Hypothesis, Theory, and Law

[Progenitor A]

A scientific hypothesis is a theory that does not have enough evidence for anything approaching validity
A scientific theory is an hypothesis raised to a slightly higher level because it has been validated (to a small extent)
A scientific Law has had so much validation that it appears
self-evidently correct.

But wait!

There are NO scientific Laws that have withstood the test of time. One of the greatest scientific intellectual accomplishments, Newton's Laws of Motion have been shown to have only a limited range of validity; the theory that 'replaced' (NASA still uses Newton's Laws) Newton, Einstein's has only a limited range of validity, and 'the greatest accomplishment of the 20th. C, QM still raises passions of dissension, has no answer to Gravity and is not yet understood in any meaningful sense.

So perhaps the dogmatists on this board that insist that various scientific rules are invariant and correct should think again

All scientific theories and laws boil down to hypotheses; some are more reliable than others, that's all

So when you state that space is expanding (and not galaxies rushing apart) are you sure that you are not simply reciting a mantra, a dogma; when you state that an electron is in two (or more)places at once, you cannot know that - you are definitely repeating a dogma

Only one thing is sure about our present state of scientific knowledge; our theories, hypotheses, laws, are almost certainly wrong.

That is self-evident.

Why?
Not just historical evidence, but simply because there is no theory that can account for the BB.
ALL scientific theories collapse in ignominy at t=0; when we have a theory that accounts for that moment, and that theory can be validated, then we will have a complete , true, scientific body of Laws.

But of course, the events at t=0 cannot in principle, be tested

Therefore all theories and laws are hypotheses, and eventually will be cast aside for newer, more fashionable models

[speakertoanimals]

A scientific hypothesis is a theory that does not have enough evidence for anything approaching validity
A scientific theory is an hypothesis raised to a slightly higher level because it has been validated (to a small extent)
A scientific Law has had so much validation that it appears
self-evidently correct.

What is important is not trying to play word games, and attack anything that is any way hypothetical..............I also disagree with your categories, in that the point about a hypothesis is that it is still waiting to be tested, and usually best makes sense in the context of competing hypotheses.

A theory is much more than a bare hypothesis, tested or not. It usually explains something as well. So, whether or not magnetic fields could be produced by currents, could be viewed historically as a hypothesis, just requiring a simple experiment to demonstrate that was the case. Then we have various laws (NOT self-evident) that relate current to magnetic field (Biot-Savart law, faradays law of induction and so on), and finally Maxwells em theory, that explains those laws, and predicts a lot else (such as light as em radiation, and the speed of light).

All scientific theories and laws boil down to hypotheses; some are more reliable than others, that's all

Wrong. Theories generate hyptheses, but they aren't the SAME. SO, we might have a hypothesis (All A's are C's), which might or might not be true. If true, a theory would then explain WHY all A's are C's, and might also predict that all B's are also C's, or only some -- further hypotheses to be tested.

But a hypothesis, if you like, is a fairly straightforward statement, whereas a theory is much more, it explains the WHY behind various hypotheses, as well as (usually) presenting new ones to be tested.

All science is updates therfore all theories are provisional is just the usual nonsense. THe point being that the evidence behind a particular theory is still there -- Newtons laws of gravitationa still hold approximately in various circumstances, just that relativity shows how things need to be modified to cope with more extreme situations. So Newtons laws are not WRONG, in the sense that they are a useful approximation. What is WRONG are things such as Newtons assumption that time is absolute.

But given that time is not absolute, that isn't going to change, we won't suddenly discover that that is wrong as well, that time is going to be absolute again. What MAY change is the exact relation between time, space, motion and gravity. So, theories get refined, but some hypotheses (is time absolute or not) are a one-shot thing, once we have found the answer we are stuck with it.

In short, facts don't change, but their interpretation in the light of new facts, or new knowledge, may.

None that have stood the test of time? The second law of thermodynamics still going strong, and that is one that most physicists would be loathe to tinker with.....................

I think you misunderstand what physics IS, it isn't some rag-bag of laws, hypotheses and theories, where some are thrown out in favour of new ones. Just because a theory cannot explain EVERYTHING (like the Big Bang), doesn't make it useless, or wrong, in the sense that some theories are only meant to be correct and useful in some regimes.

[robinpike]

I don't think laws and theories are necessarily linked in that way.

Laws relate to measurements, they generalize about what has happened, from which we can predict about what we expect to happen in similar situations. For example, the ability of the ancients to predict eclipses had nothing to do with whether they knew what an eclipse is or not.

Theories tend to start off as explaining why something happens. For example, we explain eclipses as occurring when the moon passes in front of our line of sight to the sun.

Both a law or a theory can be tested by checking against what they predict. For example, do eclipses always and only occur on the dates when the 'law of eclipses predicts'. And do eclipses always and only occur when the moon passes between the earth and the sun.

[speakertoanimals]

I agree that hypothesis, theeory, and law are often misused, and the commonest misunderstanding is lay people that state 'evolution is JUST a theory', when they intend to imply that we aren't THAT sure, else it would be a law.

As regards laws, one good example is Keplers laws of planetary motion -- it says something about how planets move, but doesn't explain why they move in accordance with these laws, rather than some other possible law. Newtons theory of gravitation is much simpler (there is a force, varies according to masses and distances like this), and then starting from that, you can derive Keplers laws.

Another good one are the laws of conservation of momentum and energy and angular momentum. The laws are that they are conserved, but the law doesn't say WHY they are conserved, or tell you how to write down the energy of an unfamiliar system. Whereas theoretically (ie in terms of theoretical physics), the WHY for the conservation is seen as a consequence of the uniformity of physics under translation (doesn't matter where you put the origin, hence translation in time gives energy conservation, translation in space gives conservation of momentum), or under rotation (doesn't matter where you assign the axes positions to be, gives conservation of angular momentum). This is a slightly odd one, in that it boils down to maths, but it isn't itself a physical theory, in that it doesn't tell you WHAT the laws of physics are that are invariant under rotation/translation, just that laws that have that symmetry will lead to conserved quantities.

The point about quantum theory (leaving aside the usual quip about it not being understood), is not that it might not be replaced (actually, it's a general class of theories, not a particular single theory), but that even if it is, it probably won't be replaced by the commonsense sort of classical theory that some seem to want -- the results for tests of Bells inequality show that.

As is usual in science, we need to focus on what has been disproved (that the world is classical), rather than worrying about proof, which is the realm of maths, not physics.

[robinpike]

Nov 29, 2010 15:10:36 GMT 1 speakertoanimals said:

The point about quantum theory (leaving aside the usual quip about it not being understood), is not that it might not be replaced (actually, it's a general class of theories, not a particular single theory), but that even if it is, it probably won't be replaced by the commonsense sort of classical theory that some seem to want -- the results for tests of Bells inequality show that.

I suspect that the reason why quantum theory isn't understood, is that it is not a fundamental explanation, i.e. there is a layer beneath what we currently know. As to whether the new layer is a classical theory or not, should still be debatable, despite test results apparently supporting Bells inequality. The reason being that it is hard to absolutely prove or disprove the reason for something happening.

[nickcosmosonde]

The vocabulary of scientific laws, hypotheses and theories is not a scientific or logical taxonomy, it's a vague and centuries old conceptual scheme that, clearly, does not refer to "natural kinds" so that one term is correct, the other incorrect. At most they refer in more or less the manner that the OP has outlined.

StA makes some interesting responses, but they don't really address the issue. There is no such animal as a "bare hypothesis" for a start; nor is it possible to differentiate a "theory" from the nest of hypothetical explanations that it attempts to elaborate. One needs only a bare glimmering of awareness of the history of science to know that all theories are provisional. Nor is it possible, logically, to show how the axioms of one theory can be wrong from the corpus of another theory that does not share those axioms. It is not possible by any observation or logical implication to demonstrate that Newton was wrong to propose that time is absolute, for example. The claim that it is "given", or that this can never change, is quite evidently mistaken, for exactly the reasons robinpike outlines. There will always be "layers beneath" we don't know, the discovery of which will almost certainly entail a complete overhaul of any theoriy's axioms, as well as a reinterpretation of all those previous observations that it was once held to explain.

The OP's argument is perfectly correct, therefore. Unless StA knows of some means to logically demonstrate that one theory is closer to the truth than another? Now that would be a fascinating discovery.

[speakertoanimals]

Nor is it possible, logically, to show how the axioms of one theory can be wrong from the corpus of another theory that does not share those axioms.

Eh? What is that supposed to mean?

Another theory -- well, I can't show that classical gravity is wrong using the axioms of quantum theory, but I can still show that the two are incompatible, and that we MUST have a quantum theory of gravity.

Wrong, provisional, yes, but in what way? I would bet a considerable sum of money that given tests of Bells inequalities, an underlying classical theory IS impossible (using classical in the usual sense of local and deterministic and all that). I don't know why some people keep flogging this one, as if there were a real chance that all the weirdness of quantum theory could somehow be done away with --it ain't gonna happen, and Bells theorem tests say that.

As regards conventional theoretical physics, the current advances are not so much quantum versus non-quantum (or alternatives to quantum), but possibilities for the ultimate structure of either matter (it's all quantum strings, rather than quantum point-particles), or the ultimate structure of spacetime (it's discrete at the lowest level, say the loop quantum gravity guys).

It is not possible by any observation or logical implication to demonstrate that Newton was wrong to propose that time is absolute, for example.

Uh, what about experimental tests of special relativity, which seem to show quite clearly that time is not absolute? you are not really disagreeing with those are you? Ditto that space is not absolute. I wasn't saying that it wasn't a sensible suggestion at the time.

Unless StA knows of some means to logically demonstrate that one theory is closer to the truth than another? Now that would be a fascinating discovery.

Why logic?You seem to be obsessed by maths! What's wrong with -- it fits the data, and encompasses a wider range of phenomena than previous theories. Hence by any sensible definition of closeness, it is closer to the truth than the theory it replaced. Relativity is closer to the truth than Newtonian gravity, just based on the simple fact that Newtonian gravity gets more stuff wrong than relativity.

Think of it like this -- at one range limited range of scales/speeds etc, Newtonian gravity works.

At a larger range of scales/speeds/masses, relativity works and works better than Newtonian gravity, AND explains why Newtonian gravity exists as a limit iof relativity.

Relativity may still have to be modified, but inlooking for that modificiation, you'd better have a theory that reproduces relativity in the limitm, rather than Newtonian gravity. Hence relativity is CLOSER to the truth than Newtonian gravity. Ditto axioms behind relativity are closer/better than those behind Newtonian gravity.

I fail to see why anyone with any sense would claim otherwise...........

[nickcosmosonde]

Nor is it possible, logically, to show how the axioms of one theory can be wrong from the corpus of another theory that does not share those axioms.

Eh? What is that supposed to mean?

It's supposed to mean exactly what it says. It's a quite straightforward truth of logic, accepted by everyone since at least Riemann's non-Euclidean geometry.

Another theory -- well, I can't show that classical gravity is wrong using the axioms of quantum theory, but I can still show that the two are incompatible, and that we MUST have a quantum theory of gravity.

Exactly. Apart from the last assertion. That's just a dogmatic guess.

Wrong, provisional, yes, but in what way? I would bet a considerable sum of money that given tests of Bells inequalities, an underlying classical theory IS impossible (using classical in the usual sense of local and deterministic and all that).

It's entirely irrelevant how much money you'd bet. I would bet an equal amount the opposite way - so what?

I don't know why some people keep flogging this one, as if there were a real chance that all the weirdness of quantum theory could somehow be done away with --it ain't gonna happen, and Bells theorem tests say that.

They don't say what you think they do. I don't know what you mean by "all the weirdness", but if you mean the quite absurd philosophical confusions and juvenile misinterpretations that a great many physicists and commentators have made about QM since the 1920s then, yes, there's no doubt whatever that this will be done away with.

As regards conventional theoretical physics, the current advances are not so much quantum versus non-quantum (or alternatives to quantum), but possibilities for the ultimate structure of either matter (it's all quantum strings, rather than quantum point-particles), or the ultimate structure of spacetime (it's discrete at the lowest level, say the loop quantum gravity guys).

It is not for you to dictate what is an advance by predetermining it can only be what you idiosyncratically take to be "quantum" rather than "non-quantum".

It is not possible by any observation or logical implication to demonstrate that Newton was wrong to propose that time is absolute, for example.

Uh, what about experimental tests of special relativity, which seem to show quite clearly that time is not absolute? you are not really disagreeing with those are you? Ditto that space is not absolute. I wasn't saying that it wasn't a sensible suggestion at the time.

The clue to your misaddressing my point is in your "seem". There has been no test of SR that contradicts Newton's axiom that space and time are absolute. On the contrary, all such "tests" have merely confirmed that all measurements of velocities are conformable to a Minkowski SpaceTime diagram. Newton would not have disagreed with this in the slightest - you have a fixed and invariant axis for time, and another for space. Einstein's equally unverifiable and ungrounded axiom that c is invariant doesn't contradict Newton's axioms, they're simply different. As was all the more evident with the reintroduction of absoluteness with GR - a point that Einstein was the first to acknowledge.

But I seem to recall we've had this argument before, on R4? As I pointed out to you then, there is no sense to be made of the Twins Paradox - the experimental fact of time dilation - unless time is absolute, in Newton's sense. Lorentz saw this, Einstein for a long time - because he was mistakenly a Machian positivist until GR - did not.



Unless StA knows of some means to logically demonstrate that one theory is closer to the truth than another? Now that would be a fascinating discovery.

Why logic?You seem to be obsessed by maths!

Why logic? That's the OP's point. I don't see what maths has to do with it. Maths is a branch of logic, not the other way around.

What's wrong with -- it fits the data, and encompasses a wider range of phenomena than previous theories.

What's wrong with this as a logical criteria of truth is that it's inadequate. That's what the OP is saying. The Copernican model of the Solar System fit the data and encompassed a wider range of phenomena than the Ptolemaic. Kepler's was more so. Einstein's was more so than Newton's mathematisation of this. They all contradict each other to some extent, and sometimes have radically differing foundations. The OP is pointing out that there is no reason to suppose this process has finished with Einstein.

Hence by any sensible definition of closeness, it is closer to the truth than the theory it replaced.

I repeat, logically this is an unsupported contention. How do you measure such "closeness" when you have no theory-independent God's Eye viewpoint from which to judge? Indeed, when all your observations are interpreted according to the theory you're presently employing?

Relativity is closer to the truth than Newtonian gravity, just based on the simple fact that Newtonian gravity gets more stuff wrong than relativity.

Impossible to give this assertion any meaningful measurement. It's an understandable intuition, but that's all it can be. When an alternative theory of gravitation comes along, predicting other observations that GR doesn't encompass, and probably throwing out the whole mathematical corpus of GR in favour of an entirely different philosophical foundation, what sense is there left in saying GR was "closer to the truth" than Newton? They were both wrong, that's all we'd be able to say. One enabled us to be wrong about a wider range of things, granted, but how is this closer to the "truth"?

Think of it like this -- at one range limited range of scales/speeds etc, Newtonian gravity works.

Whether something "works" is saying something radically different to saying it's "true".

Relativity may still have to be modified, but inlooking for that modificiation, you'd better have a theory that reproduces relativity in the limitm, rather than Newtonian gravity. Hence relativity is CLOSER to the truth than Newtonian gravity.

This is nothing more than a declaration of Faith. Not even Einstein went this far - he was fully aware that there was something seriously amiss with this "closer to the truth" idea. His own epistemology is considerably more sophisticated.

I fail to see why anyone with any sense would claim otherwise...........

That's why you'd be wise to reassess your dismissive attitude to philosophy. Read some. You'll discover that no serious philosopher of science holds your view any more, for the simple reason it's untenable.

[abacus9900]

This is nothing more than a declaration of Faith. Not even Einstein went this far - he was fully aware that there was something seriously amiss with this "closer to the truth" idea. His own epistemology is considerably more sophisticated.

Unfortunately, his epistemology lead nowhere because he could not accept the non-intuitive results of QM. Despite possessing a brilliant mind he was, nonetheless, a dinosaur. In order to get closer to the truth we must be prepared to abandon what we thought was the truth.

[nickcosmosonde]

Unfortunately, his epistemology lead nowhere

I know of no philosophy of science that has come anywhere close to seriously contradicting it. It's by far the best analysis of how the scientific process works.

because he could not accept the non-intuitive results of QM.

He didn't accept them because they were nonsensical, undemonstrated, and entirely unnecessary.

Despite possessing a brilliant mind he was, nonetheless, a dinosaur.

Ridiculous.

In order to get closer to the truth we must be prepared to abandon what we thought was the truth.

Of course. That wasn't the issue in the least for Einstein - neither was it for Bohr, Heisenberg, or any other supporter of the Copenhagen Interpretation. They would have laughed at any such notion as "getting closer to the truth."

[robinpike]

Nov 29, 2010 21:06:18 GMT 1 speakertoanimals said:

I would bet a considerable sum of money that given tests of Bells inequalities, an underlying classical theory IS impossible (using classical in the usual sense of local and deterministic and all that). I don't know why some people keep flogging this one, as if there were a real chance that all the weirdness of quantum theory could somehow be done away with --it ain't gonna happen, and Bells theorem tests say that.

But there could still be an underlying layer that is based on classical mechanics.

For example, there is a cork bobbing up and down on some choppy water. You are too far away to see the water and to touch the cork, but indirectly you can measure the position of the cork. After many measurements, you put together a theory that the cork not only has no shape, but it has no definite position, nor definite energy - its position and energy vary around a value, these things (at the small scale of the cork) cannot have absolute values. And yet the cork does have a fixed size and shape, and it never exists in two places at the same time, it never exists as a wave / particle, it is always a particle, for it is the water that has the wave properties, etc etc.

Nov 29, 2010 21:06:18 GMT 1 speakertoanimals said:

Uh, what about experimental tests of special relativity, which seem to show quite clearly that time is not absolute? you are not really disagreeing with those are you?

And again, there could still be an underlying layer, which is based on absolute time.

For example, someone notices that grass grows slower in winter than in summer, and they put forward a theory that (for grass at least), time runs slower when the temperature of grass moves closer to absolute zero. The theory of changing rates of time agrees perfectly well with the experimental evidence on the rates of growth of grass. And if they don't have the knowledge of atoms, and understand that chemical reactions run slower with lower temperatures, then why should they ever think that their theory is incomplete?

[abacus9900]

Nov 30, 2010 2:02:28 GMT 1 nickcosmosonde said:

Unfortunately, his epistemology lead nowhere

I know of no philosophy of science that has come anywhere close to seriously contradicting it. It's by far the best analysis of how the scientific process works.

If you are going to play the 'troll' you could at least try to be more convincing. Einstein stuck to the 'hidden variables' idea of QM to account for 'spooky action at a distance' which has been conclusively shown to be incorrect time and again. Do the research.

He didn't accept them because they were nonsensical, undemonstrated, and entirely unnecessary.

QM has been demonstrated repeatedly to umpteen decimal places.

In order to get closer to the truth we must be prepared to abandon what we thought was the truth.

Of course. That wasn't the issue in the least for Einstein - neither was it for Bohr, Heisenberg, or any other supporter of the Copenhagen Interpretation. They would have laughed at any such notion as "getting closer to the truth."

Then why 'do' science if not to reveal nature? Now you're really being irrational.

[abacus9900]

But there could still be an underlying layer that is based on classical mechanics.

I'm afraid not because Bell showed that statistically, the notion of locality is a dead duck and so classical mechanics cannot possibility be used to underpin reality.

[speakertoanimals]

Nov 29, 2010 21:44:08 GMT 1 nickcosmosonde said:

It's supposed to mean exactly what it says. It's a quite straightforward truth of logic, accepted by everyone since at least Riemann's non-Euclidean geometry.

Ah, so you are talking maths, do be clear! Except maths is not theoretical physics. So, we have Euclidean geometry, and non-Euclidean geometry. Except things aren't that simple in physics, because we can also consider a space which is flat, but with non-trivial toplogy. No physicists would try to use maths to prove that the world must be non-Euclidean, that would be totally stupid, since mathematically, both are sound. But based on measurements, we could quite simply decide what the geometry of spacetime actually is.

Dearie me, if all we had to do was do some maths, theoretical physics would be easy, we wouldn't need experiment, we'd just decide what possible physical theories were mathematically consistent, and that would be that.

Ditto absolute time -- we've don't the measurements, and it isn't. This isn't a logical proof, because of course an absolute time is mathematically consistent in and of itself, it just isn't consistent with our actual MEASURMENTS.

Exactly. Apart from the last assertion. That's just a dogmatic guess.

Nope, far from it. Experiment has shown (Bells theorem tests for example) that some parts of nature are quantum (or at least not classical). Best theory of gravity is classical (relativity). Why not leave it at that? Because it is quite simple to show that quantum theories for some interactions, plus classical ones for gravity, are fundamentally incompatible. It's not a GUESS, that at least is a simple mathematical statement. Nor is it some fashion statement, that some of our most successful theories are quantum theories, so we ought to make gravity quantum as well.

Something has to give -- either gravity has to be quantum as well, or quantum theory will have to be modified to include gravitational effects, but what we certainly CANNOT stick with is quantum theories plus classical gravity, the two are fundamentally incompatible and lead to inconistencies and contradictions.

xxx.lanl.gov/PS_cache/gr-qc/pdf/0108/0108040v1.pdf

So, it might be quantum theory of everything, or it might be quantumish theory of everything (ie quantum theory modified in some way to include gravity, but it won't be a classical theory, and it can't be, given experimental results.)

They don't say what you think they do. I don't know what you mean by "all the weirdness", but if you mean the quite absurd philosophical confusions and juvenile misinterpretations that a great many physicists and commentators have made about QM since the 1920s then, yes, there's no doubt whatever that this will be done away with.

Bugger the philosophy, I was talking about basic physical measurements, which show that the universe just is weird. This is what tests of Bells theorem show, that we cannot keep all our classical concepts, that something has to give and the world just is NOT classical.

It is not for you to dictate what is an advance by predetermining it can only be what you idiosyncratically take to be "quantum" rather than "non-quantum".

O, go do some reading abhout physics, before you start making such daft statements! I was under the impression you knew a little physics, but that seems to have been a mistake..................

It's entirely irrelevant how much money you'd bet. I would bet an equal amount the opposite way - so what?

Put your stake up then, cos you'd loose! Perhaps you could explain WHY you think current results still leave a classical loophole open?

Uh, what about experimental tests of special relativity, which seem to show quite clearly that time is not absolute? you are not really disagreeing with those are you? Ditto that space is not absolute. I wasn't saying that it wasn't a sensible suggestion at the time.

The clue to your misaddressing my point is in your "seem". There has been no test of SR that contradicts Newton's axiom that space and time are absolute. On the contrary, all such "tests" have merely confirmed that all measurements of velocities are conformable to a Minkowski SpaceTime diagram. Newton would not have disagreed with this in the slightest - you have a fixed and invariant axis for time, and another for space. Einstein's equally unverifiable and ungrounded axiom that c is invariant doesn't contradict Newton's axioms, they're simply different.

Velocities? you seem to think this is all that has been measured! The whole point about Minkowski diagrams is that there is MORE than one, that you can take any observer (different time axis, different spatial axis), and get the same physics.

As regards Newton, he said that all inertial frames are equivalent. Which is not quite the same as space being absolute (no preferred rest frame), although distances are. But with time absolute, you get the Galilean transformation between frames, which is different to the Lorentz transform of relativity. Hence a simple experimental task -- either observations accord with Galilean relativity, or they do not, and observations show the latter. Since they were derived based on the concept of absolute time, then that goes out the window.

As regards c, we have simple tests with moving sources, which show that it is invariant. So I have no idea why you claim this is unverifiable and ungrounded.................And it is certainly incompatible with Galilean relativity, hence the axioms that Newton was using.

As I pointed out to you then, there is no sense to be made of the Twins Paradox - the experimental fact of time dilation - unless time is absolute, in Newton's sense.

Except the Twins paradox isn't the experimental fact of time dilation, but the experimental fact, if you like, that whereas velocity is relative, acceleration is not. This is what separates the twins, and allows a different elapsed time.

[robinpike]

Nov 30, 2010 11:21:10 GMT 1 abacus9900 said:

But there could still be an underlying layer that is based on classical mechanics.

I'm afraid not because Bell showed that statistically, the notion of locality is a dead duck and so classical mechanics cannot possibly be used to underpin reality.

But that is the whole issue, it is very difficult to prove or disprove the reasons for the outcome of an experiment when it involves things at the atomic scale, such as the results of the EPR experiment.

I recently attended a lecture by Alain Aspect, who talked about his EPR experiment involving entangled pairs of polarised photons. The results agreed perfectly with the QM prediction and not the classical prediction.

So why my concern? Because the experiment was also repeated where the entangled photons were let fly for 30 kilometers before their polarisation was measured, and exactly the same results were found - agreement with QM and disagreement with classical theory.

So you might say, that is even further proof that QM is the right explanation - but to me it is the opposite. I could understand if at such a distance the results failed to agree with QM, but since the distance had no impact on the results, it suggests to me that distance is a red herring and therefore interaction over distance is also a red herring.

So it leads me to consider this: although it seems impossible, is there any way that a classical explanation could give the found results?

That I don't know. But what I would do, is try to break the conclusion that QM is correct, by performing the opposite experiment - i.e. use two matching polarised photons that are not entangled, and then repeat that same experiment by Alain Aspect and see if the results agree with the classical prediction. If they don't agree with the classical prediction (but still agree with QM), then that would make the 'proof' that QM is the explanation unsound.

So a question: does anybody know if Alain Aspect did indeed perform the opposite confirmation of his EPR experiment? (For example, by extending the experiment that after measuring the polarisation of one entangled photon and collapsing the wave function of both photons - and therefore dis-entangling them, then re-measuring that photon's polarisation and measuring the other photon's polarisation). And if so, what were the results?