Slashdot Mirror
2 Futures Can Explain Time's Mysterious Past
cyberspittle sends this excerpt from Scientific American:
Tentative new work ... suggests that perhaps the arrow of time doesn't really require a fine-tuned, low-entropy initial state at all but is instead the inevitable product of the fundamental laws of physics. Barbour and his colleagues argue that it is gravity, rather than thermodynamics, that draws the bowstring to let time's arrow fly. Their findings were published in October in Physical Review Letters.
I read the linked article, but not the original paper, and here's my summary (sorry for not limiting it to four sentences, but I can only do so much):
One of the outstanding problems of physics is to explain the directionality of time. It's easy to distinguish the future from the past, even though the fundamental laws of physics operate the same forward and backward. This "arrow of time" is normally explained in terms of increasing entropy, but that doesn't answer the question of why the universe started out in a low entropy state.
The researchers here made a computer simulation of a simple system of particles interacting with gravity, and a metric of entropy based on how close the particles were together. What they observed is that gravity can (briefly) pull the particles together into a low "entropy" state. You can then post-rationalize this as having a single low-entropy "starting point", from which you get two high entropy "futures", which consist of earlier and later states by external simulation time. But both would be "future" according to a creature living in the simulation, seeing the entropy increase as one moves away from the "big bang". It's the whole "what is north of the north pole" thing - if you walk north toward the north pole, you eventually start heading south, even though you haven't turned around.
The researchers claim that it's gravity which is the key to giving you this starting point, but other researchers quoted disagree, saying simply that having an entropy metric which doesn't have an upper limit is sufficient to give similar behavior, and claiming they'll be publishing a paper soon which will show this.
What does it mean to you? Nothing, really. It's still rather preliminary and esoteric at this point. Maybe in another few years they'll be some science special on TV which will discuss it in a "isn't science mind-boggling?" sense. It (or rather similar such arguments) will eventually inform basic theories about physics and how the universe works, but right now it's not even at the "one day sir, you may tax it" stage.
The actual scientific article was published on arXiv in september. Gravity does not appear to be central to the problem, it is just used as an example here. They basically look at a toy problem where a large set of particles with simple interactions give rise to solutions where they can identify variables that increase monotonically away from a minimum, and hence can be used as a time variable. It is basically an entropic argument worked out in detail for a simple system.
Carrol et al have published related ideas, and here is a popular science talk by Carrol (the main argument starts around 19 minutes into it).
Actually we can do better than that. The arrow of time is baked into fundamental particle physics and we have known this since the 1990's when an experiment, CPLEAR, showed that kaons turn into anti-kaons at a different rate than they switch back. This is completely independent of entropy and the result was further improved on by the Babar experiment only a few years ago showing that the 'T violation' occurs in B-mesons as well.
The article is wrong when it says that the laws of physics work the same going forwards or backwards in time. They do not and there is data to prove it. So the 'arrow of time' does not need any entropy to define it - it is baked into fundamental particle physics.