I think the window analogy is quite apt – as long as I can see the whole system, it’s closed; if I can’t, it isn’t, strictly speaking (though it still might be usefully describable as isomorphic to some closed system).
Though of course on another level, you can’t really disentangle the film that’s being viewed from the viewing of the film – for instance, one could think about, say, the data on a DVD as being the initial conditions that starts a particular evolution of the film viewing system, i.e. the DVD player, said data itself being the result of the evolution of some system – the film making system. You keep trying to view the film as a system in itself, but it’s rather more like the output of a system – which can of course be viewed as another sort of system, but then one must be careful to steer clear of level confusions. It’s like the value of pi being the output of some computer program – of course you can view the specific decimal expansion itself just as a series of numbers, but determinism in this sequence doesn’t mean that the next digit must follow from the preceding one(s); similarly, determinism for the film doesn’t mean that the next frame or scene must follow from its predecessor. You can of course create an algorithm that takes the first n digits of pi and then produces the n + 1st – but then, you can just as well create an algorithm that reproduces a given film (something like what is done in fractal compression schemes).
Not sure if I’m getting across what I mean… Say, a computer produces the sequence 6899817250753234 – that there’s no obvious way to predict the next number from the previous ones doesn’t mean that it’s an example of indeterminism; it was in fact generated in the most deterministic way imaginable. Now of course, given just the first ten numbers, say, one would be incapable of predicting the eleventh; but I can’t really consider this sufficient for a judgement of indeterminism. The sequence doesn’t produce itself out of itself, and neither does the film – hence, the information needed to predict the next step just isn’t there; while any closed system would have to, in some way, derive its next state from the previous one (or else do one of those strange indeterministic ‘jumps’ where there’s no way to decide between the alternatives, and yet, the system still does).
I think, to me, the most telling definition is that there ought to exist sufficient reason for any event. So for any digit in the number sequence, sufficient reason is given by the algorithm creating the sequence; but that doesn’t necessitate that the reason should be contained in the sequence itself.
And as a matter of fact it does matter, since the truth is not known,
and as long as the truth is not known, I may assume what I favor to
be the truth.
And if by some chance what I favor is falsified by presently unavailable
evidence, OK then, I will learn to cope.
Right after you tell me how, in an indeterministic universe, anything ever happens at all. Or how the randomness generated by approximating certain deterministic models, which you criticize as being ‘mere approximations’ implies that the fundamental dynamics is stochastic. Or how matching EPR results don’t imply that special relativity necessitates a deterministic universe…
ISTM this the nub of the OP. It’s an interesting and important question. My answer is sometimes yes and sometimes no. That is, sometimes people who do bad things are responsible and other times not. I distinguish the two based something I call soft or psychological compatibalism. I explained the theory in this thread and expounded on it here.
In sum, as relevant to the OP, I submit we may reasonably assume people generally have volition and thus are responsible for their actions. This assumption sometimes breaks down, though, and we should view their behavior in that light. Which isn’t to say the behavior is okay or should be condoned. But rather that our response in such cases should be therapeutic and goal oriented not punitive. IMHO, if the debate over determinism means anything, it means this.
Eh, that post was a bit more catty than called for; I’m sorry.
As for logical consistency, I merely meant it’s logically consistent as an answer to the first cause problem: if there’s no ‘first cause’, then there’s no problem, obviously, and such a universe may be deterministic for all times without running into any logical troubles. Obviously, I have no means of proving an infinite universe logically consistent in itself; but then, neither do I for a finite universe, or any other one, and I’m not sure on what terms such a proof should work, even. Certainly, there are infinite structures which are logically consistent, such as the whole numbers, or the reals – so if I’m allowed to choose either as a model for a timeline, I’m good, but of course, one may argue about whether or not I am.
The thing is that going down this road, we’ll get bogged down in specifics – having now a page or two of argument about infinite universes, then maybe about quantum creation, etc., where at some point, I probably won’t be able to answer questions that open themselves up with any good arguments; the larger issues remaining unresolved all the while. But that doesn’t mean that therefore, indeterminism must be right! For while it may be difficult to follow determinism all the way down to its logical conclusion, and see whether it’s turtles or something else, indeterminism doesn’t even seem to be capable of getting off the ground.
In particular, one can give a closed, mathematical description of a deterministic system: going from the initial conditions, via algebraic manipulations, one can derive its evolution, in principle for all times. Up to now, no well-acknowledged phenomenon has resisted being formulated in these terms. However, for indeterminism, such a description does not exist in principle: at some point, one reaches a fork for which there is not enough information to decide how the system’s evolution is supposed to proceed; one would then have to ‘jump out’ of the formalism and arbitrarily impose a choice to continue.
That’s why I claim deterministic explanations are more parsimonious: Some things clearly are deterministic, we can model determinism well, and there is no substantial observational evidence towards the inadequacy of such modelling; whereas in the case of indeterminism, we don’t know of a single phenomenon necessitating it, and have no idea how to model it – how it ought to work that one path is taken, when there is not enough information for determining which one.
Of course, that might just be our problem – reality may be fundamentally indeterministic, and our math, modelling and logic simply inadequate to describe it; that’s at least a conceivable, if somewhat depressing, possibility. But at present, there is nothing that forces this possibility upon us, so the best strategy is to stick with what we know is possible, and hope that nature doesn’t play tricks with us.
It just does. Sometimes there’s a causal connection to something that’s gone before, sometimes not.
The models don’t “generate” randomness at all, there’s always a randomness factor you have to add in. Like say in one of you beloved particle models - the initial vector for each particle has to randomized before letting classic mechanics do its work.
So the root of the whole thing is stochastic.
By “EPR” results I’m going to assume you mean Bell’s experiment results - how does entanglement necessitate determinism? It does the opposite, by booting locality out the window.
That doesn’t answer anything. I’m not asking why, but merely how it happens – and anything that happens has a how of its happening, merely by virtue of happening. Consider a tree falling in the woods. How did it fall? By traversing a set of states (which might or might not be continuous) in between ‘standing up’ and ‘lying flat’. At first it stood upright, then it tilted a little, then it tilted a little more, etc. That’s how it happened – the why of it, I’m not considering right now. The important thing to realize is that anything that does happen has such a how – as anything happening can be partitioned into a sequence of sub-events, if you will. ‘Happenings’ don’t happen at singular space-time points, they’re processes.
But there’s no such how in an indeterministic world. The sequence of events bifurcates at every indeterminacy. So there are several possible ways to continue this sequence, with no way of choosing any particular one; thus, there’s no way for anything to happen. ‘It just happens’ poses no way out, because every randomness posed behind this just in turn depends on such a bifurcation itself, and we’re trapped in an infinite regress. For the same reason that one never can generate true randomness analytically, no indeterminate choices can be made.
Only in order to account for the typically unknown history of the system; one can just as well start with a gas in a very definite state (say, all atoms in a corner, with a well-defined momentum vector for each), ‘turn on’ the evolution, and compute every successive state exactly. If one then uses the usual macroscopic approximations, one will again get the usual statistical descriptions back. The randomness in the resulting effective description always comes about from ‘papering over’ our ignorance of the microdynamics.
No, I meant the results obtained in my ‘entangled-with-Andromeda’ example. In order to explain the correlation, one must conclude that each team measures the same quantum system at the same time, and that thus, the viewpoint of each observer is equally valid, which means that the outcome of the Andromedan choice, and by extension, every choice, must be fixed.
What does locality have to do with determinism? (And besides, local theories are perfectly compatible with Bell tests as long as they don’t contain hidden variables – i.e. the insistence that any quantity has an underlying, ‘real’ value – as well; Bell’s theorem only excludes the combination of locality and realism.)
this is true only in a universe with strict causation. ours need not be such a universe. indeed, quantum mechanics (in its many interpretations) tends to compel one towards the conclusion that there is no strict causation in our universe, merely a likelihood of causation.
Well, the argument is that I don’t see how, in a universe without causation, anything then ever happens – so it’s an argument against the possibility of universes without strict causation.
True for some specific interpretations – Copenhagen and most others that propose an actual ‘collapse of the wavefunction’ --, but not in others, like many worlds, relative state, or decoherence, for instance. Indeed, it’s only if you impose something extraneous to the quantum formalism – like a collapse – that indeterminism arises; ‘bare bones’ quantum mechanics, as in Schrödinger evolution, is unitary, and hence, deterministic. So I’d say that here – admittedly somewhat contrary to the mainstream opinion on the topic – parsimony leads us to determinism, too.
that’s because you are assuming things must be caused in order to happen. such need not be the case in a universe without strict causation.
this is something i’ve been unable to understand regarding the philosophy of science: if any measurement enforces apparent indeterminism, and measurement is the only way the universe is knowable, what good is it saying that a particular interpretation is deterministic? by what mechanism does a measurement cause me to find myself in this particular one of the many worlds? it sure doesn’t appear deterministic.
No, I’m not, but judging by your and MrDibble’s reply, that seems to be a subtler point than I originally thought… With regards to the tree falling, I explicitly don’t care about the cause of its falling, merely about the sequence of events that describe its falling. One event follows the other – not causally, but merely spatiotemporally. However, this spatiotemporal sequence is enforced by causality; if you then remove causality, there is also no spatiotemporal sequence. So, for the falling tree, there is no way to get from ‘standing up’ to ‘lying flat’ – there’s no way to get to one from the other. As I’ve tried to emphasize, I’m not asking for the why – which would be asking for causality – but merely how – which just asks for the way it happens. Such a ‘way of it happening’ seems not to exist in an indeterministic universe; but then, how does anything happen at all? How does it happen, if A and B are two possible choices, and choice between them is not determined by all the information in the universe, that still A (for example) gets chosen?
The thing is, Everett’s original relative state formulation consisted of little more than just letting quantum mechanics do its thing. That’s manifestly deterministic; the state of the system at initial and final times are connected by a unitary transformation, and applying the Schrödinger equation, from the initial state, the final state can be computed exactly (if you don’t introduce collapse, that is). So if, for instance, you carry out measurement on a spin-1/2 system in coherent superposition of its basis states |0>, |1>, and if we denote with |U,r> the state ‘you, ready to measure’, with |U,0> the state ‘you, having obtained outcome 0’ and with |U,1> the state ‘you, having obtained outcome 1’, you’ll get an evolution like this:
|U,r>(|0> + |1>) –> |U,0>|0> + |U,1>|1>
Now, this is perfectly deterministic – it’s just Schrödinger evolution; indeterminism would only come in if one were to, say, drop the last term, thereby effecting a ‘collapse’. U will have evolved to a state in which you have both observed the determinate outcome 1, and the determinate outcome 0. Now, this right hand side has been interpreted as describing somehow distinct ‘worlds’ – from which perspective your question, effectively: ‘What put me into this world, rather than that other one?’, is sensible at first brush. But it’s not like a garden of forking paths, where you go down either one; you go down both. There’s no two classical worlds, and your consciousness somehow continued on into this one rather than that one – there’s only a quantum world, and you and your consciousness exist in superposition alongside with it. The reason you don’t notice anything weird is that there’s no state corresponding to an awareness of the superposition – there’s no |U, 0 + 1>. Whatever you perceive or experience, you perceive or experience determinately – it’s just that there’s a coherent superposition of all these experiences and perceptions.
I tried to be very careful to point out that I’m not talking about causality if I ask about the how, merely about the actual sequence of events that generally accompany ‘a thing happening’, even things that ‘just’ happen. Perhaps the clarification I gave above will help.
Yes, and that’s exactly the problem. Sans determinism, sans causality, there is no way for one single path to be selected; all possible paths stand on equal grounds, mixed as that metaphor may be.
No, you can just give them any momentum you like. Of course, if you assume all of them have the same momentum, and assume a container with perfectly smooth, perfectly aligned walls, you’ll get a pathological case of a bunch of gas atoms bouncing around the room like in that old screensaver, but in reality, walls are not perfectly smooth, they’re not perfectly aligned, and the atoms won’t have perfectly identical momenta (uncertainty alone will take care of that). If one does set up the momentum distribution as a random one, it is merely to account for those real world imperfections. But I can give every atom a momentum according to a well defined rule (after all, otherwise, such calculations would be impossible to implement on a computer), and have all the usual macroscopic statistical features emerge, if I don’t look at the microdynamics.
There is, well defined by each observer’s hyperplane of simultaneity.
No. It’s entirely possible that the state of the universe in the next moment depends on the state of the universe in the previous one in such a way that each point’s future doesn’t merely depend on its immediate surroundings, but on any other, distant region, or even the state in total, and still have determinism.
this is not so. there is just no enforced spatiotemporal sequence. there is merely a sequence of events.
does it not seem absurd to you to ask “how” something happens in a universe without causality? the question of “how” is aimed at determining the chain of causality. if no such chain exists, there can be no “how”. it’s like asking what happened “before time”. or dividing by 0. there is no sequence of events and a rule that leads from one to the next, there is merely a sequence of events.
The ‘enforced’ only referred to a deterministic universe; that it is not enforced in an indeterministic one is precisely the problem.
Consider a thrown ball. During its flight, it is at a certain time in a certain height and at a certain distance. This sequence of positions is what I’m talking about, nothing else. This must exist, because it’s what it means for the ball to have been thrown – the happening (that the ball has been thrown) implies the how (that it traversed a path in space and time). This is the case both in a deterministic and an indeterministic universe.
Now, in a deterministic universe, the existence of the how, of the sequence of positions, is easily explained: each new position follows causally from the previous one. But in an indeterministic universe, as you point out, we can’t appeal to this reasoning – so how can there be such a sequence, such a how, at all?
No, the question of why is aimed at determining the chain of causality. How is merely descriptive.
A doesn’t “get chosen”. A is “what happened”. There’s a vast semantic difference between those two statements.
Until they happen. Then they’re “what happened”. It doesn’t matter that there were other possibilities - although it seems to paralyze you, it has no such effect on the Universe. The Universe doesn’t care about “selecting a single path”, in a non-causal Universe that is a broken metaphor.
So you’re moving the randomness out of the initial momentum choice and putting it into the boundary conditions. That’s just moving goalposts. Anyway, the modelling I’ve seen (that you’ve shown) looks like perfect elastic collisions with smooth bounding walls, AFAICT.
And that’s your calculable classic mechanics, is it? “Now with Built-In Uncertainty!”
They may be microdynamics to you, they’re the inherent stochastic fabric of the Universe to me, you know, the bit I spoke about when I first posted to this thread? The bit you said didn’t actually exist?
Only if you assume there exists a causal chain between position now and future position, like I said already. If you don’t accept that causal chain, you can’t even construct such a hyperplane. Instead you get a complex ramifying hypersurface. Not so easy.
But not relativity, because you’re in the “spooky action at a distance” zone and we all know what Einstein thought of that.
Basically, you’re arguing for superdeterminism. Which really means arguing for Local Hidden Variables, because that’s the only way to have superdeterminism post-Bell experiments.
The problem with this objection is that it applies with equal force to all the proposed theories (libertarian free will, strict determinism, traditional compatibilism and my own little theory of soft or psychological compatibilism). The simple truth is that we don’t know how the mind works, yet still have to make judgments about people and their behavior. One has to pick an answer with the mechanism unknown. I explain why I reject determinism in the links above. You’re picking it (not stated whether of the strict or traditional compatibilist stripe) on grounds of parsimony. Okay, then. What say you about the issue of responsibility? Which, after all, is the question raised by the OP.