In this week's "Ask a Physicist," we deal with a crazy and exciting question. Is the Fine Structure Constant really a constant? Maybe. Okay, probably.

I have a secret to tell you. More than 90% of the cutting edge science that you read in the news is wrong. There's great fanfare when somebody announces that they have a theory which would overturn Einstein or quantum mechanics, but complete silence when it turns out that the theory makes no predictions, or worse, is already demonstrated to be wrong by what we know. Scientific revolutions happen pretty infrequently, which is why they're a big deal. This is why I tend not to report on new observations or ostensibly groundbreaking theories.


I'll make an exception today. For one thing, it's Mad Science Week here at io9, so why not go out on a limb? But for another, an interesting set of observations was announced a few weeks ago that got my attention.

A number of readers including a former student of mine, Brad Hubartt, wrote to ask me whether the report about a variable "fine structure constant" is, in fact, the real deal:

I was wondering what your take was on a recent article. It is about a variable fine structure constant which would imply at least one of our important constants might change depending on your location in the universe.


Since Brad didn't bother to post his question in the form of a question, I'm going to do a Q&A of everything you might want to know about the fine structure "constant" debate.

Let's start off easy.

What is the fine structure constant?

Before I get into that, I'm going to give you a freebie answer to another frequently asked question: Why is the speed of light 300,000 km/s rather than something else?


Not to be contrarian, but that's not the right question. The real question is why are kilometers defined so that a light travels 300,000 of them in a second? What I'm getting at is that many of the things that we think of as constants in nature only have particular values because of the units we choose to use. If we used, instead, seconds and lightseconds, then the speed of light would be a satisfying 1 lightsecond/second.

I bring all of this up because there are a few ways of combining physical constants so that all of the constants drop out. In particular, the number:

alpha=2*pi e2/hc

is completely unitless. That's your one equation (already one more than I normally present), so I'd better explain what the numbers mean. "alpha" is the thing we're talking about, the fine structure constant (FSC). It has a value of about 1/137.03599908. I've included all of those extra decimal places because we can measure this number to ridiculous accuracy, at least on earth.


On the right hand side, "h" is known as Planck's constant, and it's used whenever quantum mechanics is involved. "c" is the speed of light, and "e" is the charge of the electron.

So the FSC is a big mashup of all of these other constants. More importantly, this number is basically the magical constant that tells us how strong electromagnetism is. For example, hydrogen (and all other atoms) only emits light at certain energies. Those energies are completely determined by taking the square of the FSC.

But here's the weird thing: we have absolutely no idea where the value of the FSC comes from.


You might be tempted to think that we don't know nothing about nothing, but that's not really the case. If you've taken a few years of physics and decide to dip your toe into field theory, you'll discover something amazing. We can start with a mathematically simple idea that the physics of the universe won't change if we multiply all of my quantum mechanical wave functions by a completely arbitrary complex field. Never mind why we'd want to do this — we just can. It's called a "Local U(1) gauge transformation" if you're looking to try it out in bars. From this fairly simple assumption, we can literally derive all of electricity and magnetism. It is awesome.

There's just one missing detail — everything that comes out of this calculation is only true up to some constant of proportionality. In other words, we have absolutely no way of determining what the FSC (or the electron charge, for that matter) is except by measuring it.

What does this mean? It means that either the FSC is something we're just not smart enough to predict yet, or it means that our universe (or our part of the universe) just happens to have an FSC suitable for life. Vary the FSC by just a few per cent and we wouldn't be able to produce heavy elements. This is an "anthropic principle" argument, and as I've said before, most physicists really don't like this line of reasoning. But if the FSC varies a lot in space and time then there's the unsettling fact that only a few patches might be suitable for life, and naturally we're going to be living in one of the special places or we wouldn't be here asking the questions.


What's the wrench in the works?

So first, let's take this with a grain of salt. I'm going to tell you about a result, but that doesn't mean it's right. (And worse yet, you're still not going to know by the end of this article).


So first, a bit of physics. If you heat up a bunch of atoms, they radiate light, but only at particular wavelengths, depending on which kinds of atoms they are. This is why, for example, neon signs are a particular shade of red. On the other hand, if you take a very bright light source and shine it through a cloud of gas, light will only be absorbed at particular wavelengths.

Space is filled with clouds of gas, and there are very distant very bright sources of radiation, called quasars. John Webb of the University of New South Wales and his collaborators looked at closely spaced absorption lines from various ions of iron and magnesium in very distant clouds. By measuring the relative wavelengths of the two absorption lines, they could compare the FSC at cosmological distances to what we measure in the lab here on earth.

I know what you're thinking (at least if you're a part-time astronomer). How do you know if the shift in the spectral lines is due to a change in the FSC rather than just the ordinary redshifts that we measure for all distant objects? My, you're clever! The answer is that the energy of the average of the two lines in the doublet more or less tells you the redshift, and the separation between the two tells you the shift in the FSC from here on earth.


To make things clearer, Julian Berengut of UNSW made the following diagram:

About a decade ago, the UNSW team found, much to everyone's surprise, that billions of light years away, the FSC was slightly smaller than it is here on earth. The difference is pretty miniscule, however, only about 1 part in a hundred thousand. In other words, the physics at the other end of the physical universe would look nearly (but not exactly) like it does here on earth. That means that the diagram above shows an effect about 10,000 times larger than the group actually observed. The signal is small enough that people are right to be concerned about whether or not the UNSW team got their errorbars right.


Most physicists didn't really buy their original result (and, spoiler alert, many still don't), since the constraints were pretty weak and the effect quite small. There was a paper by a competing group which purported to find no result at all. And for the most part, the result was forgotten until recently. In their latest work, they look in the opposite direction in the sky and found that the FSC was a tiny bit higher than it is here on earth. Originally, the UNSW team claimed that the FSC changed with time, but both earth-based laboratory experiments and their own recent evidence seemed to suggest that the bigger effect is variation in space.

If the result is true, it's a very big deal. It means that somehow (and remember, we don't know where the value of the FSC comes from in the first place) the FSC varies throughout the universe. This flies in the face of the "Cosmological Principle" which says, more or less, "Wherever you go, there you are." Or, to put it another way, it assumes that on large scales, the properties of the universe are the same.

There are almost certainly those who will pounce on this in the comments section as a refutation of dark matter, dark energy, or some other part of the standard cosmological model. It isn't. If the result is real, it's still insanely small and says more about how fundamental physics works than what's going on in our local patch of the universe.


Are they right?

If you look around the interwebs or talk to your friendly neighborhood astrophysicists, you'll hear a lot of skepticism. You'll see comments to the effect that, "We don't know how the fine structure constant could vary," or ,"We'd have to rethink a lot of what we know about physics." This is true. While we wouldn't have to throw out the laws of physics, it would mean that the parameters of those laws would become very much a function of our local universe, and we don't know how that could possibly work.

And that would be too damn bad. A lot of you gave me grief about my dark energy article, because you thought we were just being lazy in explaining the universe that way. Well, when you have a set of observations that don't fit your model, you have to change your model. That's just how science works.


The biggest criticisms that most people have made is that it's possible that Webb and his collaborators haven't correctly analyzed the errors in their measurements and his "signal" is really "noise." This is a totally fair criticism, and if you ask me, it's one most likely to shake out of all of this. One possibility (which the team discusses but dismisses) is that it's possible to get the same sort of result that the UNSW team found if some of the clouds had very unusual isotope ratios. Very unusual.

Others have noticed that the UNSW team got very lucky with the orientation of their telescopes and the variation of the FSC. The first set of observations are taken in the North, were the FSC is apparently lower than on earth, while the second set of observations are taken in the South where the opposite is true. As it happens, where the two sets of measurements overlap (near the equator) the signal is just about zero. Of course, this doesn't disprove anything, but it's suspicious because it's unlikely that the FSC signal just happened to line up with the orientation of how the team took their observations.


But the reason that I'm talking about this at all is that it's not sufficient to nitpick at what might cause problems. You wanna disprove the effect, you take some observations and show that the signal isn't repeatable. Or, if you don't want to go through all of that, you could at very least show where Webb screwed up in their error analysis. And don't look at me. I'm a theorist.

Dave Goldberg is the author, with Jeff Blomquist, of "A User's Guide to the Universe: Surviving the Perils of Black Holes, Time Paradoxes, and Quantum Uncertainty." (follow us on twitter, facebook or our blog.) He is an Associate Professor of Physics at Drexel University. Feel free to send email to with any questions about the universe.