Hi William,
I think I need to emphasize a point: There is a measurable difference
between relative rotation and "absolute" rotation (take this latter phrase
to mean rotation relative to the fixed stars). To paraphrase Mach, you can
only produce a curved meniscus surface in a pail of water by rotating the
water in this "absolute" sense; you will not observe this meniscus if you
leave the water still and put only yourself into a spin, while standing on
the extended axis of the pail. The spinning water develops a meniscus only
when IT is rotating, whether you are rotating or not.
Events occur among the constituent particles of water when the water is
spinning which do not occur when it is not spinning. This has to do with
properties of the water (internal stresses and the water's response thereto)
which come into play when it rotates - this is part of the physics which SR
cannot be expected to divine simply from its stress free behavior in
translation.
If relativity alone determined the details of rotational behavior, it would
be identical for ALL rotating materials (water, steel, etc); there would be
no room for variation in the relevant properties of different materials.
(In the same way, beginning students will often mistakenly expect
conservation of momentum/energy alone to determine the details of a
collision in the general case, allowing no room for a variety of results
from a variety of interaction mechanisms.)
The same must be said of a spinning magnet. Its electromagnetic properties
change, TO ALL OBSERVERS, when IT spins. Internal OBJECTIVE events occur
which do not happen in a non rotating magnet. No purely kinematical
transformation theory can divine this behavior unless you feed it the
particular PHYSICS of the phenomena involved under rotation conditions.