bilby said:
A growing block universe assumes the existence of a universal now.
Relativity says that there's no such thing.
How does a growing block universe assume the existence of a universal now? It's basically a crystal growth model -- new spacetime gets added to the surface of the block of existing spacetime. So a "universal now" would presumably be a (hyper)plane that has block of spacetime on one side, and does not yet have any spacetime on the other side, that can serve as a preferred reference frame. Which is to say, it's a face of the crystal. So why would a flat face have to exist anywhere on the growing block? Who says the crystal has to grow uniformly?
How can that hyperplane have a shape that allows for events to have different sequence for different observers?
What hyperplane? That's my point: there is no hyperplane -- no "universal now" -- provided the growing block universe grows in a bumpy irregular fashion, with some locations expanding along the time dimension ahead of others. Any individual observer will perceive the time of distant events based on her own relativistic reference frame, and will construct a hyperplane in her imagination that's the set of all events that are simultaneous from her point of view. But none of those imagined hyperplanes will correspond to the actual shape of the growing block. A typical observer's mental hyperplane will slice through the block in some places where the block has already grown beyond it, and pass beyond the existing universe in other places where the growth hasn't yet caught up with that observer's imagination. (And no worries -- the growth can catch up in time. All not-yet-caught-up locations are outside the observer's light-cone, so everything she can actually see will be comfortably within existing spacetime, provided the bumps on the surface are never steeper than 45 degrees.) Consequently, none of the observers are privileged. So this model doesn't contain the "absolute rest" reference frame that relativity says doesn't exist.
As far as preferring one relativity-consistent model to another goes, the growing block model beats the eternal block model hands down, IMHO. The eternal block model has a nasty entropy problem. If the whole thing just "is", without ever "becoming", then what accounts for the state of the universe in 2018 being consistent with having evolved from a much-lower entropy state of the universe in 1918? If you take, as the "2018 state of the universe", an arbitrary slice through an arbitrary eternal block, and then you play the laws of physics backwards from that slice a hundred years, the state you compute for 1918 will with 99.9999+% probability have a greater or equal entropy than the 2018 state. Only a few very special arrangements of matter in 2018 are compatible with having resulted from running the laws of physics forward from a lower entropy 1918 state.
Here's a simplified example. Let's say you take as the state of the universe the following: a star, a gas giant, and a line of 19 dirty snowballs about to crash into the gas giant. Pick their exact positions and velocities any way you please. Now run the law of gravity backward for two years. You are not going to retrodict that all those snowballs coalesce into a single comet, unless you are ridiculously lucky in your selection of positions and velocities. The trouble with the eternal block model of time is that improbabilities on that scale happen constantly,
every single time you look, for no comprehensible reason. The state of the universe just is always consistent with having evolved from a low-entropy predecessor state. Why? No reason, it just is. By magic, apparently.
In the growing block model, there's a perfectly straightforward comprehensible explanation for this otherwise insane coincidence: the 2018 state is exactly consistent with the 1918 state
precisely because it was actually computed from it. That's what growth of the block is: it's the laws of physics incrementally computing the future, from the past, epsilon of spacetime by epsilon of spacetime.