Update: What is the difference between Lorentz Ether Theory, which spawned the Lorentz Transformation, and Special Relativity that uses the transformation as its basis as well?Saying that LET was ad hoc and bottom-up means that it was based on Michelson-Morley and other experiments. Saying that Einstein's SR was top-down means that he used principles deduced from those experiments, and not the experiments themselves. Lorentz described this difference as Einstein postulating what was previously proved.Mark Barton
PhD in Physics, researcher at University of GlasgowAuthor has 17.9K answers and 23.7M answer views 9yIt's a bit hard to say precisely, because LET was never an entirely finished project, but it was clearly converging on being identical to SR, and if you include Poincare's corrections then it was essentially the same for practical purposes. The main difference is that LET took an ad-hoc and bottom-up approach which ended up grudgingly backing into the relativity principle and the Lorentz transformation, whereas SR took a top-down approach that cheerfully assumed the relativity principle from the beginning, immediately derived the Lorentz transformation from it, and then read off a bunch of new physics that had to be true for all this to make sense.
Anyone insisting on LET today is probably holding out for some combination of the following ideas: (i) there is absolute space consisting of points with well-defined identities and well-defined and constant distances between them, (ii) there is a fact of the matter as to whether any object is travelling past the points that make up space, (iii) there is an absolute time in the sense that history is cleaves naturally into well-defined instants spanning all of space, and (iv) there are well-defined time intervals between the instants.
The trouble is that to the extent the relativity principle is true, all of the above four points are unfalsifiable by any experiment. It's analogous to claiming that ordinary 2D space has a One True X Coordinate, that all the points with the same One True X Coordinate are invisibly linked so as to form a sort of grain structure, and that likewise there's only One True Y Coordinate, and all the points with the same One True Y Coordinate are linked as well.
So the people insisting on LET then fall into two broad groups (with some overlap): (i) the ones who say, yes, it may be unfalsifable, but it's just complete nonsense philosophically for it to be any other way so SR has to be an illusion, and (ii) the ones who maintain that some marginal ancient result proves that absolute space and time do exist after all.
There are philosophers who argue that top-down is greatly superior to bottom-up. An empiricist might prefer bottom-up. I am not sure there is much practical difference.
In the top-down view, the Lorentz contraction is exactly what is needed for the relativity principle. An anti-positivist might prefer that. In the bottom-up view, it is what is needed for Michelson-Morley. A positivist would prefer that.
I quibble with Barton's last point. The discovery of the Cosmic microwave background was not "some marginal ancient result" and it does indeed define a frame for determining absolute time and whether an object is moving.
The radiation is remarkably uniform across the sky, very unlike the almost point-like structure of stars or clumps of stars in galaxies.[6] The radiation is isotropic to roughly one part in 25,000: the root mean square variations are just over 100 μK,[7] after subtracting a dipole anisotropy from the Doppler shift of the background radiation. The latter is caused by the peculiar velocity of the Sun relative to the comoving cosmic rest frame as it moves at 369.82 ± 0.11 km/s towards the constellation Crater near its boundary with the constellation Leo[8]So we can say that our Sun is moving with velocity 370 km/sec towards Crater.
Sometimes it is said that special relativity is based on there being no privileged frame. But that is clearly false, as the CMB forms a privileged frame, and it has no effect on the predictions of special relativity.
Back in the early 1900s, LET was known as Lorentz-Einstein Theory. It was superseded by the Poincare-Minkowski theory of a 4D spacetime with the non-euclidean geometry of the metric +dx2+dy2+dz2-dt2.
Ultimately, the problem of the preferred frame is one which requires standardization; it is not a problem of finding the natural occurrence of any such a frame.
ReplyDeleteWhen two masses interact, say by the direct contact, they exchange forces that are proportional to their masses. Starting from the knowledge of this law, there is no way to conclude that an object of unit mass exists and the only problem remaining is to find it. So, the only rational choice left is which particular object to adopt as the standard unit of mass. It's the *standard* unit, not the *natural* unit.
Since all objects interact, via various known (and even unknown) forces, there cannot be a principle which says that there must exist some objects that don't interact at all in any way, and therefore, extensions in between such objects must remain the same at all times, and therefore, these objects can provide a preferred frame for space measurements. If they were truly not to interact with our measurement detectors, we wouldn't record that they exist. If they do interact with our detectors, what is the reason to say that they don't interact in the same way among themselves? Just because it would solve a problem that's uncomfortable to us? The very idea says that some physical objects are more special in that they are selective in their interactions such that not all laws of physics apply to them. Accepting that idea is to throw all of physics out of the window.
Indeed, because of the above interaction principle, spatial changes (in the inhomogeneities) must occur internally even to any two CMB snapshots each of which gives the state of the physical universe at two different past instants.
But, yes, the proposal does have enormous merits from the viewpoint of standardization (even if none from the viewpoint of finding a naturally occurring preferred frame).
Just the way, even if the choice of the standard unit of mass should be made from such a material that it doesn't evaporate or adsorb too much, and so, it doesn't undergo any changes to its own mass property over time, similarly, the standard frame should be of such a configuration that even if does change over time, the change must be very, very slow as compared to all the other changes we wish to measure.
In reality, of course, not just electrons but even atoms from the exposed surfaces of any material are leaving the object and being adsorbed by it all the time. But the standard unit mass still is chosen to be stable enough. Similarly, the CMB configuration is stable enough.
Ultimately, it's not its near uniformity, but precisely the fact of its inhomogeneity --- and the fact that inhomogeneity changes very slowly in the past --- which makes CMB more suitable for adoption as the standard frame, as compared to any other physically existing configuration .
Best,
--Ajit