On Oct 3, 2012, at 7:47 AM, Fred lunnon <fred.lunnon@gmail.com> wrote:

Am I missing something here, or isn't this all rather obvious?

Assuming that general relativity and quantum mechanics are (approximations to) separate portions of some undiscovered super-theory in which they are not mutually inconsistent, then spatial overlap in the frame of one observer will rotate into temporal overlap in the frame of another.

Lorentz transformations preserve the past and future light cones of an event, so that's not what makes it obvious. As far as we know, and with the exception of very tiny effects due to "CP-violating weak interactions", the laws of physics are invariant with respect to flipping the sign of time in all the equations. (In quantum mechanics we also have to apply the Galois automorphism i -> -i to all the complex amplitudes.) So if we made a "movie" of some physical process and then played it backward we wouldn't be able to find deviations from physical law. The strict notion of time-reversal symmetry (and the illusion of causality) is of course at odds with everyday experience (scrambling eggs, proving theorems). This is especially true if the main protagonists in the physical process are called "Alice" and "Bob". These scenarios are always massively time-asymmetric in their boundary conditions (e.g. Alice is radiating blackbody radiation into a cold room as a result of metabolizing ATP) and therefore look like evidence of a fundamental causality. Conversely, as experiments get better at achieving the goal of isolating a physical system from its environment -- and measurements run up against quantum principles -- the illusion of causality is increasingly exposed. -Veit