Cosmology and Time's Arrow (nerd content.)

[skywatch]

There's a simple a priori issue that bites way before one gets out into the real world that can be nicely exposed using Conway's Life. Take, say, a glider. The pattern that makes up a glider, once in place will deterministically keep going such that whatever the state at any time 'T' will determine any future states. However, there are a multitude of potential states that could have led to 'T'. In other words, even in an entirely deterministic, linear world, while the future is overdetermined the past is underdetermined. Obviously the real world is slightly more complicated than life, not to mention non linear and stochastic, (if Bell's inequality holds) but the principle remains exactly the same.

To elucidate this, imagine a simple rubber ball sitting still on the floor in a simple square room with, say, four windows. If all the physical information about the room at any given time 'T' is available to you, you will be able to predict pretty reliably what is going to happen to the ball in the next five minutes. However, from 'T' you will not be able to say what happened in the last five: did it bounce in through any particular window, has it sat there for months. You simply can't tell.

See the problem?

(if you haven't played with Life, try it, it's incredibly helpful for thinking about this sort of stuff.)

I have played a bit with "Life". Cool stuff. I love how large patterns form out of simple rules. You're probably also right that you can't be sure about backwards assumptions. I am inclined to agree with chaos theory, that given enough complexity you also can't actually predict an outcome very far into the future. What I like about the article I linked, is that they don't require any sort of specific outcome, and their observation can hold true with more complexity. All it needs is for a chaotic system to briefly hit a singularity, then become chaotic again. If time is increased entropy, then it goes in both directions through the singularity. Whether or not that happened in the big bang is another story, but I like how the basic idea is rather intuitive. (I can't do the math, personally, I will be quick to admit!)

[skywatch]

and where is the cat at this stage?

Both dead and alive of course! Betcha when those quantum dudes take on this theory they'll find that time is leaking out of the singularity ... a bit like what happens to my wallet whenever I take a watch in to get serviced.

[Der Amf]

The Angel of Death accounted for whoever carried the rubber ball in, right?

[skywatch]

The Angel of Death accounted for whoever carried the rubber ball in, right?

it was the salmon mousse.

[Der Amf]

it was the salmon mousse.

The grim reaper called the angel of death "the one whose sandal straps I am not worthy to untie"

[skywatch]

The grim reaper called the angel of death "the one whose sandal straps I am not worthy to untie"

Well, you know, he must be one of the men from the village.

[Nokie]

Very interesting reading. Keep learning new things today from this site.

[skywatch]

Very interesting reading. Keep learning new things today from this site.

You're in the Bay Area aren't you? I'm in the south bay... we've done GTGs through the old site, possibilities abound!

[Matt]

I just had a dig through the original paper. Unbelievably, they are basing all of this on an body simulation in which the N = 100. So they are trying to simulate the early universe, itself a question of some uncertainty, but certainly a time in which the only 'bodies' were hydrogen and helium (if that) as the rest of the periodic table is made by stars. Stars which didn't get to ignite for at least 300 and more probably in excess of 400 million years after the events being (and I use the word very loosely) simulated here.

N body simulators are traditionally used for modelling middle sized slow moving objects such as planetary systems. It's a signally inappropriate approach to the early universe and even if it was, the axioms being thrown about here are highly questionable. I'm sure that the maths is impeccable, but the premise set and axioms are just silly and the size of the simulation about as credible as doing it using ping pong balls and string compared the roiling complexity of the early universe.

Even if all of this were not true, and it is, the problem I highlighted earlier still bites and so when they say, in the bloody abstract, that:

Thesimplest nontrivial time-symmetric law that can be used to model a dynamically closed universe is theNewtonian N-body problem with vanishing total energy and angular momentum.

They are talking tosh. It's simply not time symmetric. Take any given orbiting body at T. With Laplacian knowledge of the rules and state of the system at T one will be able to predict the ongoing behaviour of the body (and by extension) the whole system at any future T. However, once again, from T, any retrodiction falls foul of the fact that there will be vast set of ways in which the orbiting body could have arrived in that orbit and a vast set of times at which it could have happened.

In short, their metaphysics is utterly crocked and GIGO.

Yeah, I know, famous physicists and all, but...

[skywatch]

I just had a dig through the original paper. Unbelievably, they are basing all of this on an body simulation in which the N = 100. ...

Well, actually I think the N was 1,000, but I tend to agree with your point. My first inclination is to compare their symmetrical approach to the multiverse approach, where singularities arise constantly and each one puffs out into its own set of laws. That possibility is actually testable in a device like the LHC, where they are starting to search for energy leakage in certain particle breakdowns. I don't see how the symmetrical model could jibe with a multiverse model... but then most of this is beyond my comprehension. My interests are more as a spectator, because I think the questions are fascinating, and sometimes I enjoy watching the efforts at model building, even for the sheer pleasure of seeing a glimpse of beauty in it.