Talk:Tests of general relativity

External links modified (January 2018)

Hello fellow Wikipedians,

I have just modified one external link on Tests of general relativity. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added {{dead link}} tag to http://pancake.uchicago.edu/~carroll/grbook/
• Added archive https://web.archive.org/web/20010711203839/http://www2.corepower.com:8080/~relfaq/experiments.html to http://www2.corepower.com:8080/~relfaq/experiments.html

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 19:49, 24 January 2018 (UTC)

Deflection of light from galaxies by a star

"such a ring formed by the deflection of light from distant galaxies has been observed for a nearby star.[28]"

The source quoted only refers to the deflection of the light from a background star by another star, there is no mention of distant galaxies or from a ring.

Rings from distant galaxies have been observed, but around nearby galaxies (not stars) so there seems to be a mixture of the two results and an inappropriate reference. — Preceding unsigned comment added by 78.194.166.238 (talk) 06:35, 29 March 2019 (UTC)

Edit request: Is "Gravitoelectric effects (Schwarzschild-like)" a General Relativity effect?

The first table in the article does not explicitly state that the "Gravitoelectric effects (Schwarzschild-like)" is a General Relativity effect. I am not a physicist but I suspect the two terms mean the same thing, given the similar sizes of the effects (43 arcseconds in a preceding paragraph, and 42.98 arcseconds in the table). If my suspicion is correct, please reword the table so that the words "General Relativity" appear in it. Otherwise the table is unclear to the lay reader.86.136.201.152 (talk) 17:31, 25 August 2019 (UTC)

Anomalous Mercury Perihelion Precession Rate from Classical Mechanics

A new book, "Air of Doubt" (ISBN: 979-8697917329) by Dr. Frank A Tinker, offers an argument for modifying Kepler's Third Law in the same spirit as Newton's modification adding planet masses. The proposal is to add a factor involving the orbital energy to improve the comparison of two planet's orbits. In doing so, he resolves the Mercury perihelion precession anomaly without resorting to General Relativity. As such, he shows that the Theory of General Relativity is not necessary in order to resolve the anomaly. Which means General Relativity can't really use the solution to show superiority over classical mechanics.

For completeness, it appears that reference and proposal should be included in this topic. Link: www.airofdoubt.com.

Yes, that is me. But I'm not about to edit this without discussion. AoDFT (talk) 01:11, 20 October 2020 (UTC)

Revert of edit, June 10, 2021

I reverted this edit by Ruhenheim (talk | contribs) because the grammar was sloppy, the formatting was sloppy, and it used British English, while the article is in American English. If the material is important enough to be inserted, it will need to be cleaned up first.—Anita5192 (talk) 14:42, 17 June 2021 (UTC)

I have removed the unnecessary inclusions and only kept the first part which corrects the physics.

"Under Newtonian physics, a two-body system consisting of a lone object orbiting a spherical mass like a star would trace out an ellipse in the star's frame with the star fixed at a focus." Will this work?--Ruhenheim (talk) 18:12, 19 June 2021 (UTC)

I inserted your edit, replete with citation, after cleaning it up for grammar and formatting, and corrected some other subtle inaccuracies in the existing paragraph. I omitted some of the references to stars and planets to make the treatment more general.—Anita5192 (talk) 19:08, 19 June 2021 (UTC)

Thanks--Ruhenheim (talk) 17:50, 22 June 2021 (UTC)

Precession of the Perihelion of Mercury

I have made a minor change to help explain the direction of the precession. In the reference it says that that the precession is in the direction of the plane of the orbit. Not knowing anything about celestial mechanics I assumed it was 90 degrees to the plane of the orbit, and this was only corrected when I read the original citation. I think it would be much clearer to also include this in the wikipedia article. Eparaqutam (talk) 00:59, 4 November 2023 (UTC)

A 2026 study...

This content was added A 2026 study of the binary pulsar PSR J1856–0039 measured a number of GR predicted effects. As shown in the plot, at the predicted masses for the two stars, these are all consistent with predictions of GR. by @LouScheffer, but to me this is not WP:encyclopedic knowledge. It does not tell us anything that the rest of the article does not already provide. The plot is not remarkable and unexplained so unlikely to be useful to readers. Johnjbarton (talk) 01:54, 19 September 2026 (UTC)

I agree this needs more explanation, and I have added some. I feel exactly the opposite about its usefulness to readers. The fact that these three regions intersect for a particular value of M1 and M2 shows visually how GR is confirmed, and also how GR is used to measure the two masses. I recall the first time I saw a plot of this type, and to me the fact that three different post-Kepler parameters all agree was powerful evidence. A single parameter can leave a lot of loopholes (remember the arguments over solar oblateness, or ambiguity of NS masses or inclinations of orbits?) but when three of them are right at the same time, that's more conclusive. My two cents, anyway. LouScheffer (talk) 02:59, 19 September 2026 (UTC)

Sorry, but your added caption only convinces me this is not useful. Is this type of plot discussed in any textbooks or reviews? Maybe we can make something clear from such sources. Johnjbarton (talk) 04:08, 19 September 2026 (UTC)

These types of plots are quite common. See, for example, On the unreasonable effectiveness of the post-Newtonian approximation in gravitational physics, or TESTS OF GENERAL RELATIVITY AND MODIFIED GRAVITY USING PULSAR TIMING, or Testing General Relativity with Pulsar Timing. With 2 PPN parameters, you can use assumed GR to find the masses of the stars. With 3 parameters, you can find the masses, plus see if GR is consistent with the results. LouScheffer (talk) 06:18, 19 September 2026 (UTC)

So those are the sources that should be summarized in an encyclopedia, with corresponding background explanations. These secondary sources would explain the nature and value of the diagram and justify a sentence and result from a recent primary. Johnjbarton (talk) 15:28, 19 September 2026 (UTC)

For a generic explantion of these plots and their value, see section 4.2 of the "TESTS OF" paper, which is "PPK Parameters for General Relativity and the Mass-Mass Diagrams":

[...] by direct observation of the PPK parameters one can deduce five independent curves on a mass-mass diagram using the equations (21). Because this test involves only two unknown variables, the mass of the pulsar m1 and the mass of its companion m2, but five independent equations, it is a very powerful test of the theory.

So perhaps include this explanation and diagram, if the diagram is avaiable in public form. The current illustration has 3 parameters, though the idea is the same. LouScheffer (talk) 15:38, 19 September 2026 (UTC)

Could not find a licensed version of the 5 curve plot, but used the explanation from the review paper with the plot from the recent paper. Further suggestions welcome. LouScheffer (talk) 23:50, 19 September 2026 (UTC)

Has open source diagrams and some are more than three parameters.

• Freire, Paulo C. C.; Wex, Norbert (July 22, 2024). "Gravity experiments with radio pulsars". Living Reviews in Relativity. 27 (1): 5. doi:10.1007/s41114-024-00051-y. ISSN 1433-8351.

I made some changes based on this paper, please review.

Over all the section is redundant by repeating itself and the significance of the various cited tests are unclear. Johnjbarton (talk) 03:56, 20 September 2026 (UTC)

It's OK to use a formula without citing someone else who plugged in the numbers

For the energy radiated by the Earth, see, for example, Gravitational wave#Binaries for exactly the same use of a formula + well known constants gives a result with citing someone who themselves plugged in the numbers. This is explicitly allowed by Wikipedia. LouScheffer (talk) 04:25, 22 September 2026 (UTC)

I suppose you are referring to Routine calculations do not count as original research, provided there is consensus that the results of the calculations are correct, and a meaningful reflection of the sources. and my removal of

• Gravitational radiation from a circular orbit is ${\frac {32}{5}}{\frac {G^{4}}{c^{5}}}{\frac {(m_{1}m_{2})^{2}(m_{1}+m_{2})}{r^{5}}}$. For Earth, this is about 200 watts, or 6x10⁹ joules/year. In contrast the Earth's kinetic energy is about ${\frac {GMm}{2r}}$, or about 3x10³³ joules, by classical mechanics.

Gravitational radiation formula is very far from a routine calculation and no sources are cited to verify if this is a meaningful reflection. More important to me, the calculation does not verify the content on many levels. To start there is no compelling reason to compare the gravitational radiation energy to Earth kinetic energy since we have no way to detect KE. In fact I think the solar system is not relevant here beyond a verifiable claim that GR tests based on GW can not be done using the Earth/Sun system. This is just an extra can of worms sitting around. Johnjbarton (talk) 16:25, 22 September 2026 (UTC)

See, for example, Two-body problem in general relativity#Gravitational radiation where even more complex formulas are state without apology. Surely if the formula can be presented, plugging in numbers is OK.

Since we disagree about this, I suggest getting more opinions on this point. This will be much easier if they can see and criticize (or praise) something they can see in detail. So I'd suggest leaving it it (it does not disrupt a casual reading of the article) and see what others think. LouScheffer (talk) 01:29, 23 September 2026 (UTC)

I posted the content of the footnote in my reply where it can easily be read. Please undo your revert. Mistakes in other articles do not prevent improving this one. Johnjbarton (talk) 01:54, 23 September 2026 (UTC)

I appreciate your concerns. Your preference is for terseness, where mine is for additional explanation, and both views are (in my opinion) well within the typical Wikipedia style. I believe my version is more useful to the reader, as it explains where the numbers originate, and this calculation is indeed routine for any article on GR. See also for example Gravitational waves#Binaries, or click on the link in the note itself, which in turn points the reader to the classic text of Landau and Lipshitz.

Far more people read the article page than its talk page, and hence I think we are much more likely to get additional opinions if the proposed change is visible there. An explanatory note by its nature is unobtrusive to the casual reader. I think it's a good compromise, and should remain until/if others express their opinions on its suitability. LouScheffer (talk) 03:47, 23 September 2026 (UTC)

I think computing the power of the gravitational radiation emitted by the Earth-Sun system is indeed a routine calculation and can be included without a source. I'm not sure about the comparison with the kinetic energy of Earth's orbit. What's the point? To give the reader an idea of how long it would take for the orbit to collapse due to gravitational waves? In this case it would be better to give the actual number, which will be very different as the power increases as the orbit decays. Tercer (talk) 06:21, 23 September 2026 (UTC)

At least it is reassuring to know that the solar system will be destroyed by something else long before gravitational radiation has an appreciable effect on it. JRSpriggs (talk) 11:39, 23 September 2026 (UTC)

Hi! The object was to show why this effect is almost impossibe to measure in the Solar System, even though it is there. But maybe dr/dt or dP/dt would be better. Plus see below about the meaning of routine calculation in the age of AI. LouScheffer (talk) 14:14, 23 September 2026 (UTC)

The footnote provides zero information about the measurement of GW in the Solar System. It gives no verifiable information on the detection limit for GW, so numerical values of the radiation are not relevant. It gives no reference to an analysis of the impracticality of measuring Earth's orbital decay. No editor supplied formula or skill with a calculator will solve this problem. Johnjbarton (talk) 15:58, 23 September 2026 (UTC)

I've changed the main text to attempt to show more directly why GR radiation effects are hard (practically impossible) to see in the solar system. Then the footnote should show (in my opinion of course) how these numbers were determined. Footnote still needs work but I think the main idea is helpful. LouScheffer (talk) 20:57, 23 September 2026 (UTC)

Here is another consideration about what it means to be a routine calculation and the assumptions behind it. I gave google Gemini the prompt: "What is the rate of change of the period of the Earth's orbit around the sun due to the emission of gravitational waves? Please provide the formula and numbers used. A circular orbit can be assumed." and got back the following astonishing (to me) response in a matter of a few seconds. It figured out what formulas to use, chained them together, and then plugged in numbers. It explained what it was doing and why. It made approximations and justified them. It even gave the per second and per year numbers without me asking. I was more than a little impressed. If at some point in the future, you could click on a result in Wikipedia, and get an explanation of this quality as to how it was derived, that could be a huge improvement to physics in Wikipedia. LouScheffer (talk) 14:25, 23 September 2026 (UTC)

Emphatically no. Anybody that wants - and trusts - an LLM explanation can query it themselves, there's no point including it here. Wikipedia is about summarising information published in reliable sources. LLMs are not reliable sources. Tercer (talk) 15:21, 23 September 2026 (UTC)

I get your point, but am not sure this will be continue to be true. For these types of routine physics calculations, especially when they show their work, LLMs are reasonably trustworthy. They will soon (if not already) become more trustworthy than a typical wikipedia editor. Provided the LLM can provide reliable sources for the equations and values used, and explain how and why the calculation is done, the reader/editor can verify the results and the LLM itself does not need to be trusted. Overall, I'm seeing fewer and fewer differences between an LLM explanation and a human explanation. LouScheffer (talk) 21:12, 23 September 2026 (UTC)

If the LLM provides reliable sources for the calculation, we put those in the article, not whatever the LLM has calculated. And we do need to check these sources, because LLMs often hallucinate them. Tercer (talk) 08:34, 24 September 2026 (UTC)

Let me give a practical example. In the Wikipedia page Gravitational wave#Binaries, the orbital mechanical energy is stated as 1.16x10³⁶ joules, and the rate of orbital shrinking is given as 10^-15 meters per day. Both numbers were inserted by a human editor without explanation. I believe the orbital energy is wrong, but the LLM calculation above is right, precisely because the LLM showed its work (and I checked the arithmetic). The shrinkage seems correct (LLM gets 9.56x10^-16, in agreement with the text). In this case I believe the LLM is MORE trustworthy than the human editor. LouScheffer (talk) 21:30, 23 September 2026 (UTC)

Neither calculations of human editors nor LLM are relevant to Wikipedia. That is exactly my point in objecting to the calculation. If the calculation is not in a source, it is not notable. If it is, cite it. That is all. Johnjbarton (talk) 00:40, 24 September 2026 (UTC)

Please explain the diagram in the binary pulsar section

Please explain the diagram in the binary pulsar section. What do the lines mean? What do the points labeled ω and P mean? Why are those points in different positions in the insert than in the larger diagram? JRSpriggs (talk) 12:12, 20 September 2026 (UTC)

Now I see that the dots are indicating differentiation with respect to time. They are not points being labeled. The labels apply to the dashed lines in the diagram. JRSpriggs (talk) 16:49, 21 September 2026 (UTC)

I agree that the diagram needs more explanation. Please see the previous topic. I think progress is being made. Johnjbarton (talk) 20:04, 21 September 2026 (UTC)

I fixed the math symbols in the caption quote to agree with the plot. LouScheffer (talk) 21:15, 21 September 2026 (UTC)

Finding and including citations

Quite often (in this and other articles) editors add citation needed for facts that are easily available in the literature. Of course there is nothing wrong with pure criticism, but modern technology makes it much easier to be constructive.

LLMs (no matter what you think of them) make this task much easier. For example, a simple prompt such as "What's a formal paper that shows a quadrapole moment is needed for radiation of gravitational waves" will yield several likely reliable sources. Then of course YOU NEED TO READ THE PAPER TO BE SURE IT SUPPORTS THE STATEMENT. Then IF it supports the statement, you can ask "Please format this reference for use in Wikipedia". This will handle the conversion into Wikipedia proprietary format (why they could not have used Bibtex for references, like they use LaTex for math, is beyond me.). And of course check the reference on the final Wikipedia page once it is included.

Overall these make it much, much easier to add a reference to a reliable source. Note that no trust of the LLM is necessary. LouScheffer (talk) 23:13, 24 September 2026 (UTC)

Sometimes editors add content without add sources. I generally revert these additions as not verifiable, but that may discourage editors. Thus sometimes I add {{cn}} in hopes that an editor who is familiar with the sources will be able to pick a good one quickly.

When I use AI to look for sources I ask for "Textbooks, reviews, or highly cited papers that verify the claim that a quadrupole moment is needed for radiation of gravitational waves". I used a canned script (called a "skill" in google-speak) for formatting references in to {{cite}} templates. Numbers, especially ISBN can be off, but I use the AI side bar in Chrome so it is easy to check the numbers. Johnjbarton (talk) 23:41, 24 September 2026 (UTC)

At least with Google Gemini, you can give it standing instructions to double check any ISBN or DOI before it includes it. At least for me, that has solved the bad DOI/ISBN problem. LouScheffer (talk) 00:23, 25 September 2026 (UTC)

Thanks I'll try that. Johnjbarton (talk) 16:02, 25 September 2026 (UTC)