General relativity is rather solved in time symmetric way, like the least action principle condition in Einstein's field equations, what as in e.g. Wheeler-Feynman absorber theory requires symmetrically both retarded and advanced solutions.

So why seems there are only considered retarded gravitational waves?

Can we exclude being advanced wave for all observed events ( https://en.wikipedia.org/wiki/List_of_gravitational_wave_observations )? If not, should they use original chirp shapes, or maybe time reversed?

3 hours ago, Duda Jarek said:

So why seems there are only considered retarded gravitational waves?

Causality would be the primary reason.

But isn't general relativity solved in time symmetric way like the least action principle? How would you like to restrict causality to uni-directional in such 4D block universe view?

E.g. below black hole horizon time is switched with space, or even worse: Klein-botte-like wormhole ( https://en.wikipedia.org/wiki/Non-orientable_wormhole ) could in theory apply P or T symmetry to a rocket flying through it - how do you see causality there?

Solutions can exist but be considered unphysical. e.g. there are times you solve a quadratic and discard the negative results because it’s not possible or otherwise makes no sense - it violates a boundary condition of the problem, although it might not be an explicitly stated one.

You’re getting dangerously close to this being a reopening of a closed topic.

The equations of general relativity possess time-reversal symmetry, as do the general solutions of those equations, taken as sets of all solutions for each of the equations. But the individual solutions taken from a set of all solutions, perhaps specific solutions resulting from specifying auxiliary conditions, often do not share the symmetries of the equation, such as time-reversal symmetry. In other words, applying a symmetry transformation often transforms one solution to a different solution, not the same solution.

It is my understanding that for a wave equation, the initial conditions contain all the information required for the retarded wave solution, rendering the advanced wave solution redundant.

Also, the retarded wave solution is an expanding wave solution in the future direction while the advanced wave solution is an expanding wave solution in the past direction. But an expanding wave solution in the past direction is a contracting wave solution in the future direction and would require carefully arranged emitters to produce such a solution, noting that reality possesses an arrow of time which enforces the notion of causality and the second law of thermodynamics. This does not violate general relativity as covariance is maintained in spite of the asymmetry implied by the arrow of time. I suppose the simplest way to explain this is that the existence of symmetry in the laws of physics does not imply the absence of asymmetry in the laws of physics.

swansont, the big question is which boundary conditions?

For Euler-Lagrange we use values and derivative to start evolution in one direction - usually toward future, but can also use it toward past by t -> -t.

For the least action principle we use boundary conditions symmetrically from both directions, e.g. Big Bang and Big Crunch.

I haven't seen general relativity solved with Euler-Lagrange: "unrolling" 3D spacetime (?) Instead, they search for static 4D shape of spacetime satisfying least action equations - Einstein's block Universe, eternalism philosophy of time: https://en.wikipedia.org/wiki/Eternalism_(philosophy_of_time)

And in such least action principle view, there are allowed closed timelike curves, but the found static 4D solution (we travel through) has already resolved all potential time paradoxes: https://en.wikipedia.org/wiki/Novikov_self-consistency_principle

KJW, modern physics is believed to be CPT symmetric, so asymmetries like entropy gradient need to be properties of specific solution we live in, like throwing a rock to lake (symmetric in equations).

E.g. LIGO rather mostly sees retarded waves, as "collisions from our past" seem more like in our solution, but maybe it also contains reversed (?) - e.g. assuming Big Crunch and evolving backward, getting black hole collisions this way for us would be "collisions from our future" with our advanced waves - maybe LIGO could see, probably for time-reversed chirp shapes (?)

Edited December 3Dec 3 by Duda Jarek

17 minutes ago, Duda Jarek said:

KJW, modern physics is believed to be CPT symmetric

But what does that actually mean? I've already said that symmetric equations do not necessarily lead to symmetric solutions. Symmetry in physics is actually quite subtle.

17 minutes ago, Duda Jarek said:

so asymmetries like entropy gradient need to be properties of specific solution we live in, like throwing a rock to lake (symmetric in equations).

No, there is a fundamental asymmetry in physics: Transition probabilities do not possess time-reversal symmetry. Suppose one has a microscopically reversible reaction: X transitions to Y with a given probability over a given interval of time, and Y transitions to X with the same probability over the same interval of time. Here, reversibility refers to either species transitioning to the other species. But both transitions are in the future direction. If one has more X than Y (or more Y than X) at a particular time and look at the composition a given interval of time later, it would be seen than the later composition obeys the statistics of the transition probability. But if one starts with the composition at the later time and considers the statistics of the transition to the earlier time, one would see that the earlier composition does not obey the statistics of the transition probability or any transition probability (seen by considering the time-reversed statistics at various times).

And the reason why transition probabilities do not possess time-reversal symmetry is because probability is a positive number.

But note that the rate equation of the reaction is covariant with respect to time reversal. That is because the rate constant, unlike the transition probability, has a negative value for the time-reversed reaction.

Edited December 3Dec 3 by KJW

KJW, CPT symmetry is in equations - that "running movie backward" e.g. evolving back from Big Crunch (+CP), the equations governing physics should be the same - solving physics with the least action principle for GR or Feynman ensembles for QFT - assuming boundary conditions in Big Bang and Big Crunch.

Therefore, asymmetries cannot be in equations, need to be only in solution - like low entropy of Big Bang/Big Crunch - everything is localized, so entropy is low.

Or emission asymmetry: that circulating electron now loses energy (gains in CPT) - seems because there is now more absorbers than emitters, but it can be reversed e.g. in tabletop particle accelerators using lasers ( https://newscenter.lbl.gov/2016/02/01/2-stage-laser-plasma-accelerator/ ), or applying CPT symmetry ... or maybe also in far future (?)

Or there is now tendency to form black holes, which can collide - observed by LIGO ... but before Big Crunch shouldn't also this tendency be reversed? Could LIGO observe such advanced waves?

Edited December 3Dec 3 by Duda Jarek

46 minutes ago, Duda Jarek said:

Could LIGO observe such advanced waves?

I think these would look just like ordinary retarded waves to LIGO - what is measured there are essentially just tidal effects. But crucially, causality would be violated - we’d see the wave front arriving here before the source event actually happens in its own rest frame (ie all other signals would arrive much later). Needless to say this has not been observed.

Sure, I also suspect LIGO should see retarded and advanced waves as looking similar distortions of spacetime.

However, there might be ways to distinguish observed events like collisions, e.g.:

• separate observation e.g. visual - if based on retarded waves, their lack could indicate advanced gravitational waves,

• chirp shape - backward evolving from Big Crunch, shouldn't its collisions have time-reversed chirp shapes for us?

Maybe it is worth to include such mirrored chirp shapes into their database they use for search?

Edited December 3Dec 3 by Duda Jarek

ps. diagram for such hypothetical advanced gravitational wave assuming Big Crunch:

Diagra

11 hours ago, Duda Jarek said:

swansont, the big question is which boundary conditions?

You mentioned gravitational waves. We see them from black hole mergers. For a time-reversed solution, the boundary condition would include where the waves came from, and how they could all arrive at the site of the merger at the same time.

The wave itself exhibits time-reversal symmetry, but the entire scenario does not, an issue you continue to ignore.

Imagine there will be Big Crunch, so we can view the history of our eon as solution of the least action (GR)/Feynman ensemble(QFT) for boundary conditions in Big Bang and Big Crunch.

In such solution evolve backward from Big Crunch - shouldn't there also be formed black holes?

If they merge, for us they would generate advanced waves - should LIGO see them?

1 hour ago, Duda Jarek said:

Imagine there will be Big Crunch, so we can view the history of our eon as solution of the least action (GR)/Feynman ensemble(QFT) for boundary conditions in Big Bang and Big Crunch.

In such solution evolve backward from Big Crunch - shouldn't there also be formed black holes?

If they merge, for us they would generate advanced waves - should LIGO see them?

I have no idea why you think the big crunch would be a time-reversal of big bang.

Least action only works for conservative forces. Once you get away from that, you can’t apply it.

So how would you like to solve general relativity for boundary conditions in Big Bang and Big Crunch?

The standard universal answer seems: the least action principle, or Feynman ensemble for QFT.

If so, what would be the difference between "hot soup" just after Big Bang and "hot soup" just before Big Crunch? - to make their surrounding very different: that one has tendency to form black holes, but the second doesn't?

25 minutes ago, Duda Jarek said:

So how would you like to solve general relativity for boundary conditions in Big Bang and Big Crunch?

The standard universal answer seems: the least action principle, or Feynman ensemble for QFT.

If so, what would be the difference between "hot soup" just after Big Bang and "hot soup" just before Big Crunch? - to make their surrounding very different: that one has tendency to form black holes, but the second doesn't?

I don’t know what the BC of the big crunch would be, if that happens to be the fate of the universe, and neither do you.

Least action ceases to apply once there’s an energy that you can’t write as potential energy, so I don’t see how that’s the “standard” answer. The issues of cosmology involve more than general relativity, so “solving GR” is insufficient.

11 hours ago, Duda Jarek said:

So how would you like to solve general relativity for boundary conditions in Big Bang and Big Crunch?

These are scenarios where quantum effects become important, so just GR alone isn’t going to be enough here.

Also, the Big Crunch is hypothetical.

On 12/3/2025 at 3:45 PM, Duda Jarek said:

KJW, CPT symmetry is in equations - that "running movie backward" e.g. evolving back from Big Crunch (+CP), the equations governing physics should be the same - solving physics with the least action principle for GR or Feynman ensembles for QFT - assuming boundary conditions in Big Bang and Big Crunch.

Therefore, asymmetries cannot be in equations, need to be only in solution - like low entropy of Big Bang/Big Crunch - everything is localized, so entropy is low.

Although the equations are symmetric, that "running movie backward" is a solution that is not symmetric. It illustrates the point that a symmetric equation need not have a symmetric solution, although the symmetry-transformation of a solution is a solution. However, auxiliary conditions that select a particular solution from the set of all solutions need not possess the same symmetries as the equations, and indeed must not possess the symmetries of the equation that are not possessed by the particular solution.

You didn't respond to what I said about transition probabilities not being symmetric to time reversal. It should also be noted that covariance is not the same as invariance. For example, the rate equation mentioned above is covariant. But because the rate constant changes sign, along with rate of change, under time reversal, the arrow of time is maintained as the direction for which the rate constant is positive.

Why are you often mentioning the Big Crunch? The current evidence is that the expansion of the universe is accelerating and that the Big Crunch is unlikely.

15 minutes ago, KJW said:

Although the equations are symmetric, that "running movie backward" is a solution that is not symmetric. It illustrates the point that a symmetric equation need not have a symmetric solution, although the symmetry-transformation of a solution is a solution. However, auxiliary conditions that select a particular solution from the set of all solutions need not possess the same symmetries as the equations, and indeed must not possess the symmetries of the equation that are not possessed by the particular solution.

Another way of looking at this is that there’s a limit to how long you can run the movie forward or backward before something happens that is not accounted for in the equations, and you no longer have the required symmetry.

Between the big bang and big crunch is a whole lot of stuff not covered by GR.

8 hours ago, KJW said:

Although the equations are symmetric, that "running movie backward" is a solution that is not symmetric.

Sure, e.g. throwing a rock to a lake symmetric in equations, there appear asymmetries of solutions ... for physics we know many such asymmetries , like entropy gradient, emission asymmetry (e.g. circulating electron losing energy), tendency for black hole formation from direction of Big Bang.

But solving physics by the least action(GR)/Feynman ensemble(QFT) using e.g. boundary conditions in Big Bang and Big Crunch, they seem very similar just hot soups - symmetric ways of solving from symmetric boundary conditions, shouldn't the solution be also symmetric? Reversing e.g. mentioned 3 asymmetries near Big Crunch, like in diagram above ( https://scienceforums.net/topic/140178-why-we-observe-only-retarded-gravitational-waves-not-advanced/#findComment-1304506 )?

8 hours ago, KJW said:

Why are you often mentioning the Big Crunch? The current evidence is that the expansion of the universe is accelerating and that the Big Crunch is unlikely.

Both DESI observations and revisiting of supernova data (e.g. Sabine's https://www.youtube.com/watch?v=2VpP-qXuJMc ) suggest slowing down acceleration ... but sure Big Crunch is only hypothesis now, just convening to get intuitions - valuable also if it is not true.

8 hours ago, swansont said:

Another way of looking at this is that there’s a limit to how long you can run the movie forward or backward before something happens that is not accounted for in the equations, and you no longer have the required symmetry.

Between the big bang and big crunch is a whole lot of stuff not covered by GR.

Least action GR is deterministic, but sure there is also QFT - solved e.g. by <psi_f |U| psi_i> S-matrix, also using boundary conditions in both time directions, between them assuming Feynman ensemble - e.g. of histories of Universe between Big Bang and Big Crunch.

Returning to the topic, LIGO measures lengths - which are T/CPT invariant, so in theory should also see advances waves ... if only there are such events, scientists should test instead of assuming, but the big question is what to search for e.g. in LIGO data? Reversed chirps is just a first guess ...

In contrast, EM receivers are rather focused on absorption - of retarded waves. For advanced would need to be sensitive for stimulated emission - requiring excitation of sensor, what is rarely used now.

6 hours ago, Duda Jarek said:

Least action GR is deterministic, but sure there is also QFT - solved e.g. by <psi_f |U| psi_i> S-matrix, also using boundary conditions in both time directions, between them assuming Feynman ensemble - e.g. of histories of Universe between Big Bang and Big Crunch.

And thermodynamics, which you continually ignore, along with other areas of nuclear and atomic physics and mechanics.

Do you use QFT to explain galaxy/star/planet formation?

As I pointed out earlier, least action requires use of potential energy, which means the forces must be conservative. As soon as a dissipative force appears in a process, it doesn’t work. That limits the scope where you can apply it, and time symmetry.

Returning to the topic, LIGO measures lengths - which are T/CPT invariant, so in theory should also see advances waves ... if only there are such events, scientists should test instead of assuming, but the big question is what to search for e.g. in LIGO data? Reversed chirps is just a first guess ...

LIGO detects gravitational waves; it can’t distinguish between advanced and retarded. So we are testing, but that’s not the issue. The advanced waves have to exist in order to detect them, and the reasons they don’t exist are a separate problem and have nothing to do with LIGO.

On 12/7/2025 at 1:11 PM, swansont said:

LIGO detects gravitational waves; it can’t distinguish between advanced and retarded. So we are testing, but that’s not the issue. The advanced waves have to exist in order to detect them, and the reasons they don’t exist are a separate problem and have nothing to do with LIGO.

You have written that we are testing - if retarded vs advanced, how is it done?

For GR there is used the least action principle.

Gathered some more arguments.
As the main source of gravitational wave events is just orbiting of e.g. two black holes, and evolving toward minus time orbiting remains orbiting, so using Euler-Lagrange toward minus time (t -> -t), or the least action principle, there should be generated similar waves - for us being advanced of similar chirp shape as retarded. LIGO just measures lengths - invariant to time symmetry, so should see both retarded and advanced waves.

Therefore, maybe some of current ~300 events ( https://catalog.cardiffgravity.org/ ) could turn out advanced? Some arguments:

- ultimate confirmation should be certain lack of (retarded) EM counterpart when required (e.g. neutron star merger), still only 1 per ~300 observed, leaving advanced wave possibility (?),

- some events are believed to happen too early, like 66 + 85 -> 142 merger starting in 50-120 black hole Mass Gap, e.g. https://www.symmetrymagazine.org/ar...st-scale-could-explain-impossible-black-holes - advanced would have more time,

- pulsar arrays show vibrations of the Universe requiring more than expected orbiting supermassive black holes - https://theconversation.com/to-map-...uilt-a-detector-the-size-of-the-galaxy-244157 - advanced could add them,

- the largest observed luminosity distance is ~27Gly: twice the age of the Universe - maybe it is worth to consider advanced?

6 hours ago, Duda Jarek said:

- the largest observed luminosity distance is ~27Gly: twice the age of the Universe - maybe it is worth to consider advanced?

LY is not a unit of age. The size of the observable universe being larger than 15 LY is not an issue of advanced waves.

Sure, while being older than the age of the Universe would be a nice argument for advanced, luminosity distance is only one of 3 types of distances they are using, being only suggestion ... it is the last argument I have used, the remaining 3 are much stronger.

The clearest would be certain lack of (retarded) EM counterpart when required - if excluding retarded, wouldn't we have to consider it was advanced? Such cases might be a matter of months now. What other ways could we use to distinguish retarded from advanced?

Observing "too early to happen events" like with black hole mass gap seems such a way - could e.g. this 66 + 85 -> 142 merger be advanced wave?

Then there are Timing Pulsar Arrays seeing vibrations of the Universe - claiming they require orbiting SMBH ... if retarded seem insufficient, couldn't they be also advanced?

Any other arguments for/against advanced GW?

On 12/2/2025 at 3:32 AM, Duda Jarek said:

So why seems there are only considered retarded gravitational waves?

As with other waves, so take the analogy of EM for example: observation is of light moving away from a star, but not photons out of nowhere converging onto a star, splitting helium into hydrogen in the core. Both are valid solutions but....well, thermodynamics. Entropic arrow.