Looking for explanation of randomness.

[moving finger]

I think choice in this context ought to be taken to mean - the possibility of having more than one outcome for the same initial conditions.

Say we have two options on the dinner menu, chicken and beef.

In a deterministic world this choice is pre-programmed. We would pick one over the other no matter what.

In a world where we have freedom of choice, either outcome is possible.

This definition of choice shares many parallels with the definition of randomness. The difficulty people have with the notion of choice is that they assume that there must be an underlying reason for choosing, say the chicken over the beef. If there is a pre-programmed REASON for it, how can it be CHOICE? But who is to say that a reason is needed to make a choice? Photons and electrons certainly don't need a reason when they 'choose' states.

Claude.

How do you make your choice between chicken or beef - do you do it deterministically, or do you do it non-deterministically?

If non-deterministically, what's the difference between making a choice and flipping a coin?

Best Regards

 

[moving finger]

Ok, we'll have to go into the definition of "choice" then. You can indeed say that apples always "choose" to fall to the earth, even though they "could", if they wanted to, rise up to the moon, but, for some reason or another, they simply don't make use of that possibility, otherwise entirely open to them.

I hope you are not putting this forward as a serious definition of choice.

A choice is made when an agent is able to evaluate the relative consequences of two or more logically alternative courses of action, to decide (based on comparison against a set of internal values which it owns as “its value system”) which course of action is the best, and then to determine which course of action occurs. This is precisely what humans do all the time, it is also what a deterministic machine can do, it is also entirely compatible with determinism. It’s a deterministic choice. And no mention of free will or non-determinism anywhere.

An apple falling from a tree does not choose in the sense described above, because an apple (as far as we know) does not evaluate the consequences of different ways of falling, does not decide which way to fall, and does not thereby determine which route to follow in its falling. (the option of rising up to the moon would not be open to them, any more than it's open to a human being, even if apples did make choices about which way to fall).

It is already determined since about 15 billion years. Of course it is not (purely) external, but "internal" if you want. Nevertheless, you have no real "choice".

Again, you are assuming choice is defined as free will choice.– which is why you cannot accept that a deterministic machine makes a choice. I do not agree with your definition (basically because I do not believe free will choices exist, thus the entire concept of such a choice would be meaningless). Perhaps you would also say that no deterministic machine can make “decisions”, nor make “selections”?

There are such things as deterministic choices, decisions and selections. There are also (allegedly) such things as free will choices, decisions and selections (though I believe the latter are illusions). You seem to believe that all choices (and decisions and selections?) must by definition be free will choices (and decisions and selections?).

In any deterministic or non-deterministic world, alternatives exist in a counterfactual sense – “if X had wanted Y instead of Z, then X could have chosen Y instead of Z”. The whole point is that freedom is defined as : I act according to my will. And this is determinism. Whether the alternatives exist in a real sense or in a counterfactual sense is not relevant, what is relevant is that I act according to my will – and that is freedom in anyone’s language.

When we look at the supposed free-will explanation of choice – how does it work? How does adding non-determinism into the selection process suddenly generate a free will choice where there was no free will choice before? Having “freely” chosen X, for example, given a particular set of antecedent states, why would your choice be any different (what logical or rational reason could you provide for your choice being any different) if you could repeat that exact same decision again under identical antecedent states? The only rational reason it might turn out any differently is if there is indeed a random element in the selection process – but are you seriously suggesting this randomness is the source of our so-called free will?

Your only "choice" is to "desire" what will actually, unavoidably, happen. Like the train, who "desires" to go exactly the way the track is laid out.

Again this confuses fatalism with determinism. It is not the case that I am first forced to act in certain ways, and only then am I subsequently forced to want to act in those ways. My wishes/wants/desires determine my subsequent actions, not the other way about. I am not constrained to act in a certain way against my will; instead my will determines how I act. Yes it is true under determinism that if I do X then it was always going to be the case, ever since the Big Bang, that I would do X (because it was always going to be the case that I would want to do X) – but so what? The important thing is that doing X is what I want to do at the time – I am not forced against my will – and doing what I want to do is freedom pure and simple.

Saying it is unavoidable leads to misunderstanding – because there are two types of unavoidable and many people get them confused. There is the “fatalism” type of unavoidable (where “what is going to happen is going to happen, whether you like it or not!”), but such things are not necessarily unavoidable under determinism (where “what is going to happen only happens because you want it to happen”). If I want to avoid something, and I am physically unconstrained, then I can avoid it, even under determinism. What, then, do you mean by unavoidable in this context – can you explain? What is unavoidable under determinism is that my will determines my actions – I do just what I want to do – but this is freedom! – and the same applies to the alleged “free will” case!

How is it any different under your free will scheme? You claim that you have free will, that you do what you want to do, thus it follows that it is unavoidable that you do what you want to do. “Ah” you say “but I could choose to do something different”. The same is true under determinism – if I had chosen to do something different then I would have done something different. But under both “free will” and determinism what rational reason would one have for choosing to do otherwise than what one actually did choose to do given that the antecedent states would be identical?

The free will advocates would have us believe that there must be some mystical but unidentified and mysterious introduction of non-determinism at some point – but nobody has any idea of how this would work, or how this could magically turn a supposedly non-free will act into a free will act. Introducing non-determinism simply means that we lose control over what we want to happen – non-determinism does not enhance our ability to make things happen according to our will, it destroys it! The very notion is simply incoherent.

Non-determinism means that the future is “open”, but how does this create any kind of “free will choice” (in any real meaning of the word) where none existed before? If a deterministic machine, by your definition, cannot make a choice, then by what mechanism are we able to magically create the conditions for a choice by simply adding non-determinism into the equation? How exactly does this work? If a deterministic choice is (by your definition) not a choice, then how is it that a non-deterministic choice is a choice? If my choice is non-deterministic then by definition I have no control over the choice – is that supposed to be a genuinely rational and reasoned choice when a deterministic choice, over which I DO have control, is not?

The free will believer is stuck between the horns of a dilemma. Either determinism, or non-determinism. Determinism doesn’t look too attractive – after all it’s determined! To the free will believer, therefore, there is no choice but to plump for non-determinism, not really knowing whether it offers any escape, but at least hoping that it cannot be any worse than determinism. It is inevitable that the free will believer believes the magic potion of free will is buried somewhere within non-determinism – he has no choice!

You perhaps do not like the idea that your will is determined. But what choice do you have? If your will is not determined, then what is it? If you believe a determined will is not “free”, then how on Earth do you think adding a splash of non-determinism to the mix makes it “free”? The free will believer is taking a leap in the dark, feeling that determinism is unacceptable and hoping that the perceived open door of non-determinism offers some means of escape. Escape to what? In what possible sense can a non-deterministic will be more “free” than a deterministic will? Simply because it avoids determinism? But how is avoidance of determinism rationally equated with freedom of will?

The free will believer is clutching at straws, not able to cling to the raft of determinism, and hoping against hope that there may be some solution, some salvation, but he knows not what form that might take, in the unknown mystical influences of non-determinism. The very notion is irrational and incoherent, but the free will believer has no other choice – irrationality and incoherency must be more comforting than the evils of determinism.

Well, no, the tracks are there already since 15 billion years in a purely deterministic view. You will only "pick" your actions so that they are compatible with what had to happen anyways (and which anybody with sufficient computing power and knowledge of initial conditions could have predicted, even before you were even born).

Once again it seems you are describing fatalism. I do not pick my actions so that they are compatible with what subsequently happens – this explanation has cause and effect back to front – it is basically saying that what “must happen” in the future somehow determines my actions toady, outside of my control, rather than the other way about. This is totally wrong – it is the fatalist belief.

I am the driver of my train, and the train goes where I want it to go – my will is not determined by the tracks and terrain ahead of me, but instead those tracks and terrain are determined by my will. All perfectly compatible with determinism. What possible purpose would be served by introducing non-determinism, so that the train might go where I did NOT want it to go?

Right - and deterministic machines make choices between alternatives all the time. Again, don't assume that choice entails free will choice.

Well, I don't see how you can call that "making choices between alternatives", in as far that there are no alternatives.
If V = 5, and the machine has, in its memory, an instruction:
If V > 4, then do A, else do B
then B is not a possible alternative to the machine, given this instruction. Only A can happen. So there are no alternatives, and hence no choices.

Again you insist that all choice must be free will choice – you do not acknowledge the existence of deterministic choice. What about random choice – would you accept that such a thing as random choice exists? If yes, then why not deterministic choice?

You claim the machine is not making a choice, because it is acting deterministically. Where does the ability to make a choice come from then – from non-determinism? How can adding non-determinism into the machine’s decision-making process turn a non-choice into a choice? How does it work? How can{/b] it work? It’s incoherent.

What if the machine has developed its own decision-making algorithm, based on many years of experience of making decisions and evaluating the outcomes? Just like a human baby, a machine need not be fully pre-programmed, it can learn which decisions are better than others through experiment and trial and error, and as it learns it can modify its decision-making algorithms accordingly. The decision-making algorithms then eventually (after much learning and experience) become “owned” by the machine, the machine takes ownership for its decision making processes because it accepts that these processes allow it to act reasonably and to make rational decisions in accordance with the way that it thinks it ought to make decisions and act – and this is just the same way that humans learn how to make responsible decisions. The difference in human agents is that our fundamental decision-making processes are buried so deep within the complexity of our brains that we really have no idea exactly how we make the decisions that we do – and this allows us to cling to the illusion that we act free from determinism. We believe that there is no deterministic algorithm in there that is making these decisions - but then how are they made? Non-deterministically? How does that work? Keep the issue shrouded in darkness and mystery and you can delude yourself that you are not acting deterministically – but ignoring the question does not make your belief in free will a sound belief.

How else do you think you make choices? Given a morally responsible decision, you weigh up possible consequences of different possible courses of action, and you evaluate them internally against a set of moral values or standards (your moral and ethical beliefs), and then you determine which course of action is the best according to your moral values. Are you claiming that we need to deliberately somehow inject a random or non-deterministic element into our otherwise deterministic morally responsible decision-making process in order to claim that our decisions are morally responsible decisions? How would randomizing any part of a deterministic decision magically make it a free will choice, when it is not a free will choice before – can you explain how this notion is coherent?

Yes, but the point is, that even before you're even confronted to the question, or even before you exist, you could only want Y and do so, and not X, in a deterministic universe, because that follows from the initial state and the evolution equations (and was hence fixed for at least 15 billion years). The option "X" is simply not open to you, so you were NOT free to pick it, to desire it, and to do it.

If I had wanted to choose X, I could have chosen X. That is freedom. The fact is that I did not want to choose X, and the reason X is not open to me is because I did not want to choose X, not the other way around.

It is NOT the case that “X is not open to me, therefore this causes me to not want X” (that would be a kind of extreme fatalism), instead it is the case that “I do not want X, therefore this causes me to not choose X”.

The counterfactual is always logically there – if I had wanted X then I could have chosen X, but the simple fact is that I did not want X.

What you are focusing on is the fact that my will is determined – and this leads you to claim that I am not free.
What I am focusing on is the fact that I determine my actions according to my will – and this leads me to claim that I am free.

Both are correct, because determinism entails both. My will is determined, and I determine my actions according to my will. You call that lack of freedom, and I call it freedom.

Why do I call it freedom? Because freedom IS the ability to do what you want to do.
Why do I reject your notion of free will? Because it is irrational and incoherent – free will rests on a naive intuitive desire to think that we are somehow free of determinism, and that there is some form of magical property inherent in non-determinism which can give us our free will - but free will avoids, rather than addresses, the questions of logical reality.

The point is that at the time of my decision there is absolutely nothing external to me which is forcing me to choose one way or another. Whether I choose X or Y may indeed be determined since the Big Bang (if determinism is true then my choice indeed is determined), but the point is that the only reason that I choose Y over X is because, at that moment, I want to choose Y. This is all completely deterministic – what possible advantage would I have by adding in randomness or non-determinism, what possible benefit could such randomness provide to me, since what happens already happens because I want it to happen? The only difference that randomizing any of this might have is that I end up doing something that I did NOT want to do - why on Earth would I rationally want to do something which I did not want to do?

If you believe that this notion of free will adds anything useful to our ability to act according to our will, could you explain how it is supposed to work?

Positing some nebulous or dualistic soul doesn’t solve anything – it simply moves the logical problem from the physical domain to the non-physical - in the end we still need to decide whether the source of our will is determined or not determined. Perhaps free will believers hope that by shifting the problem to the non-physical they can avoid addressing the issues of rationality and coherency, and therefore ignore them? I don’t see how having a random source of will endows freedom!

You only had the *illusion* that the option X was open to you but that you deliberately choose not to pick it. Just as the machine might have the illusion that it could, if it wanted, do B, but "picked" A. But the instruction in its memory made it pick A and it could never have picked B in the first place. In the same way as you could not have opted for X.

And I would say you only have the illusion that you possesses free will.

The option X is logically open to me in a counterfactual sense – if I had wanted to do X then I could have done X. That is not an illusion, it is a logical truth. And that is exactly what freedom means! What other sense is there that has any rational meaning?

To say that “I could not have opted for X” is logically incorrect. Counterfactually, if I had wanted to opt for X then I would have opted for X. The fact is that in this particular physical world, I did not want to opt for X, therefore I did not opt for X, thus I am free. Simple as that.

Why on Earth would I wish to do what I did not want to do (which is what you seem to be advocating)? Given that I do not want to opt for X, and that I do what I want to do, where is the sense in then saying “you have no free will because given your wish not to opt for X, you cannot opt for X”? The very notion is irrational.

Yes, I agree with this. But so what? Much of what happens in my life happens becasue I want it to happen - what more freedom could I ask for than that? And this is completely compatible with determinism.

Entirely, but you could reformulate that as: much what happens in my life gives me the illusion that I wanted it to happen - what more freedom could I ask for then ?

What you have described is incoherent. By definition, “what I want to happen” is “what I want to happen”. How can I be under the illusion that “what I want to happen” is not “what I want to happen” (unless I’m somehow schizophrenic).?

The only way your so-called illusion could happen is if there are two “I”s, a kind of dualism or schizophrenia where the “real I” is the one which knows what I really want to happen, but it is being over-ruled by another “fake I” which latter is actually in control of my actions and which makes me (the real I) think that what it (the fake I) wants to happen is indeed what I (the real I) want to happen. Sorry, but such an idea is totally ludicrous.

Best Regards

 

[vanesch]

I hope you are not putting this forward as a serious definition of choice.

Nevertheless, that's the definition of choice *you* put forward, when you say that deterministic machines "evaluate, determine which course of action is best and then determine which course of action occurs".

After all, the assignment of values such as "best" etc... are just shifting around the notion we're trying to grab: free will. If you simply look at the compiled computer code in the memory of a deterministic machine, without knowing the intentions of the programmer, then you can determine that the machine will do this or that, in exactly the same way as you can write down the dynamical laws for an apple falling from a tree. Words such as "evaluate the situation, determining what is best" etc... are embellishments of simply saying: follows his deterministic dynamical laws.
Who knows ? Maybe there's a ghost in the apple, who "evaluates the situation" (calculates his lagrangian), and "determines what is best" (that is, tries to minimize its action), and hence, falls down, because the apple determined that as his "best course of action".
Any deterministic prescription of "determining the best course of action" is nothing else but other words for "follows his deterministic dynamical laws".

A choice is made when an agent is able to evaluate the relative consequences of two or more logically alternative courses of action, to decide (based on comparison against a set of internal values which it owns as “its value system”) which course of action is the best, and then to determine which course of action occurs.

It is also what an apple does. It's "value system" is "have minimal action", and his course of action is to fall down to earth. It feels also that it did exactly what it wanted to do.

An apple falling from a tree does not choose in the sense described above, because an apple (as far as we know) does not evaluate the consequences of different ways of falling, does not decide which way to fall, and does not thereby determine which route to follow in its falling. (the option of rising up to the moon would not be open to them, any more than it's open to a human being, even if apples did make choices about which way to fall).

Well, you don't know any more that an apple is not evaluating its action and tries to minimize it. And as I said, if deterministic laws (call them evalution of the situation and deciding on the best course of action, if you want) determine indeed what you are supposed to decide, then any other option than the one prescribed by the deterministic laws is just as closed to you as falling up to the moon is to the apple.
It is just that you've given a fancy name to the dynamics in certain cases, such as "evaluation of the situation followed by a decision", but deep down, it just means: follow the deterministic dynamical laws.

Again, you are assuming choice is defined as free will choice.

What else could it be ? Unless you consider "choice" as certain forms of dynamical laws, such as "if then else" statements in computer code. But they are not more than dynamical laws of course.

– which is why you cannot accept that a deterministic machine makes a choice. I do not agree with your definition (basically because I do not believe free will choices exist, thus the entire concept of such a choice would be meaningless). Perhaps you would also say that no deterministic machine can make “decisions”, nor make “selections”?

I'm not sure free will exists either (unless, we redefine it as a specific form of dynamics - which kills the essence of the concept). But we're trying to see what it would imply if it existed, and in how much it is compatible with a deterministic physics. I claim that it is not compatible.

There are such things as deterministic choices, decisions and selections.

Well, to me, that's just fancy words to describe dynamics, or laws of nature. From what point on does a dynamical rule become a "choice, decision and selection" when the "intentions" are not known, and in what way is this different from an apple "choosing, deciding" to minimize its action and "select" the movement that corresponds to this "choice" ?

There are also (allegedly) such things as free will choices, decisions and selections (though I believe the latter are illusions). You seem to believe that all choices (and decisions and selections?) must by definition be free will choices (and decisions and selections?).

I think it is only in that sense that there is "choice". If you just execute an algorithm, differential equation, minimum action principle or what ever, I don't see what "choice" there is, given that there are no allowed alternatives than what the algorithm/differential equation/minimum action principle is prescribing.

In any deterministic or non-deterministic world, alternatives exist in a counterfactual sense – “if X had wanted Y instead of Z, then X could have chosen Y instead of Z”. The whole point is that freedom is defined as : I act according to my will. And this is determinism. Whether the alternatives exist in a real sense or in a counterfactual sense is not relevant, what is relevant is that I act according to my will – and that is freedom in anyone’s language.

But the "trajectory falling up" for an apple is also an "alternative in the counterfactual sense", meaning: it was a possibility (in the space of possible trajectories) but which was not picked out by a minimum-action principle. And if we define the "will" of an apple to minimize its action, then an apple also acts according to its "will". The alternatives have to exist in a real sense for there to be a genuine choice to be possible, no ? If the choice is prescribed, by an algorithm or any other deterministic principle, it's not a choice.

When we look at the supposed free-will explanation of choice – how does it work?

The very point of a free-will choice is that one cannot say how it works, of course ! Because otherwise we would be back to case 1.

The only rational reason it might turn out any differently is if there is indeed a random element in the selection process – but are you seriously suggesting this randomness is the source of our so-called free will?

To come back to the original poster's question, randomness means "ignorance about which to pick". When one says that something is random, one simply tells you that there are different possibilities, and that one doesn't have a rule to say - even in principle - which of the possibilities should come out. A choice can hence only be random to an entity that is not making the choice, because the entity making the choice is of course not ignorant of its very choice.
However, whether that entity should follow any deterministic rules to arrive at its choice is what is debated here. From the moment that there are such rules, there's no real choice anymore ! So a genuine choice doesn't obey any rules.

Again this confuses fatalism with determinism. It is not the case that I am first forced to act in certain ways, and only then am I subsequently forced to want to act in those ways. My wishes/wants/desires determine my subsequent actions, not the other way about.

But your wishes etc... are determined by dynamical prescriptions. Your wishes ARE dynamical prescriptions. The "wishes" of a deterministic machine are simply its computer code in its memory, and the "wish" of an apple is simply to minimise its action.

I am not constrained to act in a certain way against my will; instead my will determines how I act.

So your "will" is nothing else but the deterministic law itself, but just given a fancy name, in this case.

Yes it is true under determinism that if I do X then it was always going to be the case, ever since the Big Bang, that I would do X (because it was always going to be the case that I would want to do X) – but so what?

Yes, but you will have a hard time convincing me that your "will" is anything else but the dynamical rules themselves - over which you have no choice. Like the "will" of a deterministic machine is to execute its program.

Saying it is unavoidable leads to misunderstanding – because there are two types of unavoidable and many people get them confused. There is the “fatalism” type of unavoidable (where “what is going to happen is going to happen, whether you like it or not!”), but such things are not necessarily unavoidable under determinism (where “what is going to happen only happens because you want it to happen”). If I want to avoid something, and I am physically unconstrained, then I can avoid it, even under determinism.

Hell, you don't even have the possibility not to want to avoid it ! This possibility is not even open to you. Like the possibility not to follow its computer program is not open to a deterministic machine.

Your "will" is nothing else but the dynamical law you are going to have to follow in any case. There are no alternatives open to you, and if you think so, then that's the *illusion* of free will you have. "Desiring" something else than what is determined by the deterministic dynamics since the big bang, is not an option, in the same way as *not* minimizing its action is not an option to an apple. So in as much an apple has any "desire", it must be to minimize its action.

The same is true under determinism – if I had chosen to do something different then I would have done something different.

No, you couldn't. The option was not open to you, you could not have desired anything else than what was fixed since the big bang.

BTW, I'm not advocating free will, in fact in as much as I have to pick anything, I'd rather say that we do not have free will (but only the illusion). But if we are going to require "free will" I'm only indicating that, by definition, it is not going to follow any deterministic prescriptions, and that the laws of nature need to leave some room for it by being *ignorant* of what will happen - so leaving open more than one option from which to pick.

The free will advocates would have us believe that there must be some mystical but unidentified and mysterious introduction of non-determinism at some point – but nobody has any idea of how this would work, or how this could magically turn a supposedly non-free will act into a free will act.

As I said, it is essential for an entity exercising free will that it doesn't follow any dynamics, or rules. The randomness resides not in the exercise of free will from the PoV of the entity exercising it, but rather externally: externally, one has to remain fundamentally ignorant of the decision the free will entity will take, and hence needs to introduce an element of randomness (= ignorance) in any dynamical prescription of the event.

Non-determinism means that the future is “open”, but how does this create any kind of “free will choice” (in any real meaning of the word) where none existed before? If a deterministic machine, by your definition, cannot make a choice, then by what mechanism are we able to magically create the conditions for a choice by simply adding non-determinism into the equation? How exactly does this work?

Well, the essence of the concept is that we cannot say how it works, because from the moment we can, we destroy the very concept of free choice !

It is inevitable that the free will believer believes the magic potion of free will is buried somewhere within non-determinism – he has no choice!

Yes, that was my point. Determinism doesn't allow for free will. Non-determinism doesn't provide it, but is at least a necessary condition even to start contemplating its possibility. It leaves the possibility, because the laws of nature leave open several potential alternatives (in contrast to deterministic laws). In order to have a free choice, one needs several alternatives to pick from in the first place.

The second point is that, in order not to destroy the very possibility opened up by having now different potential alternatives, the entity exercising free will cannot, itself, be subject to any regularities of laws itself (because these regularities or laws would then close the door that was opened by the non-deterministic laws in the first place). So it remains a fundamentally mystical concept.

But, again, I'm not an advocate of free will. I'm only putting down the consequences of the concept, that's all.

The very notion is irrational and incoherent, but the free will believer has no other choice – irrationality and incoherency must be more comforting than the evils of determinism.

No, not at all. Free will would simply imply that there are limits to the rational description of nature, that's all. That some things are fundamentally underscribable.

I am the driver of my train, and the train goes where I want it to go – my will is not determined by the tracks and terrain ahead of me, but instead those tracks and terrain are determined by my will. All perfectly compatible with determinism. What possible purpose would be served by introducing non-determinism, so that the train might go where I did NOT want it to go?

The tracks are nothing else but the deterministic laws of nature, and what you describe as your free will is nothing else but that same dynamics. You want the train to go where the track is, the track is where you want the train to go, and this is already determined since the big bang. Exactly as the apple wanting to fall on exactly the trajectory that is minimizing its action.

Again you insist that all choice must be free will choice – you do not acknowledge the existence of deterministic choice. What about random choice – would you accept that such a thing as random choice exists? If yes, then why not deterministic choice?

"random choice" is, as I said before, only random to the exterior of the entity exercising free will. Random means: no way to know which one. Clearly the entity making the choice knows which one: it picks it. The "random" part comes from the *non-existence* of any specific regularity of rule that could determine what the entity "has" to choose.

As I said, a free will entity must necessarily be undescribable by any form of dynamics. It's part of its essential being.

You claim the machine is not making a choice, because it is acting deterministically. Where does the ability to make a choice come from then – from non-determinism? How can adding non-determinism into the machine’s decision-making process turn a non-choice into a choice? How does it work? How can{/b] it work? It’s incoherent.

No, it is fundamentally undescribable, that's all. You can call it mystical. Yes, free will is an essentially mystical concept. Mystical in the sense of fundamentally undescribable.

 

[Mickey]

Randomness is just something that makes statistics work. That's all.

The minute statistics stops being useful, I don't think there will be too many deep questions about how randomness exists in nature.

When will that day come? Well we're already there for political polls. Give quantum physics another 30 or 40 years.

(Oh, and randomness and the question of free will are perhaps more connected than we might think... the new paradigm of AI research is now probabilistic instead of algorithmic).

 

[Hans de Vries]

What will it cost me to get this moved back to Philosophy?

It would be OK here if the starting point is that the human brain consists
out of physical things like neurons which can be described and modeled
and which are deterministic.

It should go to Philosophy however if people here insist on the existence of
indescribable meta-physics and/or religious objects like "souls" which are
supposed to work outside of any mainstream physics.Regards, Hans

 

[vanesch]

This definition of choice shares many parallels with the definition of randomness. The difficulty people have with the notion of choice is that they assume that there must be an underlying reason for choosing, say the chicken over the beef. If there is a pre-programmed REASON for it, how can it be CHOICE? But who is to say that a reason is needed to make a choice? Photons and electrons certainly don't need a reason when they 'choose' states.
.

Yes, exactly my point too. From the moment there is any reason/algorithm/dynamics/regularity that determines the "choice", it is no choice anymore in the full free sense of the word.
Hence, free choice implies postulating an entity of which the behaviour is entirely impossible to describe from the outside.

 

[vanesch]

It should go to Philosophy however if people here insist on the existence of
indescribable meta-physics and/or religious objects like "souls" which are
supposed to work outside of any mainstream physics.

Well, the importance for physics is simply the statement that got this discussion into the "free will" direction - but which is closely related to the original question of randomness: if free will is to exist somehow (which is not a necessity at all), then any purely deterministic theory of nature must somehow be wrong. This eliminates a huge class of theories.
As was pointed out in the discussion, however, our "impression of free will" is not sufficient to eliminate deterministic theories, because it might be based on an illusion.

 

[Farsight]

That's the point vanesch. Free will, and more properly consciousness is like heat, a derived effect, not real, an illusion. But it's the only thing that's real to you and me. You're just a bunch of moving atoms who think they can think, but because they do, they can. And they can choose to move other atoms. Because reality is like energy. It isn't something you can hold it in the palm of your hand. It's a property of a system, and it's relative.

 

[superweirdo]

Alright, earlier in this thread, somebody spoke about irreducible randomness. I searched it on internet, I couldn't find it. My believe was that everything can be described by natural laws but he said that there exists some things that are irreducible randomness. Now, first of all, what evidence is there that it exists, second, give me an example.

 

[vanesch]

Alright, earlier in this thread, somebody spoke about irreducible randomness. I searched it on internet, I couldn't find it. My believe was that everything can be described by natural laws but he said that there exists some things that are irreducible randomness. Now, first of all, what evidence is there that it exists, second, give me an example.

This brings us back on topic
I'm claiming that randomness is "several alternatives" or "one out of many" and no way to know which one. But this can be seen on different levels, and the main distinction here has been illustrated by a poster who emphasized, quite correctly, the difference between ontic and epistic randomness.
Mostly known/understood/used is epistic randomness. That means, there are several potential alternatives, because of lack of information on our side.
Because we have incomplete information, we are not able to say which, of several alternatives, will happen, is happening, happened. This is well described using probabilities indicating our degree of ignorance. It is the Bayesian view on probabilities. So it is sufficient to increase our knowledge, to decrease randomness. So randomness, from this PoV, is not a concept inherent in nature, but just inherent in an observer that doesn't possesses complete information about the situation and hence cannot discriminate between different alternatives.

However, it is conceivable (though of course not provable) that there is some ontic randomness - it is what I called irreducible randomness. What does that mean ? It means that nothing in nature determines the event to happen. That several alternatives are open to the laws of nature, and that there is nothing at all in the fundamental state of nature that can determine which, of the several alternatives, will actually happen.
It is of course irreducible randomness, because if *nature itself* doesn't "know" what will happen, then of course there's no hope for any observer to know !

We can have an intermediate case, where there are "hidden variables" ; there are hidden quantities in nature, which will determine what will happen, but for one or other reason, they are not available to observation. This means that there is some kind of irreducible epistic randomness, but it is not ontic randomness. Nature, "knowing" of these hidden variables, can determine precisely what will happen, but we can't, no matter how we try.

A theory that does not contain any ontic, irreducible randomness is called a deterministic theory. The laws of nature determine exactly what will happen, if the current (or past) state of nature is known. In such a universe, since the big bang, everything that happens and will happen is determined, and there are no alternatives possible.

If a theory has ontic randomness, that means that the past doesn't completely determine the future: certain alternatives are possible and there's nothing in nature that tells us which one of the alternatives is going to happen. Not simply because there's something that we ignore, but simply because the laws of nature do not determine it.

There's no discussion that classical physics is deterministic. The discussion is with quantum theory of course, which seems to be irreducibly random. However, it depends on how you interpret it to say whether the randomness is truly ontic, or whether the randomness is epistic, or whether, after all, there is no randomness.

Bohr's views seem to tend to irreducible randomness (ontic randomness). Bohmians follow Einstein, and say that it is only epistic randomness (we are simply ignorant of an underlying deterministic process).
Everettians (like me), say that the randomness is an illusion of subjective observation.

 

[MeJennifer]

This may perhaps be a weird comment (for some) but to me the strength and persistency of the argumentation in favor of "free will" is to me an indication of how strongly rooted this idée fixe is in the human mind.

 

[superweirdo]

ok... I am just going to repeat my question, I know what all those things are. What I can't find is any evidence that irreducable randomness exists and any example of it.

 

[MeJennifer]

ok... I am just going to repeat my question, I know what all those things are. What I can't find is any evidence that irreducable randomness exists and any example of it.

You can't find it because it is impossible to prove.

By the way, what is the difference between irreducable randomness and just plain randomness.

 

[Mickey]

Randomness that is in principle irreducible would, it seems, make randomness an ontological precept of nature.

Randomness that is only irreducible in practice would make randomness a provisional precept of nature.

Science, being provisional, would cover the latter, and the former would be covered by philosophy.

So, superweirdo, the way I gather this is that you're looking for evidence of an ontological principle of irreducible randomness. If so, science is a poor method for gathering such evidence, since science studies provisional claims.

If, on the other hand, you are looking for evidence of a randomness that is irreducible in practice, quantum mechanics is your thing. After all, the central role of randomness in quantum theory exists because it was developed so closely around its practice.

Roger Penrose believes that the greatest problem in physics is the question, "how are quantum probabilities converted to an actual, sharply well-defined outcome?" or what's called the measurement problem. Each interpretation of QM attempts to address it, and there are other musings like Wolfram's cellular automata, but so far there are no definite answers.

Right now, I personally don't think the philosophical question of whether nature has such a thing as principally irreducible randomness matters. I find that even approaching such questions seriously is too daunting for mortals, especially without having something like billions of years of experience with nature. Randomness just gives a working model for our practical considerations of certain fields at the moment, so questions in those fields are questions we can handle.

Anyone who doesn't like that randomness has such a central role in those models can work on developing non-probablistic theory that works better. Goodness knows everyone wants something that works better!

 

[vanesch]

Right now, I personally don't think the philosophical question of whether nature has such a thing as principally irreducible randomness matters. I find that even approaching such questions seriously is too daunting for mortals, especially without having something like billions of years of experience with nature. Randomness just gives a working model for our practical considerations of certain fields at the moment, so questions in those fields are questions we can handle.

Well, I would think it matters for people looking for potential deeper theories beyond the quantum theory we have today. Should they require determinism (at least ontic determinisim), or can they live with ontological randomness in the formulation of the "ultimate and final" theory they are going to formulate next week ?

Of course you're right that for the working physicist, who works *within* a certain paradigm, it doesn't matter at all. But for those trying to think up new paradigms, it might play a role.

 

[moving finger]

If it is established, by the ontological state of the (deterministic) universe at the time of Julius Ceasar, that you will, 15 years from now, plant a knife in the chest of your grandma, and that this is predictable with enough computing power, then what's still left of your choice to kill or not to kill your grandma ?

We assume here that if I kill my grandma then I do so because that is what I “will” to happen at the time, and not because I am forced by factors external to myself to kill her against my will.

My will is determined, yes, but it’s still my will. If I kill my grandma then I do it because I want to do it at the time. That’s freedom in anyone’s language.

The only thing we can conclude then, is that 15 years from now, you will have a desire to kill your grandma and act accordingly. Or that you will have no such desire but do it anyways. But it will happen, no matter what. There is no alternative in a deterministic universe.

No, not “no matter what”. In a logical counterfactual sense, if I did not want to kill my grandma then it follows that I would not kill my grandma. There exist logically possible worlds (where the antecedent conditions are different to our world) where I do not want to kill my grandma, and in those worlds I do not kill her. If I kill my grandma in this world, then the killing of my grandma occurs only because I want it to occur. Mine is a determined will, yes, but what happens happens only because I want it to happen at the time. That’s freedom.

The "randomness" I talked about was not in the "choice" but in the laws of nature, leaving open the way to really pick amongst alternatives, left open by the laws of nature.

This is incoherent. How does one pick the laws of nature? What logical process could carry this out to give a controlled and desired result, unless it was a deterministic process?

In that, at the time of Julius Ceasar, given the laws of nature (in a stochastic universe) and its ontic state back then, both the possibility that you will kill your grandma, and that you won't, are possibilities left open (and hence, "random"). If that's the case, there is *room* to say that something (outside of the laws of nature of course), such as your "free will", might ultimately decide to kill her or not.

You haven’t thought this through to completion – the explanation as given is incomplete and therefore incoherent. How does this thing called “free will” take that decision? Is it a random decision, or a deterministic decision? If random, then why not simply toss a coin to decide? If deterministic then I agree with you.

Plugging randomness into the mix guarantees that there can be “alternative possibilities”, yes. But any rational analysis shows that all this does is to make the outcome random instead of determined. You have not shown how your “free will” controls this random element to prevent the outcome from being simply random – your explanation is incoherent. If the free will controls the outcome, then by definition that control must be deterministic, otherwise it is not in control. You want the best of both worlds, you want the alternative possibilities afforded by random behaviour combined with the control afforded by determinism – but you cannot have the two together. It doesn’t work. Either the outcome provides for alternative possibilities (is random or stochastic, therefore not under the control of anything), or it does not (therefore can be controlled by the will). There is no third way (except in the illusion of an incomplete and therefore incoherent explanation). Your explanation is the equivalent of “and then a miracle occurs” – because you have not explained how this thing called free will controls the outcome without also making that outcome deterministic.

Uh, no, if we were able to formulate any form of regularity of this soul, it would simply become part of the physical world, right ? It would simply be an extra degree of freedom, to be described by laws (them being deterministic or not).

You are thinking of physical regularity and physical laws. I am talking of logical regularity and logical laws. Or perhaps you are claiming that this soul operates outside of logic? If so then end of discussion, because we have no way of logically discussing something which is not logical (it becomes something akin to metaphysical faith which is not amenable to any rational or logical analysis, like the existence of God). This “explanation” of free will then simply avoids giving an explanation by moving the free will into a realm where it is beyond rational explanation – it’s like the explanation which concludes with “and then a miracle occurs”……

If it is outside the physical world, it doesn't operate according to any principles of course. That's why it is outside, mystical, spiritual, or whatever qualifier one needs to add to say exactly that.

Thus you are saying that this thing called free will is located in the soul, and the soul is not amenable to any kind of logical or rational understanding? OK, as I said, end of rational discussion.

I'm not saying anything of the kind. First of all, I don't postulate any soul in this case, I'm only indicating that such a meta-physical concept is necessary to even start to address the concept of "free will" - as far as I'm concerned, the concept is not needed, and I can live with the illusion of free will, entirely generated by physical laws.

I also can live with the illusion of free will, entirely generated by physical laws. We don’t disagree then.

Free will is something that operates ON the physical universe, without the entity expressing it, being part of it. If it were part of it, it wouldn't be "free will" but just "following the laws of nature". If those laws of nature are deterministic, all future events are already determined, and there are no alternatives possible (although, with restricted knowledge, one might THINK that there are alternatives and hence a choice). If the laws of nature contain irreducible random elements, then this simply means that different events are possible. So alternatives exist. This is the place where the alternatives might be picked by something extra-physical without being in contradiction with the laws of nature - or not. But this possibility has already to exist in the first place before such extra-physical thing can even start to consider alternatives.

Agreed. And this is the premise of the free will believer – the only chance for alternative possibilities lies in random behaviour – therefore the free will believer denies determinism (he has no choice). But simply denying determinism does not solve the problem – there is still no rational, logical or coherent way to explain how this thing called free will controls an outcome without thereby making that outcome deterministic. Moving the free will to a realm where logic and rationality do not apply simply means that we end up with a “hocus pocus” explanation shrouded in smoke and mirrors which ends with “and then a miracle occurs”. Some people may be happy with that kind of mumbo-jumbo approach to philosophy, but it’s not for me.

As I explained elsewhere, I don't think that deterministic machines "choose". They have no alternatives from which to choose, because their programme leaves no alternatives. They CANNOT do what is not prescribed by their programme. If you don't know the details of their programme, it might look to you (because of lack of information) that there are alternatives open to them of which they pick one. But if you know the programme, you don't think that these alternatives are open.

We will have to agree to disagree then. Do you also deny that there is such a thing as a “random choice” or a “random decision”? I am sure you agree there is such a thing as a “free will choice” and a “free will decision”. Then why not a “determined choice” and a “determined decision”?

I accept that you define choice as “free will choice”. However, at the same time I recognise the possibility of deterministic choices, random choices and free will choices. Clearly we disagree on the meaning of the word choice. To move on, perhaps we need to agree some common terminology.

Clearly a deterministic agent (such as a human being) is doing “something” when it provides a particular output from a series of given inputs. We simply need to come up with a word that we both accept that describes that process. How about deterministic selection?

Would you agree that a deterministic agent can make deterministic selections? If yes, then perhaps we can use that word instead. In which case we have the possibility of deterministic selections, random selections, and free will selections. Is that OK?

They can "choose" to do things in the same way as apples can choose to fall to the earth.
The alternatives are like apples deciding not to fall up to the moon. Or, another analogy, if it is foggy, and you don't see the tracks, you might think that the train "decides" where to go, while if you know the track, you don't consider that the train has alternatives.

And as I explained above, these examples do not meet the necessary conditions for choice as in my suggested definition. But let’s try to get out of this rut of the definition of “choice” and move on – if you agree with my suggestion to use selection instead?

I re-read some of your posts in this thread, and think we're talking next to each other on some point. The "randomness" is not in the "free will" of course, it is in the laws of nature that concern the event to be decided upon.

Yes, I understood this – but this implies that the free will operates deterministically (you have said the randomness is not in the free will – even though earlier you seem to have implied that it makes no sense to talk of regularity or irregularity when discussing the soul). If the free will operates deterministically, then what it “wills” is deterministic. How do you get from this deterministic will to a physical act which is not deterministic by adding in physical randomness, unless that act is not in accordance with what the “free will” wills?

As I outlined above, the entity that is going to act its supposed free will (call it a soul) is outside of the physical universe, and is not a degree of freedom to which any regularities/laws/... apply.

There seems to be an inconsistency here - how can you say randomness is not in the free will if no regularities/laws apply to the free will? Randomness is simply absence of any regularity – that is how it is defined (and I am not restricting myself to physical regularities here – the same logic applies to the non-physical realm, unless one wishes to claim that the non-physical realm in question does not obey logical laws – but then as I have said end of rational discussion).

But if it is to act "freely" upon the physical universe, in order to steer things one way or another, it must have alternatives to pick from. So the physical universe itself must "leave open" some room, which it doesn't if it is deterministic.

Again, this explanation is incomplete hence incoherent – you have not shown how the “act freely” arises. I agree that physical randomness allows for alternative possibilities, but you have not shown how the “free will” can control those alternative possibilities to result in an act which is in accordance with its will without that act also then being deterministic.

I didn't mean to "add noise to the soul" or something of the kind so that it decides in "unpredictable ways - even to itself".

If it is predictable to itself then it is operating deterministically, by definition. Predictable means that, at the time of prediction, there are no alternative possibilities. If you are now claiming that the soul does NOT decide in unpredictable ways (to itself) then logically it follows that the soul is operating (logically) deterministically (deterministic at least from the point of prediction to the point of actual decision).

(I actually believe that most of the time we do not decide in predictable ways, even to ourselves, and that is one of the main reasons why we have the illusion of acting with free will – if we could always predict ourselves what we were going to do then it would be difficult to rationally maintain that illusion).

Yes, it is the same misunderstanding as with moving finger. I didn't mean the choice to be "random" as with "noise", or "unreflected" or anything of the kind. I meant: one out of many, without an instruction in advance of which to pick (because that would be no choice, would it ? If they say to you: "what icecream do you like, vanilla or chocolate, but you have to take chocolate", you don't really consider that you have a choice, right). And in order to even have physical laws which ALLOW for many possible alternatives, they cannot be deterministic (and hence, have to be random).
The choice is "random" for nature, in that case, in that *nature* doesn't "know" what you will pick (amongst the different possibilities left open). In a deterministic universe, nature does know what you will pick, so there's no choice left.

This seems inconsistent. If a choice can be “random for nature”, as you claim above, then in a similar way a choice can be “deterministic for nature”. How can you allow that it is acceptable to talk of random choice but not of deterministic choice?

I think choice in this context ought to be taken to mean - the possibility of having more than one outcome for the same initial conditions.

Say we have two options on the dinner menu, chicken and beef.

In a deterministic world this choice is pre-programmed. We would pick one over the other no matter what.

How can you say “In a deterministic world this choice is pre-programmed” if a deterministic machine does not (by your definition) make choices? What you should say (to be consistent) is “In a deterministic world there are no choices”.

You can indeed say that apples always "choose" to fall to the earth, even though they "could", if they wanted to, rise up to the moon, but, for some reason or another, they simply don't make use of that possibility, otherwise entirely open to them.

I hope you are not putting this forward as a serious definition of choice.

Nevertheless, that's the definition of choice *you* put forward, when you say that deterministic machines "evaluate, determine which course of action is best and then determine which course of action occurs".

Are you seriously suggesting that an apple falling from a tree really does "evaluate, determine which course of action is best and then determine which course of action occurs"?

After all, the assignment of values such as "best" etc... are just shifting around the notion we're trying to grab: free will. If you simply look at the compiled computer code in the memory of a deterministic machine, without knowing the intentions of the programmer, then you can determine that the machine will do this or that, in exactly the same way as you can write down the dynamical laws for an apple falling from a tree. Words such as "evaluate the situation, determining what is best" etc... are embellishments of simply saying: follows his deterministic dynamical laws.

I agree. And if I could look at and understand the decision making processes in your brain, then I can determine that you will do this or that.

The point here is that an apple falling from a tree does not (as far as we know) undertake any rational analysis of its situation, and does not determine which “way of falling” would be best, whereas a human being, or a deterministic machine with goals, objectives, ways of rationally analyzing how to fulfil those goals and objectives, does. The issue where we disagree is whether that process is ever “free” or not, so I don’t understand why you are confusing the discussion with talk of apples falling from trees.

Who knows ? Maybe there's a ghost in the apple, who "evaluates the situation" (calculates his lagrangian), and "determines what is best" (that is, tries to minimize its action), and hence, falls down, because the apple determined that as his "best course of action".

Agreed this is a logical possibility – which is why I qualified my original statement about the non-existence of “choice” in apples with “as far as we know”. If you wish to lead the discussion onto talk of ghosts in apples then (as interesting as I am sure the subject is) I’ll have to say goodbye at this point.

Any deterministic prescription of "determining the best course of action" is nothing else but other words for "follows his deterministic dynamical laws".

The two phrases are not synonymous. You are partially right in the sense that "determining the best course of action" entails "follows deterministic dynamical laws", but it does NOT follow that "follows deterministic dynamical laws" entails "determining the best course of action" (because an apple follows deterministic dynamical laws, but - as far as we know - it does not determine the best course of action).

It is also what an apple does. It's "value system" is "have minimal action", and his course of action is to fall down to earth. It feels also that it did exactly what it wanted to do.

If you wish to discuss whether an apple “feels” then you’ll need to find someone else to take that one up.

Well, you don't know any more that an apple is not evaluating its action and tries to minimize it. And as I said, if deterministic laws (call them evalution of the situation and deciding on the best course of action, if you want) determine indeed what you are supposed to decide, then any other option than the one prescribed by the deterministic laws is just as closed to you as falling up to the moon is to the apple.

I am not saying that my deterministic actions are not closed. My actions are just as closed as the apple’s actions. That’s a non-issue. The difference between me and the apple is that I have a “will” whereas the apple (as far as we know) does not. The issue is whether that will is “free” or not. If you wish to discuss whether an apple has a “will” or not then you’ll need to find someone else to do that with.

(continued below)

Best Regards

 

[moving finger]

(continued from previous post)

It is just that you've given a fancy name to the dynamics in certain cases, such as "evaluation of the situation followed by a decision", but deep down, it just means: follow the deterministic dynamical laws.

I agree completely – but the “dynamics in certain cases” is part of the difference between an apple and a human. The issue is not whether these dynamical laws exist, the issue is whether our “will” is “free” or not. In neither the apple case nor the human case is there any free will.

What else could it be ? Unless you consider "choice" as certain forms of dynamical laws, such as "if then else" statements in computer code. But they are not more than dynamical laws of course.

Agreed they are not. A deterministic choice is deterministic. A random choice is random. And a free will choice (if such exist) is free.

I'm not sure free will exists either (unless, we redefine it as a specific form of dynamics - which kills the essence of the concept). But we're trying to see what it would imply if it existed, and in how much it is compatible with a deterministic physics. I claim that it is not compatible.

I agree free will (in the libertarian sense of free will) is not compatible with determinism (but there are some such as Norman Swartz who argue that it is). This is precisely the reason why the supporters of free will (generally) do not accept that the world (where here I am talking of the entire causal world including the soul, if the soul is not epiphenomenal) is completely deterministic.

Well, to me, that's just fancy words to describe dynamics, or laws of nature. From what point on does a dynamical rule become a "choice, decision and selection" when the "intentions" are not known, and in what way is this different from an apple "choosing, deciding" to minimize its action and "select" the movement that corresponds to this "choice" ?

All words are “just fancy words to describe” something else – that’s what semantics is about. The important issue in rational debate is to agree on the meanings of words. As I have said, I can accept that you define choice to be “free will choice”, thus we need to find another word to describe what it is that humans do when they evaluate the relative consequences of two or more logically alternative courses of action, decide (based on comparison against a set of internal values which the human owns as “its value system”) which course of action is the best, and then to determine which course of action occurs - because if I apply your definition of choice then in my philosophy humans do not make choices.

I think it is only in that sense that there is "choice". If you just execute an algorithm, differential equation, minimum action principle or what ever, I don't see what "choice" there is, given that there are no allowed alternatives than what the algorithm/differential equation/minimum action principle is prescribing.

Let us, for the sake of argument, accept your definition of choice. Then in a world without free will (whether it be a deterministic or random world), choice does not exist. I do not believe free will is a coherent notion, therefore I do not believe choice (as defined by you) exists. Over to you.

But the "trajectory falling up" for an apple is also an "alternative in the counterfactual sense", meaning: it was a possibility (in the space of possible trajectories) but which was not picked out by a minimum-action principle. And if we define the "will" of an apple to minimize its action, then an apple also acts according to its "will". The alternatives have to exist in a real sense for there to be a genuine choice to be possible, no ? If the choice is prescribed, by an algorithm or any other deterministic principle, it's not a choice.

This is a non-issue if we accept your definition of choice. Neither the human nor the apple is making a choice.

The very point of a free-will choice is that one cannot say how it works, of course ! Because otherwise we would be back to case 1.

I can’t say how Santa Claus works either, but it is not safe to conclude from this that there is any coherency in the notion of Santa Claus.
Case 1?

To come back to the original poster's question, randomness means "ignorance about which to pick". When one says that something is random, one simply tells you that there are different possibilities, and that one doesn't have a rule to say - even in principle - which of the possibilities should come out. A choice can hence only be random to an entity that is not making the choice, because the entity making the choice is of course not ignorant of its very choice.

At what point is the entity making the choice “not ignorant” of its choice? Does it know the choice in advance of the choice, or is it ignorant all the way up to the very moment of choice?

If the former, then there is a deterministic link (ie regularity) between the entity’s “knowledge of the choice” with “the choice” (from the moment of knowing, there are no alternative possibilities) – hence the choice (using your definition) is no longer a choice; if the latter then the entity is just as ignorant about the choice as any observer.

In other words, if we have prior knowledge of our own choice, it follows that the choice is not a choice (because prior knowledge entails no alternative possibilities; and choice is defined, by you, as entailing alternative possibilities).

However, whether that entity should follow any deterministic rules to arrive at its choice is what is debated here. From the moment that there are such rules, there's no real choice anymore ! So a genuine choice doesn't obey any rules.

There seems to be a rule linking “choice” with “knowledge of the choice”, where that knowledge is (temporally) in advance of the choice. What comes first – does the entity “choose” what it wants to do, and thus has knowledge of that choice, or does it first “have knowledge” of that choice, and then choose according to its knowledge (but if so then the choice is not a choice as per your definition)?

But your wishes etc... are determined by dynamical prescriptions. Your wishes ARE dynamical prescriptions. The "wishes" of a deterministic machine are simply its computer code in its memory, and the "wish" of an apple is simply to minimise its action.

Agreed.

So your "will" is nothing else but the deterministic law itself, but just given a fancy name, in this case.

Agreed.

Yes, but you will have a hard time convincing me that your "will" is anything else but the dynamical rules themselves - over which you have no choice. Like the "will" of a deterministic machine is to execute its program.

I’m not trying to convince you otherwise – I agree with you here (except that “will” is not necessarily emergent from a machine simply executing a program)

You seem to think that I harbour some kind of belief that “free” will emerges from determinism – I do not. Whether the world is fundamentally deterministic or not makes no difference – “free” will is a logically incoherent concept.

Saying it is unavoidable leads to misunderstanding – because there are two types of unavoidable and many people get them confused. There is the “fatalism” type of unavoidable (where “what is going to happen is going to happen, whether you like it or not!”), but such things are not necessarily unavoidable under determinism (where “what is going to happen only happens because you want it to happen”). If I want to avoid something, and I am physically unconstrained, then I can avoid it, even under determinism.

Hell, you don't even have the possibility not to want to avoid it ! This possibility is not even open to you. Like the possibility not to follow its computer program is not open to a deterministic machine.

Again, we must be very careful what kind of aviodance we are talking about here. The kind of unavoidability (inevitability) inherent in fatalism is false.

I agree that I cannot avoid certain things, such as death. But I may be able to avoid taxes for a while (if I am lucky) and I should be able to avoid being knocked over and killed in a car accident tomorrow (if I am careful). As I said, if I want to avoid something, and I am physically unconstrained, then I can avoid it, even under determinism. Do you deny this?

If you think there are cases where I cannot avoid something that I want to avoid (assuming I am physically unconstrained) then perhaps you could give an example.

If on the other hand you are simply saying that we cannot avoid things that we do not want to avoid then I agree – but so what? Why on Earth would I be concerned about not being able to avoid something that I did not want to avoid?

Your "will" is nothing else but the dynamical law you are going to have to follow in any case. There are no alternatives open to you, and if you think so, then that's the *illusion* of free will you have.

I don’t have any illusion of free will. Free will is an incoherent notion. Again I think you are ascribing beliefs to me which I do not have.

"Desiring" something else than what is determined by the deterministic dynamics since the big bang, is not an option, in the same way as *not* minimizing its action is not an option to an apple. So in as much an apple has any "desire", it must be to minimize its action.

If I am unconstrained (free) to do what I want to do, why would I desire something different? Where is the rationality in such a position? But being unconstrained to do what one wants to do is perfectly compatible with determinism, and does not entail any incoherent notion of “free” will.

The same is true under determinism – if I had chosen to do something different then I would have done something different.

No, you couldn't. The option was not open to you, you could not have desired anything else than what was fixed since the big bang.

This is clearly not what I said. I said “If I had chosen” – and you interpret this as “I could have chosen”. I never claimed that the option was open to me to do something different – I was using “if” in a counterfactual sense (in the same way that one would say “if it had not rained yesterday, we could have had our picnic” – there is nothing in here which implies we have any control over whether it rained yesterday or not – it is simply a counterfactual statement).

You claim that I could not have desired anything else than what was fixed since the Big Bang – and I agree – but again, if I am unconstrained (free) to do what I want to do, why would I desire something else (which would then be what I do not want to do)?

BTW, I'm not advocating free will, in fact in as much as I have to pick anything, I'd rather say that we do not have free will (but only the illusion).

It’s amazing that we agree on this, and yet disagree on so much else!

But if we are going to require "free will" I'm only indicating that, by definition, it is not going to follow any deterministic prescriptions, and that the laws of nature need to leave some room for it by being *ignorant* of what will happen - so leaving open more than one option from which to pick.

I agree that this is the route that free will believers take. They cannot see free will being compatible with determinism, therefore the solution (if it exists) they think must lie in indeterminism or randomness. But any attempt to explicate how this works ends up in hocus-pocus, smoke and mirrors – such as pushing free will into some mystical soul-realm where it is claimed to be beyond rational explanation – hence the reference to “and then a miracle occurs”. Nobody can fully explicate how the free will mechanism is supposed to work because any such mechanism would be logically incoherent. My point is that the solution does NOT exist – or at least that if it does exist then it is beyond rationality and logic (hence metaphysical and unworthy of further rational debate). The main reasons some of use believe we have free will are (a) because we want to believe that we are autonomous agents (we cannot accept that we are deterministic machines) (b) because we are ignorant of the details of most of our decision-making processes and (c) because we are prepared to tolerate smoke-and-mirrors explanations resulting in “then a miracle happens”. But not everyone accepts these reasons.

As I said, it is essential for an entity exercising free will that it doesn't follow any dynamics, or rules. The randomness resides not in the exercise of free will from the PoV of the entity exercising it, but rather externally: externally, one has to remain fundamentally ignorant of the decision the free will entity will take, and hence needs to introduce an element of randomness (= ignorance) in any dynamical prescription of the event.

All this does is to make actions random – without control. If you are saying that free will entails random behaviour then fine – but that’s not the kind of freedom I would want, and it’s not the kind that libertarians want. The kind of freedom I want is simply to be unconstrained by forces outside my will (which is compatible with determinism) – I have no desire to sprinkle random events into the mix – what possible good would that do? The kind of free will that libertarians want, however, is incoherent.

Well, the essence of the concept is that we cannot say how it works, because from the moment we can, we destroy the very concept of free choice !

In other words, the entire concept is incoherent.

The second point is that, in order not to destroy the very possibility opened up by having now different potential alternatives, the entity exercising free will cannot, itself, be subject to any regularities of laws itself (because these regularities or laws would then close the door that was opened by the non-deterministic laws in the first place). So it remains a fundamentally mystical concept.

Mystical perhaps – but more importantly incoherent and inconsistent. You have said above :

The randomness should then be on the side of the purely physical, so that the choice between different possible alternatives is not in contradiction with the laws of nature.

Which implies that randomness is on the physical side, but the “soul” side is regular (following logically deterministic laws).

I didn't mean to "add noise to the soul" or something of the kind so that it decides in "unpredictable ways - even to itself".

Similarly, this suggests (“I didn’t mean to add noise to the soul”) that the “soul” operates predictably (hence deterministically – one cannot have infallible predictability without determinism), at least to itself.

But now you seem to be claiming that the ‘soul” is NOT “subject to” any regularities. But by definition if it is not “subject to” any regularities then it is operating randomly…..

Which is it to be? Regularities (determinism) or no regularities (random behaviour)?

But, again, I'm not an advocate of free will. I'm only putting down the consequences of the concept, that's all.

I understand that – and I am saying that the logical consequences of the concept lead to inconsistency or incoherency. The only way the free will supporter can avoid this incoherency is by burying the explanation in inaccessible mysticism. I can justify ANY belief on that basis.

No, not at all. Free will would simply imply that there are limits to the rational description of nature, that's all. That some things are fundamentally underscribable.

It is very easy to claim that something is undescribable by simply refusing to attempt to describe it. I can justify a belief in the Tooth Fairy, Santa Claus, Tokoloshes, Vampires, Ghosts, and any manner of entities using the same “argument”. It’s not a coherent argument.

The tracks are nothing else but the deterministic laws of nature, and what you describe as your free will is nothing else but that same dynamics. You want the train to go where the track is, the track is where you want the train to go, and this is already determined since the big bang. Exactly as the apple wanting to fall on exactly the trajectory that is minimizing its action.

I want to go where I want to go, and my wanting determines my action, not the other way around. If the metaphor is “the train” = my will and “the track” = what I do (how I act), then it is not the case that “I want to go east because the track goes east”, it is the case that “the track goes east because I want to go east”.

As I said, a free will entity must necessarily be undescribable by any form of dynamics. It's part of its essential being.

If part of an essential being is being logically inconsistent then I agree. The only way to avoid logical inconsistency in free will is to claim it is undescribable.

No, it is fundamentally undescribable, that's all. You can call it mystical. Yes, free will is an essentially mystical concept. Mystical in the sense of fundamentally undescribable.

Undescribable because any attempt to try to adequately describe it entails inconsistency.

Best Regards

 

[Astronuc]

how randomness can exist?

Because there are an infinite number of directions, but something can only go in one direction at a time.

The quantization of matter and the ultimate speed limit (speed of light) are somewhat regulatory.

 

[Farsight]

What about quantum fluctuations then? Are they random?

 

[superweirdo]

When a person is born, he is being influenced by his genetics(which do continue to influence him for the rest of his life) Some even say that we can only see 10% of genetics, others can even tell you the future! Anyways, later, he starts being influenced by society, even later, he starts being influenced by his soul. But all of those are govern by laws-simple physical laws, for the soul, they have their own law that we aren't aware of. All we do is caused by chemical reactions in our brain which are caused by the diet we get, the air we breath, the genetics we have, and our surrounding environment. Beyond that, we say soul gives us will, it is true, but probably, even it is being influenced by something else that we aren't aware of and all we are is reaction. We are never the action, we always react. And just like it is predictable if you throw water on a match with fire, it wil dimish it. Our actions can be predictable too.(theoretically)

 

[moving finger]

Because there are an infinite number of directions, but something can only go in one direction at a time.

tell that to the photon which goes through both slits at the same time

The quantization of matter and the ultimate speed limit (speed of light) are somewhat regulatory.

the speed of light does not seem to be a barrier to some kind of communication between entangled states

Best Regards

 

[Mickey]

Well, I would think it matters for people looking for potential deeper theories beyond the quantum theory we have today. Should they require determinism (at least ontic determinisim), or can they live with ontological randomness in the formulation of the "ultimate and final" theory they are going to formulate next week ?

Of course you're right that for the working physicist, who works *within* a certain paradigm, it doesn't matter at all. But for those trying to think up new paradigms, it might play a role.

Hmm. Maybe, it's a question of whether we can have such a thing as intelligible randomness?

Randomness, being what it is, is not rationally intelligible, but maybe we're still capable of modelling it, without violating occam's razor, resorting to philosophical speculation, or any non-empirical method.

Entanglement appears to play a role here. Looking at one list of unpredictable measurements, we can make a prediction for another list of measurements that are otherwise unpredictable. Does entanglement offer us a glimpse at a non-rationally intelligible randomness?

Perhaps there are other phenomena like entanglement that preserve randomness but otherwise make it intelligible, as opposed to just statistical.

 

[moving finger]

Hmm. Maybe, it's a question of whether we can have such a thing as intelligible randomness?

That's an oxymoron. Randomness by definition contains no useful information whatsoever - if it did then it wouldn't be random. If there is no useful information, how can it be intelligible?

Randomness, being what it is, is not rationally intelligible, but maybe we're still capable of modelling it, without violating occam's razor, resorting to philosophical speculation, or any non-empirical method.

Entanglement appears to play a role here. Looking at one list of unpredictable measurements, we can make a prediction for another list of measurements that are otherwise unpredictable. Does entanglement offer us a glimpse at a non-rationally intelligible randomness?

Entanglement may be counter-intuitive, but is not "non-rational", and when we do finally understand what is going on there is no reason to think that the explanation will be non-rational. If one could make intelligible sense of something which one thought was random, then it would no longer be random.

At the end of the day - there is no infallible test for absolute randomness. The best we can ever do is to say "well it sure looks like it is random, but it is always possible there is some regularity hidden in there that we just cannot see yet".

Best Regards

 

[Mickey]

I was very careful not to say that entanglement was non-rational. I only asked if entanglement would allow us to begin to understand randomness as it exists in nature while preserving its status as practically irreducible randomness.

Pure mathemetical randomness may contain no useful information whatsoever, but natural randomness might.

Don't worry, I already said I thought this was idle conversation. Just trying to spin some of it.

 

[moving finger]

I was very careful not to say that entanglement was non-rational. I only asked if entanglement would allow us to begin to understand randomness as it exists in nature while preserving its status as practically irreducible randomness.

My apologies, you are quite right.

My point regarding randomness and intelligibility stands, however. If a certain piece of information is intelligible then by definition that information is not random. If we thought it was random then this simply reflects our ignorance about it's true nature.

Pure mathemetical randomness may contain no useful information whatsoever, but natural randomness might.

By definition, if its random it contains no useful information.
If it contains useful information, then calling it "natural randomness" is an incorrect use of the word random. It would be better to say "the appearance or illusion of natural randomness".

Best Regards

 

[wierd101]

Whoa. Alot of talk on what would seem to be a very simple thing (for some). I have kinda skipped several pages because of all the talk that seems to go on and on without producing a result.
One point of view would be (which seems most logical to me):
In order for us to predict (with absolute certainty) something happening, you need to know all the variables. When it comes to predicting if something will fall if you pick it up and drop it, you can guess what the result will be (it will fall until it can't any more) since the variables that are most effecting the object are ones that we can know to a fair degree of accuracy - they are the ones that govern how we live, so we need to be able to get them allmost right. As the end result of the problem gets smaller, the lesser variables have more effect on the whole, making it more important that you know everything in order to predict that result.

Provided you can know all the variables in an equation, you can predict the result. It is most likely however, that since we are part of the equation, it makes it impossible to know every other part of the equation. Even if we did manage know just about every variable, or at least enough to calculate everything we can think of, there may be more variables out there that affect things that it is impossible for us to know of.

In mathematics, it is possible to create an artificial universe in which you can know every variable if you know enough of the whole to begin with. There are still problems in math that seem like randomness as we do not know all the parameters or have not spent enough time solving them, and there are equations that are random simply because we refuse to tell the rest of the equation what part of it could be.

If we then take into account the possibility of a God, we must assume that this entity is the source of all randomness, as it is the ultimate unpredictable source. Many beliefs contradict this, while others agree. It is more a topic of discussion that cannot be resolved, as we will never know if we simply don't know all the possible variables or something is out there changing them before we can tie them to paper.

And i think that there are a couple of other variables to this problem that i can't quite remember at the moment, but for now, randomness is either an illusion brought on by being stupid (as humans are) or something that we will never know the answer to. Strange how the result of the randomness equation is un-knowable or random.

 

[moving finger]

Whoa. Alot of talk on what would seem to be a very simple thing (for some). I have kinda skipped several pages because of all the talk that seems to go on and on without producing a result.
One point of view would be (which seems most logical to me):
In order for us to predict (with absolute certainty) something happening, you need to know all the variables. When it comes to predicting if something will fall if you pick it up and drop it, you can guess what the result will be (it will fall until it can't any more) since the variables that are most effecting the object are ones that we can know to a fair degree of accuracy - they are the ones that govern how we live, so we need to be able to get them allmost right. As the end result of the problem gets smaller, the lesser variables have more effect on the whole, making it more important that you know everything in order to predict that result.

It does not follow that "knowing everything" will allow you to predict the result - unless you also know the result. If there are truly random (indeterministic) processes at work then their outcome cannot be predicted.

Provided you can know all the variables in an equation, you can predict the result.

If you know all the variables, there is nothing left to "predict", is there?

It is most likely however, that since we are part of the equation, it makes it impossible to know every other part of the equation. Even if we did manage know just about every variable, or at least enough to calculate everything we can think of, there may be more variables out there that affect things that it is impossible for us to know of.

Chaos theory, HUP, and the indeterminability of self-referential systems all work to prevent infallible predictions.

In mathematics, it is possible to create an artificial universe in which you can know every variable if you know enough of the whole to begin with.

This only works because you have eliminated chaos, HUP and self-referentiality from your "artificial universe" - it doesn't work in real life.

There are still problems in math that seem like randomness as we do not know all the parameters or have not spent enough time solving them, and there are equations that are random simply because we refuse to tell the rest of the equation what part of it could be.

This has nothing to do with randomness - it's simply an "unknown"

Strange how the result of the randomness equation is un-knowable or random.

Unknowable is not the same as random.

Best Regards

 

[vanesch]

Hmm. Maybe, it's a question of whether we can have such a thing as intelligible randomness?

This brings us back to the previously discussed difference between epistemological randomness (= ignorance randomness) and ontological randomness (the irreducible kind).

It's not clear what you mean by "intelligible randomness", and I agree with moving finger's response. Maybe you mean: purely epistic randomness without ontological foundation (which is *usually* what his understood by randomness).

Entanglement appears to play a role here. Looking at one list of unpredictable measurements, we can make a prediction for another list of measurements that are otherwise unpredictable. Does entanglement offer us a glimpse at a non-rationally intelligible randomness?

Entanglement as such doesn't do anything of the kind, it just complicates an eventually underlying determinism. Although it is not the same, it is probably understandable: consider a non-local deterministic theory (such as Bohmian mechanics). The motion of a particle in the Andromeda galaxy can *seriously* affect the motion of a particle in your lab on earth, due to the non-local character of the dynamics. Bohmians use it to *mimic* entanglement, but it is not the same (although it can produce the same results). It is true that a deterministic, even Newtonian, mechanics, with strong non-locality such as in Bohmian mechanics, would make any apparent randomness untracktable, because you could never confine the system under study to a finite region of space with an (even approximate) local dynamics.

Entanglement does something similar, although it is conceptually different.

However, both phenomena (entanglement, or non-local deterministic dynamics) do not really address the issue of randomness ; they just make the follow-up of the dynamics even more hopeless by potentially including the entire universe significantly into the dynamics.

 

[Mickey]

Oh, interesting. I hadn't appreciated the distinctions between entanglement and Bohmian mechanics yet.

I suppose it just underscores my original point. All this non-local business makes approaching the randomness question even more daunting, not just for the fact that the universe is so large, but that we'd have to take into account the whole of it.