How a Total Solar Eclipse Helped Prove Einstein Right About Relativity

That historic experiment was carried out on May 29, 1919. British astronomer Sir Arthur Eddington was paying attention to Einstein's outlandish yet powerful new ideas after getting word from Dutch physicist Willem De Sitter (Holland was a neutral nation during WWI) and realized he could lead an experiment to test the theory.

In 1917, Sir Frank Watson Dyson, Astronomer Royal of Britain, had conceived an experiment that would plot the positions of background stars close to the sun's limb during an eclipse — an experiment that Eddington would lead two years later. If the positions of the stars could be precisely measured during the 1919 eclipse and then compared with their normal positions in the sky, the effects of warped space-time could be observed — beyond what Newton’s classical mechanics would predict. If the position of the stars were altered in exactly the way that Einstein's theory predicted they should be, then this might be just the test general relativity needed. Eddington most likely knew that if this test confirmed general relativity theory, it would turn the view of the Newtonian universe on its head.

Result of the Experiment: Picture that proves Einstein Correct

Here is where my question comes in to play. Einsteins theory of general relativity was accepted because of this first experiment initially. I just have questions about the method of calculation that proved this theory correct.

When this experiment took place, what was the rotational direction of the sun in this photo? And if the rotational direction was the same direction as the displacement of the stars light, then could not the suns rotational direction have effected the light, instead of the idea that space being bent around the suns mass caused this distortion in light?

If space is being bent, then would not this effect occur at all angles of the sun? If it does, are their more photos of this in action? Has anyone tried correlating the displacement of stars light to the rotational direction of the sun, or large bodies of mass that light is passing nearby?


closed as off-topic by Emilio Pisanty, stafusa, Jon Custer, ZeroTheHero, Yashas Oct 11 '17 at 8:22

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "We deal with mainstream physics here. Questions about the general correctness of unpublished personal theories are off topic, although specific questions evaluating new theories in the context of established science are usually allowed. For more information, see Is non mainstream physics appropriate for this site?." – Emilio Pisanty, stafusa, Jon Custer, Yashas
If this question can be reworded to fit the rules in the help center, please edit the question.

  • 2
    $\begingroup$ So what is the question? $\endgroup$ – DanielC Oct 10 '17 at 14:50
  • 4
    $\begingroup$ It is an exasperating mistake to assume that the initial piece of evidence for [insert counter-intuitive phenomenon] is particularly relevant to modern physics, when it has by now been buried under a mountain of additional evidence; when you say you're "concerned" that some small detail could have been missed a century ago, you just come off as a crackpot. We are not here to evaluate your personal theories. If you have a specific question about the calculations, then ask that and just that; if you have a historical question, go to History of Science and Mathematics $\endgroup$ – Emilio Pisanty Oct 10 '17 at 15:14
  • 1
    $\begingroup$ A physical theory is usually expressed using mathematical formulas that predict measurements. For your self-attracting space idea to be a theory, you need to support it with mathematics. $\endgroup$ – safesphere Oct 10 '17 at 15:15
  • 1
    $\begingroup$ The deflection of light has actually been observed for Earth-Sun-Star angles ranging from grazing angle to nearly 90 degrees. General Relativity predicts the whole curve, and this is a parameter-free prediction, by that I mean nothing to adjust to fit the data, and, measured points lie exactly on the predicted curve. See for example "S. S. Shapiro, J. L. Davis, D. E. Lebach, and J. S. Gregory. Measurement of the solar gravitational deflection of radio waves using geodetic very-long-baseline interferometry data, 1979 ̆1999. Phys. Rev. Lett., 92:121101, Mar 2004." and references therein. $\endgroup$ – user154997 Oct 10 '17 at 15:36
  • 1
    $\begingroup$ is there a justification to the “quantum mechanics” and “classical mechanics” tags? $\endgroup$ – ZeroTheHero Oct 10 '17 at 20:50

I think this is a fair question, but it's based on a misunderstanding of what physicists mean by a theory and how they set about testing it.

In physics a theory is just a mathematical model i.e. something that will take data from experiments as input and compute predictions of future experiments as output. Physicists can come up with any mathematical model they want, and over the years there have been some really crazy ones, but at the end of the day it is experiments that test the model to see if its predictions are correct.

The other key point is that experiments cannot prove a theory - they can only disprove it. You do an experiment to test your theory, and if it passes the test that's great but all that proves is that the theory hasn't failed yet. Indeed virtually all our theories ultimately fall foul of that brutal tyrant, experiment. Newton's laws fail at high velocities. Special relativity fails at high densities. General relativity may or may not fail at singularities - physicists still argue about that one.

But let's get back to the specific question you pose. General relativity predicted that the bending of light by the Sun would have a certain value. This was a genuine prediction, and it was done using a theory that on the face of it seemed absolutely bizarre so there was considerable scepticism about it. And ... astonishingly the prediction was correct!

Now, the experiment didn't prove GR was right, it just failed to disprove it. But given how complicated and bizarre the theory was it seemed ridiculous that it got the correct answer just by chance. So while strictly speaking the Eddington experiment didn't prove GR most physicists took the view that it was effectively a proof, and that was why there was so much excitement about it.

Since then we have measured gravitational deflection of light in countless different experiments, some involving rotating bodies and some not, and we can be confident that the effect is not just due to the rotation of the Sun. As it happens rotation does have some effect due to a phenomenon called frame dragging, but the effect is so small that it would only be detectable in very dense rapidly rotating bodies like a neutron star and sadly we have none of those to hand.

  • $\begingroup$ Thank you for taking my question and understanding it for what it was. I am no expert at physics by all means. Its why I come here to ask questions that are to complicated for my understanding. I was just wondering if the rotation of the sun had been taken into consideration when this measurement was taken, which lead to this bizarre question. I felt that maybe if the measurement was taken on the opposite side of the sun, the results "could" have been different. I figured the suns rotation would have "some" effect on light, but just didn't know how much of an effect. $\endgroup$ – shader2199 Oct 10 '17 at 16:46

In its restricted form, your question

could the rotational direction of the sun be partly responsible for the displacement of light in the picture?

does fit the purview of this site, and it can be conclusively answered: no.

It is a mistake to think that that initial piece of evidence, ninety-eight years ago, is the single nail on which we hang all of the support for the experimental validity of general relativity. It's the one we bring up first because when summarizing for non-experts we need to keep things brief, but if you think that there have not been any similar measurements in the intervening century then you're not thinking clearly.

Luckily, though, a quick trip to Wikipedia gives a handy reference to a modern review of similar experiments carried out using radio waves (which are conceptually equivalent to visible light, but don't need solar eclipses to work) for a wide range of sources ─ from grazing incidence to large angles, and located East, West, North and South of the Sun ─ and the results clearly depend on the angular distance between the source and the Sun (the 'solar elongation angle') and not the relative position of the source with respect to the Sun's rotational axis.

What that means is that models that explain the deviation with a nontrivial contribution from the Sun's rotation are necessarily in contradiction with the measured data.

And, on a separate track, this is also incorrect:

Einsteins theory of general relativity was accepted because of this first experiment initially.

No theory is ever fully accepted in one go (and even today we remain open to the possibility of a bigger theory that supersedes GR). Eddington's 1919 measurements were a huge piece of evidence in favour of GR, but it was not the only one (since GR already had relevant evidence, such as the precession of Mercury's perihelion) and it would be joined over the years by a vast array of other evidence from a variety of sources. It is this array, as a whole, that gives us confidence in GR, not any one piece of evidence individually.

Moreover, it's important to note that Eddington's measurements were important not just because they agreed with GR, but because they disagreed with newtonian physics, and it is this latter aspect that gives rise to the (justified) tone of e.g. the New York Times headline, or what you describe as 'turning the Newtonian Universe on its Head'. This is true regardless of the validity of GR: the experiments did not agree with the existing newtonian physics, which was otherwise very solid, and any modification (including possible alternatives, say, based on the Sun's rotation, which are ruled out by data collected later) would still have 'turned the newtonian universe on its head'.

  • $\begingroup$ Emilio, I do not find criticism, unless its constructive, very helpful. You made your comment in the original post, and I believe you got your point across the first time. When I originally thought of my question, I was wondering something. I was wondering "what if" the same kind of measurement was taken at different angles of the sun. Would that lead to data that could help me better understand the theory of general relativity? Be civil and considerate about the people that post. I am sure you understood what my question was. Instead of helping me understand, you chose to criticize me though. $\endgroup$ – shader2199 Oct 10 '17 at 21:22
  • $\begingroup$ This is the answer to the question as you posed it: you asked if a specific explanation is viable, and the answer is no, because it would have very different symmetry to the GR prediction, and that's pretty much the first thing we check when analysing data - and, as I said, there's a mountain of data in addition to Eddington's, some of it linked above, that explicitly rules out asymmetries of the kind you asked about. $\endgroup$ – Emilio Pisanty Oct 10 '17 at 23:02
  • $\begingroup$ More generally, it is extremely important that you distinguish questions about the theory as we understand it today and historical questions about how that theory was built, and that you ask them separately. (If you don't, you get a confused mess, like the current form of this one, that's much harder to answer.) Historical questions, like "what alternatives to GR are considered before it was adopted?", are not without interest but they are ultimately irrelevant to the modern understanding and they're not within our purview - take them to History of Science and Mathematics instead. $\endgroup$ – Emilio Pisanty Oct 10 '17 at 23:07

Not the answer you're looking for? Browse other questions tagged or ask your own question.