Tag Archives: DoingScience

What’s up with arXiv?

Leave a reply

First of all, I wanted to take a moment to say that this is the one-year anniversary of this blog. I’ve been posting every week, (almost always) on Friday, since I first was motivated to start blogging back in November 2012. It’s been a fun ride, through ups and downs, Ars Technica and Amplituhedra, and I hope it’s been fun for you, the reader, as well!

I’ve been giving links to arXiv since my very first post, but I haven’t gone into detail about what arXiv is. Since arXiv is a rather unique phenomenon, it could use a more full description.

arXivpic

The word arXiv is pronounced much like the normal word archive, just think of the capital X like a Greek letter Chi.

Much as the name would suggest, arXiv is an archive, specifically a preprint archive. A pre-print is in a sense a paper before it becomes a paper; more accurately, it is a scientific paper that has not yet been published in a journal. In the past, such preprints would be kept by individual universities, or passed between interested individuals. Now arXiv, for an increasing range of fields (first physics and mathematics, now also computer science, quantitative biology, quantitative finance, and statistics) puts all of the preprints in one easily accessible, free to access place.

Different fields have different conventions when it comes to using arXiv. As a theoretical physicist, I can only really speak to how we use the system.

When theoretical physicists write a paper, it is often not immediately clear which journal we should submit it to. Different journals have different standards, and a paper that will gather more interest can be published in a more prestigious journal. In order to gauge how much interest a paper will raise, most theoretical physicists will put their papers up on arXiv as preprints first, letting them sit there for a few months to drum up attention and get feedback before formally submitting the paper to a journal.

The arXiv isn’t just for preprints, though. Once a paper is published in a journal, a copy of the paper remains on arXiv. Often, the copy on arXiv will be updated when the paper is updated, changed to the journal’s preferred format and labeled with the correct journal reference. So arXiv, ultimately, contains almost all of the papers published in theoretical physics in the last decade or two, all free to read.

But it’s not just papers! The digital format of arXiv makes it much easier to post other files alongside a paper, so that many people upload not just their results, but the computer code they used to generate them, or their raw data in long files. You can also post papers too long or unwieldy to publish in a journal, making arXiv an excellent dropping-off point for information in whatever format you think is best.

We stand at the edge of a new age of freely accessible science. As more and more disciplines start to use arXiv and similar services, we’ll have more flexibility to get more information to more people, while still keeping the advantage of peer review for publication in actual journals. It’s going to be very interesting to see where things go from here.

Blackboards, Again

4 Replies

Recently I had the opportunity to give a blackboard talk. I’ve talked before about the value of blackboards, how they facilitate collaboration and can even be used to get work done. What I didn’t feel the need to explain was their advantages when giving a talk.

No, the blackboard behind me isn't my talk.

No, the blackboard behind me isn’t my talk.

When I mentioned I was giving a blackboard talk, some of my friends in other fields were incredulous.

“Why aren’t you using PowerPoint? Do you people hate technology?”

So why do theorists (and mathematicians) do blackboard talks, when many other fields don’t?

Typically, a chemist can’t bring chemicals to a talk. A biologist can’t bring a tank of fruit flies or zebrafish, and a psychologist probably shouldn’t bring in a passel of college student test subjects. As a theorist though, our test subjects are equations, and we can absolutely bring them into the room.

In the most ideal case, a talk by a theorist walks you through their calculation, reproducing it on the blackboard in enough detail that you can not only follow along, but potentially do the calculation yourself. While it’s possible to set up a calculation step by step in PowerPoint, you don’t have the same flexibility to erase and add to your equations, which becomes especially important if you need to clarify a point in response to a question.

Blackboards also often give you more space than a single slide. While your audience still only pays attention to a slide-sized area of the board at one time, you can put equations up in one area, move away, and then come back to them later. If you leave important equations up, people can remind themselves of them on their own time, without having to hold everybody up while you scroll back through the slides to the one they want to see.

Using a blackboard well is a fine art, and one I’m only beginning to learn. You have to know what to erase and what to leave up, when to pause to allow time to write or ask questions, and what to say while you’re erasing the board. You need to use all the quirks of the medium to your advantage, to show people not just what you did, but how and why you did it.

That’s why we use blackboards. And if you ask why we can’t do the same things with whiteboards, it’s because whiteboards are terrible. Everybody knows that.

Dammit Jim, I’m a Physicist not a Graphic Designer!

4 Replies

Over the last week I’ve been working with a few co-authors to get a paper ready for publication. For my part, this has mostly meant making plots of our data. (Yes, theorists have data! It’s the result of calculations, not observations, but it’s still data!)

As it turns out, making the actual plots is only the first and easiest step. We have a huge number of data points, which means the plots ended up being very large files. To fix this I had to smooth out the files so they don’t include every point, a process called rasterizing the images. I also needed to make sure that the labels of the plots matched the fonts in the paper, and that the images in the paper were of the right file type to be included, which in turn meant understanding the sort of information retained by each type of image file. I had to learn which image files include transparency and which don’t, which include fonts as text and which use images, and which fonts were included in each program I used. By the end, I learned more about graphic design than I ever intended to.

In a company, this sort of job would be given to a graphic designer on-staff, or a hired expert. In academia, however, we don’t have the resources for that sort of thing, so we have to become experts in the nitty-gritty details of how to get our work in publishable form.

As it turns out, this is part of a wider pattern in academia. Any given project doesn’t have a large staff of specialists or a budget for outside firms, so everyone involved has to become competent at tasks that a business would parcel out to experts. This is why a large part of work in physics isn’t really physics per se; rather, we theorists often spend much of our time programming, while experimentalists often have to build and repair their experimental apparatus. The end result is that much of what we do is jury-rigged together, with an amateur understanding of most of the side disciplines involved. Things work, but they aren’t as efficient or as slick as they could be if assembled by a real expert. On the other hand, it makes things much cheaper, and it’s a big contributor to the uncanny ability of physicists to know about other peoples’ disciplines.

Talks, and what they’re good for

1 Reply

It’s an ill-kept secret that basically everyone in academia is a specialist. Nobody is just a “physicist”, or just a “high energy theorist”, or even just a “string theorist”. Even when I describe myself as something as specific as an “amplitudeologist”, I’m still over-generalizing: there’s a lot of amplitudes work out there that I would be hard-pressed to understand, and even harder-pressed to reproduce.

In the end, each of us is only going to understand a small subset of the vastness of our subject. This is problematic when it comes to attending talks.

Rarely, we get to attend talks about something we completely understand. Generally, we’re the ones giving those talks. The rest of the time, even at conferences for people of our particular specialty, we’re going to miss some fraction of the content, either because we don’t understand it or because we don’t find it interesting.

The question then becomes, why attend the talk in the first place? Why spend an hour of your time when you’re not getting an hour’s worth of content?

There are a couple reasons, of varying levels of plausibility.

One is that it’s always nice to know what other subfields are doing. It lets one feel connected to one’s compatriots, and it helps one navigate one’s career. That said, it’s unclear whether going to talks is really the best way of doing this. If you just want to know what other people are doing, you can always just watch to see what they publish. That doesn’t take an hour, unless you’re really dedicated to wasting time.

A more important benefit is increasing levels of familiarity. These days, I can productively pay attention to the first quarter of a talk, half if it’s particularly good. When I first got to grad school, I’d probably tune out after the first five minutes. The more talks you see on a subject, the more of the talk makes sense, and the more you get out of it. That’s part of why even fairly specialized people who are further along in their careers can talk on a wide range of subjects: often, they’ve intentionally kept themselves aware of what’s going on in other subfields, going to talks, reading papers, and engaging in conversation. This is a valuable end goal, since there is some truth to the hype about the benefits of interdisciplinarity in providing unconventional solutions to problems. That said, this is a gradual process. The benefit of one individual talk is tiny, and it doesn’t seem worth an hour of time. Much like exercise, it’s the habit that provides the benefit, not any individual session.

So in the end, talks are almost always unsatisfying. But we keep going to them, because they make us better scientists.

You get paid to learn. How bad can that be?

Leave a reply

In my “who am I” post, I describe being a grad student as like being an apprentice. I’d like to elaborate on that.

Ph.D. programs in the sciences are different at every school, but they have a few basic features. Generally you enter them with a bachelor’s degree from another university. The program lasts for somewhere between four and six years, longer for particularly unfortunate cases. Sometimes you get a Master’s degree after the first two years, sometimes you don’t, but you don’t usually have to get it from another school. Generally the first two years mostly involve taking courses while the later years are mostly research, but this can vary as well. And in general, once you’re in the program, you get paid: either as a Teaching Assistant, in which case you help grade papers, lead lab sections, and sometimes give lectures, or as a Research Assistant, in which you are paid to do research.

This last is occasionally confusing to people. If a Ph.D. student learns by doing research, then why are they also paid to do research? That sounds like not just getting your education for free, but being paid for it, which sounds at the very least like a very good deal.

There are two ways to think about the situation. One, as I mentioned in my “who am I” post, is as an apprenticeship. An apprentice is expected to learn on the job, and provided they learn enough they are eventually certified to work on their own. Despite this, an apprenticeship is still very much a job. An apprentice is subservient to their master, and can generally be counted upon to work on the master’s projects and help the master in their job. In much the same way, a Ph.D. student is not certified to work on their own until they graduate from the program and obtain their Ph.D. In the meantime they are subservient to their advisor, and they have to take their advisor’s desires into account when choosing research projects. In general, most of a grad student’s research projects will be part of their advisor’s research in one way or another, furthering their advisor’s goals. Beyond the research itself, grad students will often have other duties, depending on the nature of their advisor’s work, especially if their advisor has a lab with complicated equipment that needs to be maintained.

The other thing to realize is that grad students are, ostensibly, part-time workers. The university pays me for 20 hours a week of work. The thing is, though, I don’t just work part-time. I work full-time. I also work at home, on the weekends…whenever I can make progress on my research (and I’m not doing some side project like this blog or taking a needed sanity break), I work. So if I work 40 hours a week and am paid for 20, that means I am effectively spending half my income on education.

Not so free, is it?

It’s not as if any of us could just work less and take on another part-time job, either. Apart from the fact that many grad students are international students on visas that don’t allow them to get other jobs, it is research itself: keeping up, making progress, working towards graduating, that takes up so much of our time. To get any education out of the process at all, we have to be involved as much as possible.  So we are, inevitably, paying for our education. And hopefully, we’re getting something out of it.

Blackboards

2 Replies

As a college student, I already knew that theoretical physicists weren’t like how they were portrayed in movies. They didn’t wear lab coats, or have universally frizzy, unkempt white hair. I knew they didn’t have labs, or plot to take over the world. And I was pretty sure that they didn’t constantly use blackboards.

After all, blackboards are a teaching tool. They’re nice for getting equations up so that the guy way in the back can see them. But if you were actually doing a real calculation, surely you’d prefer paper, or a computer, or some other method that doesn’t involve an unkempt scrawl and a heap of loose white dust all over your clothing.

Right?

Right?

Over the last few years I’ve come to appreciate the value of blackboards. Blackboards actually can be used for calculations. You don’t want to use them all the time, but there are times when it’s useful to have a lot of room on a page, to be able to make notes and structure the board around concepts. More importantly, though, there is a third function that I didn’t even consider back in college. Between calculation and teaching, there is collaboration.

Go to a physics or math department, and you’ll find blackboards on the walls. You’ll find them not just in classrooms, but in offices, and occasionally in corridors. Go to a high-class physics location like the Perimeter Institute or the Simons Center, and they’ll brag to you about how many blackboards they have strewn around their common areas.

The purpose of these blackboards is to facilitate conversation. If you want to explain your work to someone else and you aren’t using a blog post, you need space to write in a way that you can both see what you’re doing. Blackboards are ideal for that sort of conversation, and as such are essential for collaboration and communication among scientists.

What about whiteboards? Well, whiteboards are just evil, obviously.

Perimeter and Patronage

3 Replies

I’m visiting the Perimeter Institute this week. For the non-physicists in the audience, Perimeter is a very prestigious institute of theoretical physics, founded by the founder of BlackBerry. It’s quite swanky. Some first impressions:

  • This occurred to me several times: this place is what the Simons Center wants to be when it grows up.
  • You’d think that the building is impossible to navigate because it was designed by a theoretical physicist, but Freddy Cachazo assured us that he actually had to get the architect to tone down the impossibly ridiculous architecture. Looks like the only person crazier than a physicist is an artist.
  • Having table service at an institute café feels very swanky at first, but it’s actually a lot less practical than cafeteria-style dining. I think the Simons Center Café has it right on this one, even if they don’t quite understand the concept of hurricane relief (don’t have a link for that joke, but I can explain if you’re curious).
  • Perimeter has some government money, but much of its funding comes from private companies and foundations, particularly Research in Motion (or RIM, now BlackBerry). Incidentally, I’m told that PeRIMeter is supposed to be a reference to RIM.

What interests me is that you don’t see this sort of thing (private support) very often in other fields. Private donors will found efforts to solve some real-world problem, like autism or income inequality. They rarely fund basic research*. When they do fund basic research, it’s usually at a particular university. Something like Perimeter, a private institute for basic research, is rather unusual. Perimeter itself describes its motivation as something akin to a long-range strategic investment, but I think this also ties back to the concept of patronage.

Like art, physics has a history of being a fashionable thing for wealthy patrons to support, usually when the research topic is in line with their wider interests. Newton, for example, re-cast his research in terms of its implications for an understanding of the tides to interest the nautically-minded King James II, despite the fact that he couldn’t predict the tides any better than anyone else in his day. Much like supporting art, supporting physics can allow someone’s name to linger on through history, while not running a risk of competing with others’ business interests like research in biology or chemistry might.

A man who liked his sailors

*basic research is a term scientists use to refer to research that isn’t made with a particular application in mind. In terms of theoretical physics, this often means theories that aren’t “true”.

Model-Hypothesis-Experiment: Sure, Just Not All the Same Person!

Leave a reply

At some point, we were all taught how science works.

The scientific method gets described differently in different contexts, but it goes something like this:

First, a scientist proposes a model, a potential explanation for how something out in the world works. They then create a hypothesis, predicting some unobserved behavior that their model implies should exist. Finally, they perform an experiment, testing the hypothesis in the real world. Depending on the results of the experiment, the model is either supported or rejected, and the scientist begins again.

It’s a handy picture. At the very least, it’s a good way to fill time in an introductory science course before teaching the actual science.

But science is a big area. And just as no two sports have the same league setup, no two areas of science use the same method. While the central principles behind the method still hold (the idea that predictions need to be made before experiments are performed, the idea that in order to test a model you need to know something it implies that other models don’t, the idea that the question of whether a model actually describes the real world should be answered by actual experiments…), the way they are applied varies depending on the science in question.

In particular, in high-energy particle physics, we do roughly follow the steps of the method: we propose models, we form hypotheses, and we test them out with experiments. We just don’t expect the same person to do each step!

In high energy physics, models are the domain of Theorists. Occasionally referred to as “pure theorists” to distinguish them from the next category, theorists manipulate theories (some intended to describe the real world, some not). “Manipulate” here can mean anything from modifying the principles of the theory to see what works, to attempting to use the theory to calculate some quantity or another, to proving that the theory has particular properties. There’s quite a lot to do, and most of it can happen without ever interacting with the other areas.

Hypotheses, meanwhile, are the province of Phenomenologists. While theorists often study theories that don’t describe the real world, phenomenologists focus on theories that can be tested. A phenomenologist’s job is to take a theory (either proposed by a theorist or another phenomenologist) and calculate its consequences for experiments. As new data comes in, phenomenologists work to revise their theories, computing just how plausible the old proposals are given the new information. While phenomenologists often work closely with those in the next category, they also do large amounts of work internally, honing calculation techniques and looking through models to find explanations for odd behavior in the data.

That data comes, ultimately, from Experimentalists. Experimentalists run the experiments. With experiments as large as the Large Hadron Collider, they don’t actually build the machines in question. Rather, experimentalists decide how the machines are to be run, then work to analyze the data that emerges. Data from a particle collider or a neutrino detector isn’t neatly labeled by particle. Rather, it involves a vast set of statistics, energies and charges observed in a variety of detectors. An experimentalist takes this data and figures out what particles the detectors actually observed, and from that what sorts of particles were likely produced. Like the other areas, much of this process is self-contained. Rather than being concerned with one theory or another, experimentalists will generally look for general signals that could support a variety of theories (for example, leptoquarks).

If experimentalists don’t build the colliders, who does? That’s actually the job of an entirely different class of scientists, the Accelerator Physicists. Accelerator physicists not only build particle accelerators, they study how to improve them, with research just as self-contained as the other groups.

So yes, we build models, form hypotheses, and construct and perform experiments to test them. And we’ve got very specialized, talented people who focus on each step. That means a lot of internal discussion, and many papers published that only belong to one step or another. For our subfield, it’s the best way we’ve found to get science done.

Ansatz: Progress by Guesswork

Leave a reply

I’ve talked before about how hard traditional Quantum Field Theory is. Building things up step by step is slow and inefficient. And like any slow and inefficient process, there is a quicker way. An easier way. A…riskier way.

You guess.

Guess is such an ugly word, though…so let’s call it an ansatz.

Ansatz is a word of German origin. In German, it is part of various idiomatic expressions, where it can refer to an approach, an attempt, or a starting point. When physicists and mathematicians use the term ansatz, they mean a combination of all of these.

An ansatz is an approach in that it is a way of finding a solution to a problem without using more general, inefficient methods. Rather than approaching problems starting from the question, an ansatz approaches problems by starting with an answer, or rather, an attempt at an answer.

An ansatz is an attempt in that it serves as researcher’s best first guess at what the answer is, based on what they know about it. This knowledge can come from several sources. Sometimes, the question constrains the answer, ruling out some possibilities or restricting the output to a particular form. Usually, though, the attempt of an ansatz goes beyond this, incorporating the scientist’s experience as to what sorts of answers similar questions have had in the past, even if it isn’t understood yet why those sorts of answers are common. With information from both of these sources, a scientist comes up with a preliminary guess, or ansatz, as to answer to the problem at hand.

What if the answer is wrong, though? The key here is that an ansatz is only a starting point. Rather than being a full answer with all the details filled in, an ansatz generally leaves some parameters free. These free parameters represent unknowns, and it is up to further tests to fix their values and complete the answer. These tests can be experimental, but they can also be mathematical: often there are restrictions on possible answers that are difficult to apply when creating a first guess, but easier to apply when one has only a few parameters to fix. In order to avoid the risk of finding an ansatz that only works by coincidence, many more tests are done than there are parameters. That way, if the guess behind the ansatz is wrong, then some of the tests will give contradictory rules for the values of the parameters, and you’ll know that it’s time to go back and find a better guess.

In the end, this approach, using your first attempt as a starting point, should end up with only a few parameters free, ideally none at all. One way or another, you have figured out a lot about your question just by guessing the answer!

The use of ansatzes is quite common in theoretical physics. Some of the most interesting problem either can’t be solved or are tedious to solve through traditional means. The only way to make progress, to go beyond what everyone else can already do, is to notice a pattern, make a guess, and hope you get lucky. Well, not just a guess: an ansatz.

Breakthrough or Crackpot?

1 Reply

Suppose that you have an idea. Not necessarily a wonderful, awful idea, but an idea that seems like it could completely change science as we know it. And why not? It’s been done before.

My advice to you is to be very very careful. Because if you’re not careful, your revolutionary idea might force you to explain much much more than you expect.

Let’s consider an example. Suppose you believe that the universe is only six thousand years old, in contrast to the 13.772 ± 0.059 billion years that scientists who study the subject have calculated. And furthermore, imagine that you’ve gone one step further: you’ve found evidence!

Being no slouch at this sort of thing, you read the Wikipedia article linked above, and you figure you’ve got two problems to deal with: extrapolations from the expansion of the universe, and the cosmic microwave background. Let’s say your new theory is good enough that you can address both of these: you can explain why calculations based on both of these methods give 14 billion years, while you still assert that the universe is only six thousand years old. You’ve managed to explain away all of the tests that scientists used to establish the age of the universe. If you can manage that, you’re done, right?

Not quite. Explaining all the direct tests may seem like great progress, but it’s only the first step, because the age of the universe can show up indirectly as well. No stars have been observed that are 13.772 billion years old, but every star whose age has been calculated has been found to be older than six thousand years! And even if you can explain why every attempt to measure a star’s age turned out wrong, there’s more to it than that, because the age of stars is a very important part of how astronomers model stellar behavior. Every time astronomers make a prediction about a star, whether estimating its size, it’s brightness, its color, every time they make such a prediction and then the prediction turns out correct, they’re using the fact that the star is (some specific number) much much older than six thousand years. And because almost everything we can see in space either is made of stars, or orbits a star, or once was a star, changing the age of the universe means you have to explain all those results too. If you propose that the age of the universe is only six thousand, you need to explain not only the cosmic microwave background, not only the age of stars, but almost every single successful prediction made in the last fifty years of astronomy, none of which would have been successful if the age of the universe was only six thousand.

Daunting, isn’t it?

Oh, we’re not done yet!

See, it’s not just astronomy you have to contend with, because the age of the Earth specifically is also calculated to be much larger than six thousand years. And just as astronomers use the age of stars to make successful predictions about their other properties, geologists use the age of rock formations to make their own predictions. And the same is true for species of animals and plants, studied through genetic drift with known rates over time, or fossils with known ages. So in proposing that the universe is only six thousand years old, you need to explain not just two pieces of evidence, but the majority of successful predictions made in three distinct disciplines over the last fifty years. Is your evidence that the universe is only six thousand years old good enough to outweigh all of that?

This is one of the best ways to tell a genuine scientific breakthrough from ideas that can be indelicately described as crackpot. If your idea questions something that has been used to make successful predictions for decades, then it becomes your burden of proof to explain why all those results were successful, and chances are, you can’t fulfill that burden.

This test can be applied quite widely. As another example, homeopathic medicine relies on the idea that if you dilute a substance (medicine or poison) drastically then rather than getting weaker it will suddenly become stronger, sometimes with the reverse effect. While you might at first think this could be confirmed or denied merely by testing homeopathic medicines themselves, the principle would also have to apply to any other dilution, meaning that a homeopath needs to explain everything from the success of water treatment plants that wash out all but tiny traces of contaminants to high school chemistry experiments involving diluting acid to observe its pH.

This is why scientific revolutions are hard! If you want to change the way we look at the world, you need to make absolutely sure you aren’t invalidating the success of prior researchers. In fact, the successes of past research constrain new science so much, that it sometimes is possible to make predictions just from these constraints!

So whenever you think you’ve got a breakthrough, ask yourself: how much does this mean I have to explain? What is my burden of proof?