Is Einstein's theory of GR shaky

[robertjford80]

This is from Robert March's book Physics for Poets:

There have been a few somewhat shaky astronomical confirmations of some of these predictions, notably that of the bending of the path of starlight as it passes near the sun (a phenomenon observable only during solar eclipses and requiring fantastically precise measurement;, and Einstein’s explanation of some long-known irregularities in the orbit of the planet Mercury. But in both cases the measurements are not sufficiently precise to really provide a severe test of the theory, and there are those who maintain that even these comparatively crude measurements disagree with Einstein’s theory

How accurate is the above statement?

 

[jtbell]

There have been many more tests of GR. The Wikipedia page lists a bunch of them.

http://en.wikipedia.org/wiki/Tests_of_general_relativity

(Of course, it's difficult to vouch for its accuracy at any given moment)

Here's a review article published in a journal a few years ago:

http://arxiv.org/abs/0806.1731

 

[HallsofIvy]

It depends upon what you mean by "shaky". No physical theory can be perfect. There is always going to be some, as yet unknown, data that will not be accounted for by the theory. But relativity fits the known facts so well that whatever does supercede relativity will have to include it as a "special case" or "approximation".

 

[robertjford80]

The author said that the measurements were rather crude and not very precise. Is that right?

 

[Mentz114]

... Einstein’s explanation of some long-known irregularities in the orbit of the planet Mercury. But in both cases the measurements are not sufficiently precise to really provide a severe test of the theory, and there are those who maintain that even these comparatively crude measurements disagree with Einstein’s theory

Those who maintain this are wrong. The precession of the perihelion of Mercury is a very tiny relativistic effect and it is explained to a high accuracy by GR.

If Robert March cares as much for GR as I do for poetry, then you can disregard what he says about GR.

 

[phinds]

The author said that the measurements were rather crude and not very precise. Is that right?

No. The initial eclipsed-sun measurements WERE "shaky", even to the extent that the experimenters disagreed on the interpretation of the results, but that was nearly 100 years ago.

Perhaps March's knowledge of physics is 100 years old. It certainly isn't current.

 

[Dickfore]

Who in the world is Robert March?

EDIT:

Oops, he's a high-energy (emeritus) professor at University Wisconsin-Madison.

 

[Matterwave]

Is this book published before the Gravity Probe B results were? I don't think you can call those "crude" by any means.

 

[Mentz114]

Who in the world is Robert March?

EDIT:

Oops, he's a high-energy (emeritus) professor at University Wisconsin-Madison.

Ah, that explains it.

 

[Dickfore]

I don't know when 'crude' got identified with 'shaky'.

The author (supposedly) objects that some of the earlier measurements were not precise enough. Of course, if you judge by the future standards, every past experiment is too 'crude'. For example, is Coulomb's experiment with the torsion balance the end all proof of Coulomb's Law in electrostatics?

One needs to remember the distinction between 'precise' and 'accurate'. The Gran Sasso neutrino time-of-flight measurements were very precise. However, they weren't accurate, because they had a large systematic error. I would say this is more 'shaky'.

 

[Dale]

I really like "there are those who maintain that". It is a phrase which can be used to justify absolutely anything. "There are those who maintain that werewolves are real." Or "There are those who maintain that the Earth is flat."

 

[atyy]

The first measurements of the bending of starlight may have been shaky - I don't believe there is consensus on that. The current solar system tests are pretty tight, and the bending of starlight outside of the solar system has even been used to study dark matter. A very famous test of GR is the binary pulsar observations of Taylor and Hulse. Of course, if dark matter doesn't exist, that would mean GR is wrong.

Example of an evaluation of early experiments as "ambiguous": http://aapt.org/doorway/TGRUTalks/Weiss/WeissTalk1of9.htm (p3)
Pretty up-to-date Review of GR tests: http://relativity.livingreviews.org/Articles/lrr-2006-3/fulltext.html
Bending of light or "gravitational lensing" to study dark matter: http://home.slac.stanford.edu/pressreleases/2006/20060821.htm

 

[kurros]

Surely the significance of the starlight bending observations was not a precise agreement with GR, but rather that any bending occurred at all. Sure, I have heard modern arguments for the bending of light in Newtonian gravity (although still using the relativistic relationship between mass and energy), but at the time many people expected that there should be zero bending of light by gravity, if I remember correctly. Therefore the observation of any bending at all would be a big deal.

 

[robertjford80]

ok, it looks like the board is unanimous that March is wrong. Someone mentioned the gravity probe B project. I've heard of that. I looked up the measurements on wiki and it said

In an article published in the journal Physical Review Letters in 2011, the authors reported analysis of the data from all four gyroscopes results in a geodetic drift rate of −6,601.8±18.3 milliarcsecond/year (mas/yr) and a frame-dragging drift rate of −37.2±7.2 mas/yr, to be compared with the GR predictions of −6,606.1 mas/yr and −39.2 mas/yr, respectively

I think that book was written in something like 1979, so definitely before gravity probe b

 

[atyy]

I think that book was written in something like 1979, so definitely before gravity probe b

Gravity Probe B had some problems http://arxiv.org/abs/1106.1198

Among the tests of GR available in 1979 were:
http://en.wikipedia.org/wiki/Shapiro_delay
http://en.wikipedia.org/wiki/PSR_B1913+16

But there's interesting stuff about the correctness of GR from that period, eg. the solar oblateness story: http://www.astrosociety.org/pubs/mercury/9404/dicke.html

Again, a good review is http://relativity.livingreviews.org/Articles/lrr-2006-3/

 

[Dickfore]

I remember this as an abblication of the Mossbauer effect:

Pound-Rebka experiment

 

[PhilDSP]

"There have been a few somewhat shaky astronomical confirmations"
^^^^^^^^^^^

To be fair, precisely what he wrote is true (given that interpretations of the term "shaky" may vary). The initial solar eclipse measurements and comparison from experiment to theory were extremely crude. Much of the reason for that is that the refractive index around the sun is so highly variable that the deviation of the results for that reason swamped any other observables.

There was a study done by a french scientist, in the 60's I think, that applied the same methodology for determining Mercury's precession of the perihelion to other planets and satellites. None of the calculations were close to observed values. I make no judgement on the merits of that as I haven't read the book in which the results were published (and don't at the moment remember the author's name), however they were mentioned by the very eminent physicist Leon Brillouin.

 

[ParticleGrl]

To be fair, there has never been (or at least their hadn't when i last looked for such a things a few years back) a strong field test of GR. Only first-order, weak-field corrections.

 

[Dickfore]

Please predict black holes in the weak-field limit of GR.

 

[harrylin]

Please predict black holes in the weak-field limit of GR.

Please give evidence for GR's black holes (with singularities?) that cannot be called "shaky". :tongue2:
[STRIKE]Does that book say anything about black holes?[/STRIKE]

PS I looked it up at Amazon, and I see that he does write about them (but I can't access those pages)

Anyway, I see, a bit to my surprise, that it is a textbook complete with exercises and answers.
Near the end of the introduction he writes "What science should teach us is to doubt - to consider that many of the beliefs [..] may simply be wrong." - very true!
And he ends that section with: "if this book can contribute to teaching that lesson, it will have done its job".
That explains the sentence with which this thread started.

 

[Dickfore]

Please give evidence for GR's black holes (with singularities?) that cannot be called "shaky". :tongue2:

Ok, so first of all, let us define what was meant by 'ParticleGrl':

To be fair, there has never been (or at least their hadn't when i last looked for such a things a few years back) a strong field test of GR. Only first-order, weak-field corrections.

Look at the gravitational field of a point mass in Newtonian gravitation:
$$\mathbf{g}(\mathbf{x}) = -G \frac{m}{x^2} \, \hat{\mathbf{x}}$$
As $x \rightarrow 0$, $g \rightarrow \infty$, i.e. it is unbounded. At some point, what was called "weak" must fail to hold, because, how can infinity be weak (unless you compare it to another infinity)?

 

[Dale]

To be fair, there has never been (or at least their hadn't when i last looked for such a things a few years back) a strong field test of GR. Only first-order, weak-field corrections.

To be doubly fair, we still don't have a strong field test theory for GR. So even if we had everyday access to black holes we don't have a theoretical framework to test it against. All we have is the PPN.

 

[harrylin]

Ok, so first of all, let us define what was meant by 'ParticleGrl': [..]

Sorry, I remained at what was meant by robertjford80 and March.

 

[Dickfore]

Sorry, I remained at what was meant by robertjford80 and March.

My comment followed 'ParticleGrl' attempt at pedantry of what predictions of the theory of GR had been tested so far. I thought it was obvious.

 

[harrylin]

My comment followed 'ParticleGrl' attempt at pedantry of what predictions of the theory of GR had been tested so far. I thought it was obvious.

It wasn't clear, and I wonder if you missed the point... some textbooks write such "pedantry", emphasizing that GR could be found to be an approximation of a better theory - just as Newton's mechanics is an approximation of GR.

 

[Dickfore]

some textbooks write such "pedantry", emphasizing that GR could be found to be an approximation of a better theory

ORLY? What books would that be? Care to give references? Also, until such a "better theory" is proposed, it is pointless to discuss it. Do you know of any proposals of such theories?

 

[PeterDonis]

To be fair, there has never been (or at least their hadn't when i last looked for such a things a few years back) a strong field test of GR. Only first-order, weak-field corrections.

Binary pulsar systems qualify as "strong field" in at least some respects (certainly they require corrections beyond first order), and so far all observations of those have matched GR predictions. See the living reviews site here:

http://relativity.livingreviews.org/open?pubNo=lrr-2001-4&page=node23.html

 

[Whovian]

ORLY? What books would that be? Care to give references? Also, until such a "better theory" is proposed, it is pointless to discuss it. Do you know of any proposals of such theories?

Just a guess here: the point here might be Quantum Mechanics and General Relativity are incompatible, and so one might think both are approximations(?)

 

[Samshorn]

To be fair, there has never been (or at least their hadn't when i last looked for such a things a few years back) a strong field test of GR. Only first-order, weak-field corrections.

The tests aren't limited to "first-order", if by that you mean first order in m/r (in geometrical units). Of the three classical tests (or four, counting Shapiro), the redshift was basically a zeroth order test, since it must be satisfied by any metrical theory with the equivalence principle, and the light deflection and Shapiro are first order tests in m/r, but the perihelion advance is second-order in m/r. That's why it is by far the strongest test of general relativity.

 

[George Jones]

To be fair, there has never been (or at least their hadn't when i last looked for such a things a few years back) a strong field test of GR. Only first-order, weak-field corrections.

In a few years, we should be able to "image" (using microwaves) the black hole at the centre of our galaxy and stuff falling into the black hole,

https://www.cfa.harvard.edu/~loeb/sciam2.pdf.