification scheme just wasn't working. There were too many exceptions, too many outliers and nothing made sense. The person responsible for generating the vast majority of the classifications was none other than Annie Jump Cannon herself.

She was a machine, or or should I say she was a computer. By 19 13, she was skilled enough to accurately classify up to 200 stars an hour. Based on looking at the faintest spectral features through a magnifying glass, in total in her life, she manually examined and classified 350,000 stars. That is insane. She was so good at this.

And the amazing thing about Annie Jump Cannon was that she wasn't just a computer. She wasn't just sitting there dumbly processing and categorizing and moving on. She was not a machine. She was not an electronic device. She was a human being.

She was a smart, sharp, and quick human being. And because she was the closest to the data, as in her nose was literally shoved in spectra all day long, she was the first person to develop an intuition about the classification. Now this is something I don't find in any of the the biographies or histories of Annie John Cannon. This is what I think was going through her head, was the development of a scientific 6th sense. She knew that we were trying, we being astronomers at the time, were trying to classify the stars, and that if we found the right classification scheme, it might be able to tell us something interesting and powerful and useful about stars themselves.

But so far, nobody had been able to find it. That classification scheme that worked and made sense and allowed us to easily identify and classify stars and then would lead to something useful and powerful. But she was in the spectra all day long, every single day. She began to develop, I believe, an intuition about what the right classification would be. As more and more stars came in, it was becoming apparent to canon that whatever evolution that the categories were implying wasn't working.

For example, she got the intuitive sense that o stars didn't belong at the end, but at the beginning, and that a and b stars had to be swapped. I can't really describe to you how she came up with this. As far as I'm aware, there's we have no recording of her internal thought process that led to this. She didn't tell us how she came to that conclusion. She didn't tell us like why this seemed to make more sense, where you had Secchi's original classification of 3 different kinds of stars, then you got some more stars, okay, so we need to subdivide it, but then this subdivision, okay, some things need to be switched around instead of a b c, we need to go o b a.

Why? Because it made sense to Annie Jump Cannon, that's why. That was it. It just made sense. Within a few years, she developed her own classification scheme.

O stars were blue stars with strongly ionized helium. B stars were blue white stars with strong neutral helium. A stars were white stars with helium no longer visible, but h one visible, and some neutral metal lines beginning to appear. F stars were yellow white stars. Hydrogen's getting weaker.

Metal lines are getting stronger. Calcium getting the strongest. G stars are yellow stars like the Sun. Again, strongest line is calcium, is getting even stronger. K stars, orange yellow stars, where that hydrogen line gets weak again.

Many many metal lines appear, and then m stars are red stars. Again, many metal lines. Things get complicated. This was her classification scheme. This is what made sense to her to put the ordering of color and various spectral features, that's all she had.

This ordering made sense. Canon had an intuitive sense that this was telling us something not just about how stars are organized in the universe, but potentially how stars evolve, or are born. It tells us something useful. She would publish this, again, very rare at the time. Within a decade, the International Astronomical Union would ratify it as the classification scheme.

And with some small modifications, we have some more subdivisions of some more information because we can measure some more things about stars. It is the classification scheme that we use today. There's a handy mnemonic you can use to remember the ordering of stars. O stars are the blue stars, and m stars are the red stars, and then you have the colors in between. You have o b a f g k m.

If you remember o b a fine guy or girl, kiss me, you can remember the ordering of stellar classifications. And why is it so weird? Why isn't it just 1, 2, 3, 4? Well, we tried 1, 2, 3, that didn't work out. Then we went to alphabet and it was a through o, but then Annie Jump Cannon realized that there needed to be some shakeups here.

And some some of the categories were merged together, some were dropped all completely. It was Annie Jump Cannon who brought order out of the chaos. Shortly after she published her classification scheme, we realized that there was connection here to something that we couldn't directly measure which was stellar temperature. That the o stars and b stars had the hottest surface temperatures, and the k stars and the m stars had the lowest surface temperatures. And from there, our dream came true.

With this classification scheme, we were finally able to unlock stellar evolution. We were able to realize that hotter stars are larger stars and that cooler stars tend to be smaller stars. There are exceptions of course with the red giants. This tells us something oh, I see. When I see an m star I'm probably looking at a very small star, a small cool star.

If I'm looking at a B star, I'm looking at a big bright star. Using Annie Jump Cannon's classification scheme, we're able to get more information. Now I can just take a spectrum of a star, note its color, and now I know its temperature range and now I know its size range, which is stuff I can't directly measure. And once we have size and temperature, we could start playing around with models of how stars work, how they keep themselves hot. Eventually, we will come upon nuclear fusion as the answer.

Once we had nuclear fusion, we could figure out stellar evolution. We say, oh, I get it now. Stars are born from dense clumps of gas, sometimes the clumps are small, sometimes the clumps are big. If they're small, you get an m star or a k star out of it. If the clump is big, you get a b star or an o star.

If it's kinda medium, you get a g star. That size of the star determines its surface temperature, determines its color. It will then evolve and maybe become a red giant or, yeah, whatever. We could figure out the whole story of stellar evolution. And it was Annie Jump Cannon's organization that helped us unlock surface temperature, which helped us unlock size, which helped us unlock evolution of stars.

We would be lost in the interstellar weeds without Annie Jump Cannon, and it was all because she was close to the data, she could look at the data, and she began to develop a sense of what made sense and she nailed it. She totally should have won a Nobel Prize for her work, but women don't get Nobel prizes sadly. That's still kinda true today and that's a problem. Thank you so much for listening. You know, I'd love to do a follow-up episode on Henrietta Swan Levitt, another of the Harvard Computers, so feel free to ask, and I'll I'll do that episode someday soon.

Thank you. Please, please go to patreon.com/pmsutter to help contribute to this show, and thanks to my top Patreon contributors this month, Justin g, Chris l, Barbara k, Duncan m, Corey d, Justin z, Nalia Scott m, Rob h, Justin Lewis m, John w, Alexis Gilbert m, Joshua, John s, Thomas d, Simon g, Aaron j, Justin k k, and Valerie h. That's patreon.com/pmsutter. And thanks to Kevin a on YouTube and Morris a on YouTube for the questions about any jump cannon. Please keep sending me questions and if you wanna hear about another amazing woman that should have also should have won a Nobel Prize, Henrietta Swan Leavitt, feel free to ask.

That's askaspaceman.comoraskaspaceman@gmail.com atpaulmatsutter on all social channels, and I will see you next time for more complete knowledge of time and space.

Comment