Tom Van Flandern/Meta
Abstract. Lorentz contraction is not a change in the
physical length of rods or meter sticks. Rather, it is an illusion introduced in
special relativity by the lack of remote simultaneity. In Lorentzian relativity,
elysium (the light-carrying medium) is entrained and time is universal, so the
need for a Lorentz contraction vanishes.
According to Einstein's special relativity (SR), the motion of a body relative
to an observer produces time dilation – a slowing of the rate of passage of
time, and length contraction – a compression along the direction of motion. For
a moving extended body or an entire inertial frame in motion, SR also predicts a
time desynchronization effect, sometimes referred to as a lack of remote
simultaneity. The moving body or frame is unaware of any changes in itself or
its own time, but instead sees the same kind of effects happening to the
observer and all bodies in the original frame. The amount of these effects is
given by the Lorentz transformation equations.
A variety of independent experiments, 11 in all, have tested and verified that
the predicted "time dilation" effect occurs at least for atomic clocks, which do
slow their rate of ticking with motion. []
Strictly, it has not been verified that anything happens to time itself.
Likewise, the reciprocity of the effect from moving frame back to laboratory
frame and the time desynchronization effects have not been separately verified.
As for length contraction, its existence is inferred, but it has also never been
seen directly in any experiment.
We here consider the motivation for proposing length contraction and its
physical nature. In the end, we will conclude that length contraction does not
An important feature of SR is that time will appear desynchronized in any other
inertial frame with a motion relative to the observer's own inertial frame. []
Specifically, if the nearest clock in a moving frame with a train of clocks is
synchronized with a clock at rest at the observer, then receding objects in the
moving frame will be experiencing time in the observer's past, while approaching
objects in the moving frame will be experiencing time in the observer's future.
See Figure 1. The greater the distance, the greater is the time discrepancy. The
fact that the finite speed of light prevents the observer from viewing into the
past or future does not diminish the presumed reality of these time differences
Figure 1. In SR, time will
appear desynchronized with the present in any inertial frame with a motion
relative to an observer.
Consider any object (say, a rod) in the passing frame, oriented along the
direction of relative motion. Then according to SR, a snapshot of the rod taken
by the stationary observer will not show the length of the rod all at one
instant of "rod time", but will see one end at a different moment of
moving-frame time than the other end. If the rod had stripes of different colors
painted lengthwise from end-to-end and was rotating around its long axis, the
observer would see the rod's colored stripes apparently twisted – not because
the rod is twisted, but because different parts are seen at different rod-times
during the rotation, all at a single observer instant.
Effect on a moving
As this pertains to Lorentz contraction, consider the consequences of this
desynchronization of time in a relatively moving frame for the appearance of the
moving rod as seen by the fixed observer. Suppose the rod's leading end is
labeled A and its trailing end is labeled B. See Figure 2.
Figure 2. At any one instant, the observer will see a
moving, rotating rod at progressively later times along its length from A to B.
This makes the rod look shorter and its stripes appear twisted.
The rod's stripe is actually linear from one end to the other, with no wrapping.
However, because of desynchronization, the observer will see end A as it was at
some time in the moving frame (depending on location), and the observer will see
end B as it was at some later time. So time along the rod appears different from
place to place to the fixed observer. And that makes the stripe appear twisted
as the rod rotates in time. What the observer sees of the moving rigid rod in an
instant is the same as a video camera would see if it panned slowly along the
length of a non-moving-but-rotating rod from end A to end B over some finite
period of time.
A further consequence of this desynchronization is that end B will always be
seen as it was at a later time moment than is seen for end A. But because B is
moving in the same direction as A, a later moment will bring end B closer to
where end A was a bit earlier. In other words, the rod will always appear
contracted in length because the leading end is seen at an earlier time than the
trailing end. This applies whether the rod is approaching or receding from the
Pole and barn
With this thought in mind, we can now readily understand the famous "pole and
barn paradox" (Figure 3). Consider a 21-meter pole approaching a barn with doors
open at both ends, 7 meters apart. Let the pole's speed be 99% of the speed of
light so that the length contraction factor is 7. Then the observer in the
barn's frame sees the length-contracted pole as only 3 meters long, and argues
that both doors can be slammed shut while the pole is inside. But the pole
regards itself as at rest and sees the barn approaching at 99% of the speed of
light. The pole therefore thinks the barn is length-contracted to only 1 meter
long between doors, and therefore cannot possibly contain the pole. Both views
cannot be correct – hence, the paradox.
The resolution of the paradox comes from time desynchronization. The pole only
appears contracted because the moment its leading tip reaches the far door of
the barn is earlier than the moment its trailing tip passes by the near door.
There is no single moment of time in either frame when the pole could actually
be fully contained within the barn.
However, the clear implication of our considerations here is that length
contraction is not a physical shortening, but is merely an observational
consequence of time desynchronization. In SR, physical bodies do not actually
change dimensions. This also explains why “length contraction” does not flatten
spherical bodies in their direction of motion, making their rotation about a
perpendicular axis irregular (an effect called “Thomas precession”). []
in Lorentzian relativity
Our next question is: what does Lorentzian relativity (LR) say about length
LR agrees with all existing experimental evidence at least as well as SR, yet
has no time desynchronization. As we see from the above considerations, the
absence of time desynchronization means that LR has no length contraction
The original motivation for length contraction was in the interpretation of the
Michelson-Morley experiment. In it, each half of a split light-beam travels
along one of two equal-length, perpendicular arms of an interferometer and
bounces off mirrors back to an observer. Surprisingly, both beams arrived back
at the same time and produced no interference fringe shifts, even if the
interferometer was turned to different orientations. Yet the Earth is moving at
a speed of no less than 30 km/s in its orbit around the Sun, not to mention
possible higher speeds in its Galactic orbit. So the round-trip travel time for
a light beam through a light-carrying medium (aether) was expected to take
longer when moving along the direction of observer motion than when moving
perpendicular to that direction, for the same reason that a round-trip in a
canoe takes longer when the stream is flowing than when the water is still. Yet
observations show that the round-trip times for light beams in any direction
were actually the same. This was explained in SR by length contraction along the
direction of motion, making the interferometer arms shorter by just enough to
compensate for the otherwise-longer travel time through the moving aether.
In LR, because there is no length contraction, it cannot be invoked to explain
the lack of fringe shifts in the Michelson-Morley (M-M) experiment. But then the
absence of fringe shifts implies that Earth has no motion relative to the aether.
However, this is not a problem for LR. As is now becoming well known, LR
recognizes the local gravitational potential field, also known as "elysium", as
the light-carrying medium [,,].
And that local field has no motion with respect to the Earth's center of mass.
So the M-M experiment shows no fringe shifts because local elysium is entrained
by the Earth and at rest with respect to it.
An exception is that Earth does rotate with respect to its own gravity field.
However, rotation does produce fringe shifts in an M-M-type experiment.
This is known as the Sagnac effect, which was first seen in 1913 in the
laboratory using a rotating platform for the M-M experiment. The same effect was
later replicated in the Michelson-Gale experiment in 1925 using the actual
rotation of the Earth to rotate the M-M interferometer. So LR predicts the
absence and presence of fringe shifts just as experiments confirm they should
We conclude that length contraction, which has never been seen per se in any
experiment, is not a real, physical effect. In SR, it is merely a by-product of
time desynchronization (the lack of remote simultaneity), and is therefore
illusory. In LR, length contraction does not exist, but is not needed because
elysium (the light-carrying medium) is the same as the local gravitational
potential field, and is entrained by the Earth. So no fringe shifts are expected
in a Michelson-Morley-type experiment unless there is rotation, which is just
what experiments show.
Last updated 2007/08/25