Free will and determinism are both names for things that don’t exist. There is something to them, but the appearance that they describe separate real states of affairs, only one of which can be true, is an illusion. That’s why there are various ways of demonstrating that either one must be the true one. The underlying problem is that our comprehension has settled on tidy little concepts that are too clumsy.
This is like the wave - particle duality.
At least, I think so. I really really really think so.
No; no one has ever demonstrated that “free will” is true or even possible. How could they? If it isn’t deterministic and it isn’t random, then what could it possibly be even in theory?
Please, explain or model Brownian motion in such a way that you never use a random element. Einstein called it random, Langevin called it random, who, exactly (besides you), says it isn’t random?
weren’t you just arguing that the model shows it’s not random? Yet when I point out that the model assumes randomness, you say the model isn’t important?
No. I don’t think so.
But in Brownian motion, it’s individual motions that matter, not averaged-out ones. So we can’t turn to any statistical solution to average out the particle path and expect it to be descriptive. We turn to random walks instead.
Valid and true - the individual Brownian particle path is stochastic - that’s why the only way to model it is through randomness, or a facsimile thereof.
The fact that you can cover up the stochastic nature of reality by averaging out using statistical mechanics doesn’t erase the stochastic nature of reality, it just ignores it.
Actually, by the time we’re talking electrostatic forces and the like, we’re well off the “classic” path. Again, we start dealing with equations that have all these "assumptions built into them - “assume an infinite field”, “assume no gravity” etc, etc. The point is that the mere existence of messy reality is not accounted for by the neat equations of classical mechanics.
It’s all very well to talk about all of this as though it were definite, but there’s a reason that the fact that we can’t even prove that there is a solution to fluid dynamics is a Millennium Problem. Really, if you can, in fact, prove that there are always solutions to the Navier-Stokes equations, you’re short a million dollars.
But you can’t, so you take it on faith that everything is reducible to the equations. But you can’t prove it.
You can’t point to the two end-points and tell me it’s settled. Not when the Coefficent of Restitution is not fixed for non-homogenous substances and there’s all sorts of problems with it dependent on material composition.
…and many other articles on physical topics point to Brownian motion and just call it random. But you don’t see me citing that as evidence, do you, because that would be weak. No, I’m looking at the equation - if it were only “apparent”, why is there a factor for it?
I never said it wasn’t. It is also just an approximation of the truth.
So, the ones that are idealized abstractions, then/ OK.
I disagree. It matters very much to free will.
FTL communication is impossible. You know that. And how is the EPR paradox relevant? I’m not arguing for local reality here, despite what it looks like. I’m arguing for the inadvisability of trying to read the future. I’m arguing for probabilistic causation.
There’s such a model right on the wiki page I linked to earlier – this one. Pretend you don’t see the blue dots, and the red ones will trace out Brownian paths. Which you can then model using random walks, of course, because it simplifies the description. That’s all there is to it.
Generally, in statistical physics, one starts with an exact description of the microphysical degrees of freedom, which one then approximates – and in the process of this approximation, one introduces probabilistic models in order to be able to handle the computations. Say, you start with the Hamiltonian for N interacting particles, which is something like H = sum(p_i^2/2m + V), where the first term is the kinetic energy of each particle, and the second is some general interaction potential. This is a highly general ansatz – basically, any interaction you could wish for can be realized through some appropriate choice of potential. It’s also usually taken to be completely deterministic: knowledge of H and the boundary conditions allows in principle the exact calculation of the system’s dynamics, subject to the usual problems of computation of course. You could of course try to insert some sort of potential that introduces indeterminism, but you’d have to get creative – most sensible functions will yield to Hamilton’s equations being uniquely solvable. But that’s beside the point.
The point is that there’s no reason for you to try and introduce any stochasticity at this level of description, as the formalism will reproduce apparently stochastic behaviour in the appropriate limit – if you choose to ‘forget’ something about the microdynamics. For instance, in order to get the behaviour of an ideal gas, you just assume that there’s basically no interaction between the gas particles – i.e. you set V to zero. Then, since you have an intractable amount of particles going at various speeds, you say, OK, I really don’t need to know about all that stuff, and you simply average over all of those speeds – which is where the statistical description is introduced. So, each of those particles will have on average the kinetic energy <E> = m<v^2>/2. From there, it’s essentially straightforward to deduce the ideal gas law: assume the gas is contained in a certain volume, calculate the force any particle exerts (on average!) on a wall of that volume, through elastic collisions, use equipartition, and you’re done, more or less. Thus, the deterministic microdynamics gives rise to a statistical description of the macrodynamics due to simplification – there’s nothing inherently statistical about the system at bottom anywhere.
The same basic strategy can be employed for Brownian motion. All you have to do is to embed some particle into your ideal gas, and make sure that it’s small enough so the impacts don’t average out – for instance, assume that only one particle hits it in a given interval (on average!). Assume it gets hit uniformly from every direction. And you’re basically good to go. Note that nothing’s changed about the microdynamics – they’re still essentially given by the Hamiltonian above, and thus, are deterministic. As said before, the statistical nature of the process enters purely through simplifications and approximations – which is just sensible, since the full microscopic state of the system never enters into everyday considerations anyway. There’s no point in knowing the precise direction and velocity of every particle when you can make equally good (i.e. empirically verifiable) predictions just assuming a Maxwell-Boltzmann distribution (which the particles will follow – on average).
What do you mean by that? Electromagnetism is a perfectly classical theory in every sense I know to use that word.
Whether the coefficient is fixed or variable is no problem at all for the deterministic nature of the equations. And there’s no middle I have excluded: varying that very coefficient takes care of all cases between elastic and inelastic collisions.
I really don’t know what you’re talking about here.
But we’re just talking about determinism – whether or not I know what happened on the Sun 5 minutes ago has no bearing on what actually did happen.
Doesn’t matter for the thought experiment.
If both ‘ends’ of the EPR setups I detailed in the post before the last one are not simultaneously in existence, then there would seem to be no way to explain how they got matching measurement results; but if they are, then there’s determinism.
(Incidentally, I’d be very surprised if Einstein considered the microdynamics of Brownian motion – as opposed to its macroscopic, or perhaps mesoscopic, description – in any way random, or non-deterministic; I mean, that’s the guy who famously thought that even behind quantum mechanics – and thus, of course, behind all dynamics in its classical limit – we would eventually find a deterministic theory…)
Oh, and I just found a better visualization of Brownian motion emerging from completely deterministic microdynamics.
Sorry for the multi-posting, but I’d completely forgotten my resolve to ask every proponent of indeterminism how it could possibly be that an undetermined choice ever gets made – or, as I put it earlier:
Of course they are “caused by something”. Feelings are a function of the brain, there’s nothing magic about them. The same goes for your decision making processes.
In order to talk about whether a system is deterministic, we need a state space, and a evolution parameter. In a Newtonian universe, the state space would describe the positions and velocities of all the particles in the universe. In modern physics, the state space is more complex, involving wave functions and other such things. In a movie, the state space describes the colors in a two-dimensional array of pixels. In our universe, the evolution parameter is time. In a film, the evolution parameter is the frame number. Determinism is the property of a system that, given complete knowledge of all* “past” states, you can determine the next (“future”) state.
Under this definition, a film is clearly indeterministic. It’s not indeterminate on account of being open. Once a film is produced, it doesn’t interact with the world around it. Nor does the “history**” of making the film matter, since the process of making it cannot be represented in the state space. So, do you disagree with my definition of “determinism”?
If you want to change the definition of determinism to include requirements on how the system was created, you’re going to run into problems. Under my definition, Conway’s Game of Life is a deterministic system. If, however, we need to know how the game was created, then we can’t say whether it’s deterministic without also knowing whether Conway was deterministic. Consequently, we can’t know whether any system is deterministic until we know whether the entire universe is deterministic. Or worse, physicists may find several models of physics consistent with reality, some of which are deterministic and others that aren’t. If that turns out to be true, the question of whether any closed system is deterministic would be unanswerable.
If you still want to include requirements about how a system came to be, realize that these requirements will break the chain of causation, destroying the argument that free will is incompatible with determinism.
I don’t know what the mechanism is. If you figure it out there’s a Nobel, or maybe a Field’s Medal in it for you.
*A slightly weaker version of determinism would only require knowledge of a single previous state.
**In quotes because “history” refers to time, and not to frame count. Time is not the appropriate evolution parameter for a film. The question of whether a film was deterministic still makes sense if you lay the frames out side-by-side, rather than flash them on a screen one at a time.
You mean the blue dots that are treating the environment as a closed system (ignoring bbr), and whose kinetic energy is conserved “on average”? That totally abstract model? Yes, I see it. It’s irrelevant to the real world.
I’m exactly the opposite. I am more than willing to accept “it just happens”. Causality is one of those “lies-to-children” we tell ourselves to make the universe make sense, while all around us the Universe literally seethes with uncaused phenomena - every alpha particle sleeting through the air from my granite countertop, the electrons released by the C[sub]14[/sub]this desk, the neutrinos that flit through me like I wasn’t there. All we can do is average it out to control it, or point to penultimate causes. But no amount of billiard balls knocking together will ever tell you when a particular uranium atom will break down, ever. I rank this ability for simplifying things up there with the Gambler’s Fallacy or optical illusions, as one of those things just built into our brain.
And it’s determinism that leads to God (or infinite regress, if you’re OK with that, I guess), not indeterminism. For if everything has a cause…you know how this goes;)
He may have believed* in determinism, but he used randomness to describe it.
*(I’m not questioning this, I know he did - it’s the other usage of “may” - contingent?)
Once again you’re ignoring that the map is not the territory. It’s a pretty picture, but it ignores a lot of what actually goes on in reality. Again, reality is modelled as a nice closed box with perfect classical collisions. That’s not real. And what it ends up with looks like Brownian motion. But it’s not true Brownian motion any more than afractal tree is a real tree.
To put it another way, for entities with no free will, there is no particular reason that their actions ought to correspond with any sort of desires, or even that they ought to have any such thing as “desires”. But our minds nevertheless preface our actions with a conscious awareness of a desire to act. We may be just responding to stimuli as dictated by a predetermined program, but if so it’s a program which simulates an act of choice.
Well, you can draw up a box around every system and look only at what goes on within that box – for instance, you can describe a falling basketball as a succession of heights that suddenly and indeterministically inverts its direction; but the indeterminism only comes into play because you’ve failed to include the ground into your model. By this reasoning, computers also could come up with truly random numbers.
No, I don’t think so. There’s an asymmetry here: if you can draw a box around your system such that everything within it is describable deterministically, you know you have a closed system – since any influence from the outside would show up as indeterminable. So anything that can be embedded within a deterministic system isn’t fundamentally indeterministic in itself, but only due to its openness. The film can be thus embedded, and so can the sequence of dice throws.
That’s actually pretty much the situation we’re in now: under some readings, quantum mechanics is deterministic, under others, it isn’t. Both possibilities are consistent with reality; however, since I find determinism to be the more parsimonious explanation, it’s the one I think ought to be assumed.
Well, I don’t think there is such a mechanism, and in fact, that the idea of a mechanism is self-contradictory: if there is a mechanism choosing, say, A over B, then this determines the outcome; if there is no mechanism, then neither can happen. Basically, indeterminism proposes something analogue to underdetermined equations having definite solutions; like saying x + y = 5 can be solved without further information, and the solution is (say) x = 3, y = 2. If there is any way to choose that solution over the infinitely many others, then there’s enough information to solve the equation; if there isn’t, then there’s no way to call this ‘the’ solution.
No, it’s not. It reproduces real world phenomenology in a deterministic way; yes, it’s not a realistic model, but it demonstrates that there is no good reason to assume fundamental randomness. Really, the point of my lengthy argumentation above was not to show that there can’t be randomness on the fundamental level, but merely that all observations can be accounted for without assuming there is; and in practice, that’s the way it’s done. You go from deterministic microdynamics and approximate to your heart’s content, and you get the ideal gas, or the van der Waals gas, or Brownian motion, and all that. To assume randomness is just a violation of parsimony; you could as well assume that god did it, or tiny little fairies push the particles around.
That assumption doesn’t enter into electrodynamics, of course.
This, on the other hand, is just the macroscopic formulation, in which the usual approximations – averaging, assuming things behave probabilistically etc. – for macrodynamics feature; the microscopic Maxwell equations describe classical electromagnetism exactly.
The thing is that ‘it just happens’ makes no sense. If it happens and there’s a way to make it happen, then it’s deterministic; if there’s no way to make it happen, it doesn’t happen.
Just like, since every mammal’s mother is a mammal, there must have been infinitely many mammals – you know how this goes.
(Less facetiously, I do of course recognize that this is a problem, but to my mind, not nearly as insurmountable as the problems indeterminism faces – but going into plausible resolutions for this would take us too far afield.)
Yes, as a means of conveniently ignoring – averaging out – irrelevant microdynamics.
Well, this of course goes somewhat into larger issues of mathematical modelling of reality, and physical theories in general. You can always claim that an equation is not a physical system, but this does not give you license to claim that therefore, the physical system behaves ‘actually’ differently, and much less reason to assert that since the deterministic modelling is only approximate, therefore the real world is indeterministic. In fact, if you wish to argue that the mathematical modelling doesn’t tell us anything about reality, then you have no grounds to make any particular claim about reality, lacking any way to model it.
And the Hamiltonian I posted earlier is sufficiently general to cope with any kind of interactions you might come up with, and the dynamics it encodes remain deterministic as long as Hamilton’s equations are at least numerically uniquely solvable. It’s from this Hamiltonian that all ‘probabilistic’ descriptions are ultimately abstracted – true, you won’t write it down that way every single time, and may start with things like ‘let’s assume a random distribution of particle movements’, but the only thing that gives you license to do so is that for sufficiently many particles, deterministic particle movements are well approximated by certain random distributions.
I understand that an open system is automatically indeterminate. Deterministic implies closed, and conversely, open implies indeterministic. This I agree with. What you’re not getting is that a film is a closed system. Or, if you insist on describing it as open, then it is open in precisely the same way that the Game of Life, or the universe itself, is open: because we don’t have a complete account of how it came to be.
All three of my examples (the universe, the game of life, a film) have an evolution parameter (time, iteration number, frame number, respectively). All three have an initial state (big bang, initial distribution of “living” cells, first frame). None of them have had their initial states fully explained. As far as we can tell, none of them get tampered with after they’ve been completed (God doesn’t intervene in the universe, the programmer doesn’t change the game of life as its running, no one secretly tampers with the frames as you’re watching a movie).
Maybe it would be helpful to compare two films. Let’s compare a movie, say (a digital copy of) Ben-Hur, with a recording of a run of the Game of Life (RGOL). If it helps, imagine that there is an entire crew of people involved in making RGOL, who had to set up the initial state just right to play out the way they wanted it to. Imagine that both films are rolled out so that their frames are side by side, and are then put under a pane of unbreakable glass, making it is impossible to tamper with either film.
The Game of Life is deterministic, so RGOL must also be deterministic as well, and therefore RGOL is a closed system. Ben-Hur is not deterministic. Is Ben-Hur open or closed?
The way I see it, Ben-Hur must be closed. Both Ben-Hur and RGOL have the same evolution parameter, and draw from the same state space (two-dimensional arrays of grayscale pixels). Both had teams of creative people looking to produce just the right shots. Both are protected from being tampered with. The only significant difference I see is whether the films are deterministic or not.
I imagine that you’ll argue that Ben-Hur is open. Why is it open? If you think it is open because it is indeterministic, then you need to support this. No one has argued–or even stated–the proposition that closed implies deterministic (all closed systems are deterministic). I don’t see why this would be true, given my definition of determinism. If there is a different property that makes Ben-Hur open but RGOL closed, what is it?
If, on the other hand, you wish to argue that RGOL is open, then you must conclude that RGOL is indeterministic. Thus my definition of determinism must be wrong. What would a correct definition be?
Of course they are “caused by something”. Feelings are a function of the brain, there’s nothing magic about them. The same goes for your decision making processes.
[/quote]
Feelings are no more a “function of the brain” than they are a “function of oxygen”.
I actually have no idea what you’re attempting to assert here.
Actually, no, this is the Hard Problem of your brand of superdeterminism. It doesn’t take us too far afield, it is the field. If you can’t answer the First Cause question properly, you don’t really have standing to claim parsimony for your side.
As you guessed, I disagree. A film is a particular part of a system, the system that surrounded the events of the making of the film. It’s what you get if you leave out stuff from that system – the direction, for instance, which to a large extent supplies the deterministic impetus for what we later see as ‘the film’. Cutting gets rid of other parts, and some parts may just get lost; some, like a score, get added later. Out of all those interactions, ultimately a part are recorded, in a deterministic fashion, and from this recording, the appropriate equipment can recreate, again deterministically, a facsimile of this particular part of the system.
In other words, for every scene in the film, there exists sufficient reason – this wouldn’t be the case for a truly indeterministic system. A film is just like viewing a part of a system through a window, so to speak – that this viewing is delayed in time doesn’t change the fact that it is just a part of a system.
You could just as well cut up your recording of the Game of Life, leave out frames, excise parts of the grid, etc., and call that an indeterministic system – but of course, it hasn’t actually become one; you’ve merely restricted what part of the system can be seen.
Eh, logically, an infinite universe is perfectly consistent, as is a circular ‘region’ of time, or an entity whose possibility entails its necessity, or the proposal recently repopularized by Hawking that there still is ‘nothing’ in a way, so there has never been a transition from ‘nothing’ to ‘something’, which therefore needn’t have been caused, or the idea that quantum theory is actually a logical necessity, which then segues into Vilenkin-style quantum creation, or Russell Standish’s information ontology (about which I’ve talked a little here), Vic Stenger’s idea that physics ultimately derives from symmetry, and since ‘nothing’ is perfectly symmetric (i.e. invariant under all imaginable transformations), nothing simply entails physics, which is really just like Standish’s view portrayed a little differently, or of course a timeline that is finite, but without first point, such as the open interval (0,1] contains no smallest number, in which similarly no first cause exists, and probably dozens of other proposals I don’t even know about. Of course, one can quibble with each on various grounds, but as resolutions to the ‘first cause’ problem, they’re all at least plausible; which is more than I can claim to have heard regarding the question how undetermined events occur.
This thread has turned into a bit of a mess.
Firstly some people seem to be confusing determinism with fatalism.
Secondly, there was no reason to bring free will into this thread. Note that whether determinism is compatible with free will is a big discussion topic in itself. It just muddies the waters here.
I suppose rather than just threadshitting, I should also add my position.
Regarding determinism, we don’t know whether the universe is deterministic, but several aspects of quantum mechanics imply that it is not. That even if you could have perfect knowledge of the current state of the universe, you could not predict the future state of the universe, with perfect accuracy, for instance.
Regarding free will, I think the very concept is incoherent. It’s never defined fully, and it doesn’t make sense. Why would I choose differently in the same situation given the same knowledge, memories and disposition?
But, since free will doesn’t even make sense, the lack of it doesn’t bother me.
And you disagree in a way that doesn’t clarify anything. Here are some questions:
[ol]
[li]Is RGOL an open system?[/li]I assume your answer is no, but that cutting the film will make RGOL open. In that case:
[li]Suppose I cut off the last two frames. Is RGOL now open?[/li][li]Suppose I cut some things out of the middle, but then change my mind and tape it back in. Is RGOL now open?[/li][li]Suppose that RGOL eventually becomes periodic. I remove exactly 1 period somewhere from the middle of the film. Is RGOL now open?[/li][/ol]
