Felgar wrote: I think you're confused about 2 things: 1) What determinism means, and 2) what a system is.
If I look confused it's only because the written communication has some well known disadvantages, and I am not really managing it well. Adding more sentences to other sentences could be counterproductive for clarity, I imagine, and the result could even be that there are more occasions to be criticised. I would try to be shorter, but it is almost impossible if I have to reply from many sides. Apologies.
Believe it or not, I know what determinism is and what a system is. And I can give you a definition that comprehends both concepts: a deterministic system is a system for which it is possible to integrate over time its differential equations, in order to get an analytical time history in form of further integrated equations.
Felgar wrote: Ok, so when you say that a system is never predictable, this is false and I believe the reason is that you are incapable of defining a system.
I said that a real system is never 100% predictable. Therefore, a real system is never 100% deterministic, even if ruled by deterministic laws. And this is because we can't model its behaviour at 100%. And this is because of several factors:
- insufficient accuracy of the mathematical laws
- impossibility to measure initial parameters
- insufficient accuracy in calculations
etc
A theoretical system could be 100% predictable: in this case it's a deterministic system.
But, the factors that prevent us from an exact modelling could be the same as before.
But let's go on.
Felgar wrote: So let's take our solar system - the sun, the nine planets, the comets (which we know little about) and the asteriod belt are part of this system. THIS system is entirely predictable in that if we knew the location, size, mass, and velocity of every object within it, we can determine the system's exact state at a point in time in the future.
Sorry, I was always trying to distinguish between real and theoretical systems. But here, I have got the feeling that the disagreement (or misunderstanding?) is mainly on the fact that I consider chaos as intrinsic in the Solar System, and you don't.
For what I said before, we have that:
1. Newton's gravitational Law rules the Solar System according to the 1/d^2 law
2. this law is perfectly deterministic, and we can see it when we model two bodies and their evolution
3. if we apply this law to three bodies, we get a three-dimensional matrix that contains differential equations applied to the three bodies dynamics.
4. when we try to integrate analytically the differential equations, to get the evolution in time, we realise that the equations are not analytically integrable.
5. even more, when we try to integrate numerically, we find cahotic behaviour
6. even more, the chaotic behaviour is intrinsic to the three bodies system because it can be tested or observed
Item 6 demonstrates that any gravitational system (deterministic as you like) with more than two bodies contains a chaotic component. And this is independent on the fact that we are not able to know all initial parameters.
When you say that if we knew all initial parameters in an exact way we could predict the evolution of the system, then I start to be concerned. This is another key point.
7. In theoretical systems, you can always assume that you know all initial parameters.
8. In a real system you are NEVER able to know all initial parameters (Heisenberg principle, tested)
Therefore, we continue with real systems only, because there is no disagreement (or misunderstanding??) about theoretical systems.
Due to 8 above, applying it to the fact that the Solar System (based on deterministic Newton's law) is intrinsically chaotic, we get that we are never able to exactly predict the evolution of our spacecraft Cassini-Huygens.
I don't like to consider a possibility ("if we knew") that is totally impossible and not scientific.
Felgar wrote: But wait, what if another star passed close enough to influence our system? Well then your system is no longer predictable because it has been acted on by an element not contained within it.
Not exactly.
An additional body travelling into the Solar System would not change the "chaotic determinism" intrinsic of the system itself (due to Newton's law).
It would not change the fact that it is potentially unpredictable (or not 100% predictable).
It would not turn the predictability into unpredictability.
It would not change the equations.
It would not change our impossibility to know all parameters.
It would just modify its equilibrium point.
It could be a dysaster because the equilibrium of a planet could be spoiled. Here we could talk about stability and equilibrium for pages and pages. I only reply that it would be a perturbation that might be sufficient to modify the tiny stable equilibrium of the Earth or other planets.
Felgar wrote: But holding onto the fact that the solar system is unpredictable is really a ludicrous notion because the chances of anything external interferring with the system is negligable. We make an assumption and move on - and frankly I see no reason for NOT making that assumption.
As said above, I am not using the term "unpredictable" in absolute terms. It is not fully 100% predictable, therefore it is somehow "unpredictable". It's not playing with semantics: it's my attitude to avoid absolute quantities. You are probably more comfortable than me with absolute concepts. Apologies.
Felgar wrote: Only because we did not account for every element that could cause a failure - the definition of our system was incomplete. You know, the complex factors like a centimeter not being equal in length to a inch.
Your example is really important (it happened to an airplane: they run out of fuel because of confusion between gallons and liters! I think it was in Canada!!!), but it is not the type of things I was thinking of.
I am saying that failures are unpredictable. But let's go on.
Felgar wrote:I'm trying to be gentle, but it seems like you are not understanding what I'm saying, and this statement reveals your ignorance. Determinism has NOTHING to do with a system being chaotic. There's deterministic and non-deterministic, and that's it. Ordered vs. chaotic are a completely different concept altogether.
As said before, I disagree (not on the fact that you try to be gentle!!!!!! I appreciate, believe me).
There is definitely chaos within deterministic systems. Just solve your differential equations (once again, deterministic, yes).
Ah, and here I remember that I was criticized also by Dan, about the weather chaos. Write the differential equations that describe atmospheric phenomena (I don't recall them, of course: I can look for them and I will post them): when you integrate them, you find chaotic behaviour also for them. And, again, I don't care about "if we exactly knew every parameter", because this is impossible.
Weather is chaotic even if ruled by deterministic equations, and given that (for accepted/agreed reasons) we can never precisely model its behaviour, I am entitled to conclude that we can never exactly predict its evolution. And, if we can't exactly predict its evolution, in practical terms it is not 100% deterministic. You defined (correctly) determinism as 100% predictability.
If something is not 100% predictabile, for whatever reason, sorry, then it is not 100% deterministic.
Felgar wrote: See, a deterministic system is one where the future state can be predicted exactly. Even if there are no natural ones, I could make a theoretical system that IS deterministic, because within that I can define the system completely and keep it free of other influences. I recommend you research Turing State machines - Alan Turing who was a pioneer of computing (and who had a huge part of our winning WWII through his work breaking German codes) worked on the principles of determinism. And really, most computer code is nothing but glorified Turing State machines which are by definition determinstic.
Here I agree and I have no problems. But I disagree with the Turing test for distinguishing a person from a computer. But this is out of topic.
Felgar wrote: When you finally understand determinism, contrast with chaos which I feel that Marco adequately defined:
So a choatic system is more difficult to predict in practice because extremely small innacuracies in the measurements of our starting conditions causes large changes in the results of our predictions. In theory though, where we CAN define an exact starting state, a chaotic system is no less predictable than any other.
I thought Marco said something slightly different. Apologies.
I don't disagree with the fact that we can start with a set of equations, a set of exactly defined initial conditions, and therefore come up to a set of time equations that predict the system.
I am saying that there is chaos even in deterministic systems. And here, I understood Marco was of a different opinion.
Question now, please (let's forget predictability at the moment):
Would you agree that chaos is implicit in a deterministic system? As in my example of the three bodies and Solar System?
Would you agree that chaos is implicit in a non-deterministic system?
If you answer no, then I would like to know in which point exactly you disagree.
If you answerw yes, then I go on with my basic question to Marco (and to anyone):
How can you demonstrate the existence of an intelligent creator only on the basis of physical laws that contain chaotic behaviour?
Felgar wrote: If this is still not clear, I think I need to know your age and status of education - because I'm not sure that any further explanation (like demonstrating a Turing machine) will help.
I have no difficulty to give more info about myself:
born 1954 in Turin. University grade in aeronautical engineering. One year doctorate in aerospace engineering (then stopped because of military service and job). Working since 1982 as a development engineer in Flight Control Systems (fly-by-wire) for high performance aircraft. Reader of Scientific American magazine.