Does Physics disprove the existence of free will?

[Baldeee]

I don't think it's the same as the argument put forward in post #130? That argument argues from determinism, but my question is about the logical implications of the definition of the terms, and it is completely independent from determinism.

Then I'm not sure I fully follow what you're saying.
First, though, I am not assuming that all objects necessarily have actions they "must" do.
But determined ones would fit within those that (I think) do.
As such you seem to be saying that if an object is one that can only do what it must then it is not free.
And I would agree.
But I think it more likely that I am not understanding your point.

Yes, and this "multiple usages" of the word "free" is why I explicitly focused on one, and underlined it.

Noted.

But that leads to an infinite regression. Set A contains two options, 1 and 2. Now object O must choose between 1 and 2. Set B is thus constructed, of the options "choosing option 1 in set A" and "choosing option 2 in set A". Now object O must choose between those two options. Thus, set C is constructed, of the options "choosing the first option in set B" and "choosing the second option in set B". This goes on forever.

Not if there is inherent randomness that removes the regress and effectively makes the "choice".

And "which possible option it must do" is troublesome in itself, must being defined as "can not do otherwise". If it must do one option, then it by this definition cannot do the other options, thus those options are not in the set of possible options; they were never options at all. You're always left with just one option, namely the one that gets done.

I am tending to give randomness a touch of licence here, in considering the options remaining open until randomness cuts it down to one, which I would think is at the point of action itself.
But if we are talking strict determinism, in which there is only one option, then yes, I agree.

Yes, but as I pointed out, it goes further. The set of possible options can naively be based on what object O can do. However, we can also make a set of options Object O must do. It is not possible for object O to pick an option outside either of these sets, and thus it could never have done any of the options that is in the "can"-set but not in the "must"-set. These two sets turn out to be equivalent. But the definition of free demands an option outside of the set be picked, and thus no object can ever be free, by definition. That seems like a waste of a word to me.

I think you are under the assumption that I am saying (or have said) that all objects are necessarily ones that can only do what they must.
I haven't said that.
I don't know if they are or not.
I have merely referred to the case of those things that are determined (or probabilistically determined).

Further, probabilistic determinism keeps open the options in the "can" set until the point of action, at which point it is randomly identified (in accordance with the probability function) as the "must".

I don't see how randomness factors into your definition of free?

It doesn't, but in probabilistic cases I see it as bridging between the multiple "can" to the singular "must".

Sure, but if "free will" is based on your definition of free, then "free will" can't exist by definition; that's my point. All other arguments are unnecessary, because you've implicitly defined "free will" to be nonexistent in the first place.

Implied, yes, as a logical consequence, as far as determined systems go.

That's not what your definition appears to be, as I pointed out that it is more along the lines of: "if, for a given set of possible outcomes, something cannot pick an outcome not part of that set, it is not free."

And if that set is singular in content then they are equivalent, are they not?
If you have but one option, you can not pick outside, and you have no control over which you do.

And I don't disagree with your arguments and logic. My only point here is that your definition of free seems to already rule it out entirely.

Understood, although I'm not convinced.
But that may due to misunderstanding.

Which is why I'm explicitly underlining the word, to make that distinction clear.
...
Which is why I'm only discussing your definition of the word; I'm explicitly trying to stay away for that hornet's nest.
...
Completely agreed.
...
Such resistance is typically a bad omen for the discussion in general.

Noted.

I've seen incoherent definitions, useless definitions, tautologies. But I think the most egregious one I've come across is that "free will" was given to us limited creatures by an all-powerful, all-knowing divine being, and that it gives us the ability to go against the will of said divine being. Typically, this being is also all-good, and it is this "free will" given to us that allows us to do evil. It's incoherent and self-contradictory on so many levels, I was quite amazed by it!

Sounds appealing.
Care to move it to our religion forum and give it a whirl?

 

[geordief]

If the multiworlds hypothesis where all possible outcomes do in fact occur is accepted as feasible could it have any bearing on how to define determinism?

Within this hypothesis what status is accorded to outcomes which are more probable than others?

By the way I would still only regard such a diversion ás philosophical banter unless such a hypothesis had some way of being experimentally verified (and shown to be of some value too,I suppose)

Edit: have not been able to keep up with the thread so all goes if this has already been mentioned.

 

[Baldeee]

If the multiworlds hypothesis where all possible outcomes do in fact occur is accepted as feasible could it have any bearing on how to define determinism?

In the probabilistic world, I'm not sure it would have any impact on how we define it.
In multiworld hypothesis, it would presumably be that every possible outcome of an interaction happens in some universe or other, with the number of each possible occurrence following/matching the profile of the probability function.
We still wouldn't be able to determine other than probabilistically which outcome would occur in our universe, though, so nothing would change.

Within this hypothesis what status is accorded to outcomes which are more probable than others?

The status of "more likely", perhaps?

By the way I would still only regard such a diversion ás philosophical banter unless such a hypothesis had some way of being experimentally verified (and shown to be of some value too,I suppose)

Not sure you will ever consider it something else then.

 

[geordief]

Not sure you will ever consider it something else then.

True ,but I will accept that that may be a shortcoming on my part (I have to live within my limitations)

 

[Yazata]

If the multiworlds hypothesis where all possible outcomes do in fact occur is accepted as feasible could it have any bearing on how to define determinism?

The many-worlds hypothesis (whether quantum or modal) seems to me to be inconsistent with determinism. Determinism seems to insist that state A determines state B, determines state C, all the way out to state Z, so that the chain of causes can only evolve along a single pre-ordained path and hence there can only be one single world with no other alternative possibilities possible. Initial state A and the dynamical equations determine everything that can possibly happen forever into the future.

I personally consider that a metaphysical belief, one that I don't personally share.

 

[DaveC426913]

The many-worlds hypothesis (whether quantum or modal) seems to me to be inconsistent with determinism. Determinism seems to insist that state A determines state B, determines state C, all the way out to state Z, so that the chain of causes can only evolve along a single pre-ordained path and hence there can only be one single world with no other alternative possibilities possible.

It is still deterministic if state A leads to both state B and C, with B leading to B1 and B2 and C leading to C1 and C2.

I start with bread and apply my two types of meat.
Now I've got two sandwiches.
I apply two types of cheese.
Now I've got four sandwiches.
All entirely deterministic.

Apparently, it's lunchtime at chez DaveC416913...

 

[NotEinstein]

Then I'm not sure I fully follow what you're saying.
First, though, I am not assuming that all objects necessarily have actions they "must" do.

Sure, and neither am I.

But determined ones would fit within those that (I think) do.

Agreed.

As such you seem to be saying that if an object is one that can only do what it must then it is not free.

True...

And I would agree.
But I think it more likely that I am not understanding your point.

Right. Once again, my point is: with that definition of free, how can an object be free, as that seems to be a violation of logic?

Noted.

Not if there is inherent randomness that removes the regress and effectively makes the "choice".

I am tending to give randomness a touch of licence here, in considering the options remaining open until randomness cuts it down to one, which I would think is at the point of action itself.
But if we are talking strict determinism, in which there is only one option, then yes, I agree.

Hmm, this sounds a lot like fuzzy logic. You have to introduce a source of randomness into logic itself. That's a pretty big deal, but it would indeed solve that issue. Would you say this source of randomness is all that makes up "free will", or are there other components to it?

I think you are under the assumption that I am saying (or have said) that all objects are necessarily ones that can only do what they must.
I haven't said that.
I don't know if they are or not.
I have merely referred to the case of those things that are determined (or probabilistically determined).

No, I'm merely asking you how, with that definition of free, any object can be free.

Further, probabilistic determinism keeps open the options in the "can" set until the point of action, at which point it is randomly identified (in accordance with the probability function) as the "must".

Let's assume this is indeed what's going on; how then do you define "free will" from there?

It doesn't, but in probabilistic cases I see it as bridging between the multiple "can" to the singular "must".

You said, in post #217:

The former is, as I see it, a matter of randomness, and thus not free.

The "thus" implies a logical relation between the two. Can you please elaborate on your original statement, if that implication isn't what you meant?

Implied, yes, as a logical consequence, as far as determined systems go.

(I have not been talking about deterministic systems.)

And if that set is singular in content then they are equivalent, are they not?
If you have but one option, you can not pick outside, and you have no control over which you do.

Yes, as I explicitly pointed out in the very post you are quoting.

Understood, although I'm not convinced.
But that may due to misunderstanding.

Your definitions from post #188:
Free: able to do other than it must.
Must: can not do otherwise
Can: is able.
Combined: something that is free, is able to do other that it is not able to do otherwise.
Simplified: something that is free, is able to do actions that it is not able to do. (Sorry for using the word "actions"; it was the easiest way to make the grammer work.)
That seems contradictory to me, on a basic logic level. Adding randomness doesn't make the contradiction go away. Fuzzy logic might, but I'm not versed in that, and that's not something you appear to be arguing for?

Noted.

Sounds appealing.

Care to move it to our religion forum and give it a whirl?

No, I'm not interesting in discussing religion at this time.

 

[Baldeee]

Sure, and neither am I.

You seem to be, though.
I'll skip to your clarification at the end, as this may help clarify from my POV as well...
And forgive me for not responding line by line, I think/hope the points are still all covered in the below.
If not, please just point out where I have missed anything you would still like clarification on, or answers on.

Your definitions from post #188:
Free: able to do other than it must.
Must: can not do otherwise
Can: is able.
Combined: something that is free, is able to do other that it is not able to do otherwise.
Simplified: something that is free, is able to do actions that it is not able to do. (Sorry for using the word "actions"; it was the easiest way to make the grammer work.)
That seems contradictory to me, on a basic logic level.

I think the simplified version should include an "otherwise"...
E.g. ...is able to do actions than it is not otherwise able to do.
I don't think you can omit the "otherwise" and retain the same meaning.

Hmm, this sounds a lot like fuzzy logic. You have to introduce a source of randomness into logic itself. That's a pretty big deal, but it would indeed solve that issue. Would you say this source of randomness is all that makes up "free will", or are there other components to it?

I‘m not versed in fuzzy logic, but I don't see probabilistic outcomes resolving the issue.
I don't think free will exists, unless one is talking about the appearance of "choice" by our consciousness.
And I refer herein to that notion of freewill as an illusion when compared to the notion of free will from first principles (for lack of better phrase).

Maybe the issue of randomness is a distraction here.
Probability functions are actually inherently indeterministic, but are deterministic if one considers the same probability function of outputs rather than the same specific output under strict determinism.

Randomness is merely that which, not under the control of the interaction itself, converts the probability function into a specific result.
As to what drives that randomness... does anyone know?
If we did, would, or could, it still be random?
But they are questions for another thread, no doubt.

With regard freewill, I don't see it as offering too much, at least not as far as the argument I posted goes.

 

[NotEinstein]

You seem to be, though.

Since you bring it up again: what gave you that impression?

I'll skip to your clarification at the end, as this may help clarify from my POV as well...
And forgive me for not responding line by line, I think/hope the points are still all covered in the below.
If not, please just point out where I have missed anything you would still like clarification on, or answers on.

OK

I think the simplified version should include an "otherwise"...
E.g. ...is able to do actions than it is not otherwise able to do.
I don't think you can omit the "otherwise" and retain the same meaning.

Please explain the significance of the "otherwise", as this seems to be the crux to our discussion.

I‘m not versed in fuzzy logic, but I don't see probabilistic outcomes resolving the issue.
I don't think free will exists, unless one is talking about the appearance of "choice" by our consciousness.
And I refer herein to that notion of freewill as an illusion when compared to the notion of free will from first principles (for lack of better phrase).

Maybe the issue of randomness is a distraction here.
Probability functions are actually inherently indeterministic, but are deterministic if one considers the same probability function of outputs rather than the same specific output under strict determinism.

That is pretty much exactly what fuzzy logic is, if I got my terms straight.

Randomness is merely that which, not under the control of the interaction itself, converts the probability function into a specific result.

That would qualify any outside factor as "randomness", which I don't think you are arguing. For it to be true to the word "randomness", there's got to be an unpredictable, arbitrary component involved. I think that should be a necessary part of the definition; otherwise, it shouldn't be called "randomness".

As to what drives that randomness... does anyone know?
If we did, would, or could, it still be random?
But they are questions for another thread, no doubt.

Most likely, yes.

With regard freewill, I don't see it as offering too much, at least not as far as the argument I posted goes.

You said that you "don't think free will exists", and you have here a definition of "free" you are using. How do you (roughly) define this "free will" that you say doesn't exist?

 

[iceaura]

It is still deterministic if state A leads to both state B and C, with B leading to B1 and B2 and C leading to C1 and C2.

So it still requires - in your view - that natural law be contravened, overcome, defied, avoided, etc, for any freedom of will to exist.

I don't think free will exists, unless one is talking about the appearance of "choice" by our consciousness.

It's not mere appearance. The choices actually exist, and are actually made.

First, though, I am not assuming that all objects necessarily have actions they "must" do.
But determined ones would fit within those that (I think) do.

And in the physical universe cause and effect, the laws of physics, and the fact that all systems are built of smaller systems, determine all physical events.
All physical events are determined. All actions are physical events.
So all actions take place as they must, there is no freedom to will anything other than that, and
so, therefore,
there is no freedom of will.
That's your argument as it appears here.

 

[Baldeee]

Please explain the significance of the "otherwise", as this seems to be the crux to our discussion.

In the context used, it is intended to mean along the lines of "in circumstances different from that considered".
Something is free if it is able to do something other (something else) than the thing it has to (or must) do if not free.
If something is not free then, for example, doing A means it has to do B.
If something is free then it is able to do C, or D etc.
There is no "must do" for something that is free.

That is pretty much exactly what fuzzy logic is, if I got my terms straight.

Okay.

That would qualify any outside factor as "randomness", which I don't think you are arguing. For it to be true to the word "randomness", there's got to be an unpredictable, arbitrary component involved. I think that should be a necessary part of the definition; otherwise, it shouldn't be called "randomness".

Randomness I would consider to be inherent within the interaction itself.
Outside factors are just other interactions.
And yes, unpredictable and arbitrary, at least to a degree: individual interactions would be unpredictable, but an infinite such identical interactions would produce a probability function, such as that allowing us to measure radioactive half-life.
Each decay is unpredictable, but a large number follow a probability function.

You said that you "don't think free will exists", and you have here a definition of "free" you are using. How do you (roughly) define this "free will" that you say doesn't exist?

In an individual interaction it would be akin to that interaction effectively being able to pick its own output, unbeholden to any probability function and random selection therein.
This would then be truly free and appear absolutely random, with no probability function restraining the possible outcomes.
In a dynamic system that is similarly deterministic it would be the ability of that system to influence the output to something other than that required by the inputs to the system.

But this is more a work in progress.
I don't actually think it is necessary to be able to define it, though, other than to identify it as something that can not (per the logic) be free.

 

[TheFrogger]

Prediction is impossible. In the sequence, 2,4,6,8, you may predict the next number to be 10, but 12 may actually be drawn. Until the next number is known, it cannot be predicted.

 

[iceaura]

There is no "must do" for something that is free.

And there is a "must do" for anything bound by natural law, of course, which means any sequence of physical events.

In an individual interaction it would be akin to that interaction effectively being able to pick its own output, unbeholden to any probability function and random selection therein.

Not bound by natural law.

This would then be truly free and appear absolutely random, with no probability function restraining the possible outcomes.

Some confusion here. If it appears "absolutely" random, it appears to follow a probability function.

In a dynamic system that is similarly deterministic it would be the ability of that system to influence the output to something other than that required by the inputs to the system.

So, in any natural system, freedom means the ability to contravene natural law - to create effects other than those produced by available cause, to act otherwise than physical law determines, etc. There is no freedom otherwise.

No expansion or extrapolation of something like the engineer's degrees of freedom is possible. It's supernatural, or it isn't free.

By assumption/definition.

 

[Baldeee]

Prediction is impossible. In the sequence, 2,4,6,8, you may predict the next number to be 10, but 12 may actually be drawn. Until the next number is known, it cannot be predicted.

In a strictly deterministic system, prediction is theoretically possible, even if practically impossible due to complexity of system invovled, and the practical inability to identify accurately all the starting inputs.
In a probabilistic, or other indeterministic system, prediction can only be achieved on a probabilistic basis at best, not at all at worst.
Is your notion of free will therefore based upon predictability?
If so would this be theoretical predictability or practical?

 

[TheFrogger]

Practical.

 

[NotEinstein]

In the context used, it is intended to mean along the lines of "in circumstances different from that considered".
Something is free if it is able to do something other (something else) than the thing it has to (or must) do if not free.
If something is not free then, for example, doing A means it has to do B.
If something is free then it is able to do C, or D etc.
There is no "must do" for something that is free.

But that's circular; you can't define "free" by saying that objects are free if they are able to things that objects that are free can do. Your definition of "free" depends critically on the word "must"; you can't now define "must" in terms of "free"!

Okay.

Randomness I would consider to be inherent within the interaction itself.
Outside factors are just other interactions.
And yes, unpredictable and arbitrary, at least to a degree: individual interactions would be unpredictable, but an infinite such identical interactions would produce a probability function, such as that allowing us to measure radioactive half-life.
Each decay is unpredictable, but a large number follow a probability function.

I would describe that as an "unknowable" component, not fundamentally as "randomness". But I understand what you mean.

Under the assumption that such "unknowable components"/"randomness" indeed is part of interactions, that could indeed be defined quite nicely as "free". Thanks!

In an individual interaction it would be akin to that interaction effectively being able to pick its own output, unbeholden to any probability function and random selection therein.
This would then be truly free and appear absolutely random, with no probability function restraining the possible outcomes.
In a dynamic system that is similarly deterministic it would be the ability of that system to influence the output to something other than that required by the inputs to the system.

I see what you're getting at. A non-free system will always give the same output for the same input, a free system can given one of multiple (infinite?) outcomes. Free will would be the ability for the system to select which of the unpredictable outcomes results. Something along those lines?

But this is more a work in progress.

The same is true for me.

I don't actually think it is necessary to be able to define it, though, other than to identify it as something that can not (per the logic) be free.

Perhaps not a strict definition, but some description of what the concept entails would be required for a proper discussion. But I am fully aware of the difficulty of defining "free will" in a way that makes intuitive sense.

 

[DaveC426913]

So it still requires - in your view - that natural law be contravened, overcome, defied, avoided, etc, for any freedom of will to exist.

Many Worlds Interpretation - which is what Yazata was talking about - is not part of natural law. I do not advocate for it, I am simply responding to Yazata's post.

Go find someone else to bully.

 

[Baldeee]

But that's circular; you can't define "free" by saying that objects are free if they are able to things that objects that are free can do. Your definition of "free" depends critically on the word "must"; you can't now define "must" in terms of "free"!

If that is what I have done it is inadvertent.
I see "must" and "free" effectively as opposites.
One can define "must" as obligated, lacking ability to do something else, required to, etc.

I would describe that as an "unknowable" component, not fundamentally as "randomness". But I understand what you mean.

Under the assumption that such "unknowable components"/"randomness" indeed is part of interactions, that could indeed be defined quite nicely as "free". Thanks!

I'm not sure I'd agree.
I don't see a random system as necessarily being any freer than a deterministic one.
At least not in the sense I am using the term here.
Imagine a prisoner that gets let outside to exercise one hour a day.
In a deterministic world it would be the same hour each and every day.
In a random world it would be a different hour each day, with the hour decided either by some unknown or random component.
Is the prisoner ever "free" on the matter of which hour they are let out?

I see what you're getting at. A non-free system will always give the same output for the same input, a free system can given one of multiple (infinite?) outcomes.

No, I am not saying that, although with what you have taken to be "free" above I can see how you might say it.
With (probabilistic) random outcomes I still see it not being free, even though it could theoretically end up in a number of outcomes (per the example above).
But I'm finding it difficult to put into words what is clearly an additional element to the notion, such that it rules out random selection as being "free".

The same is true for me.

Perhaps not a strict definition, but some description of what the concept entails would be required for a proper discussion. But I am fully aware of the difficulty of defining "free will" in a way that makes intuitive sense.

I think we all have an intuitive idea of what the concept entails, do we not?
Something like the ability to make conscious decisions, to choose from between multiple possible outcomes according to whatever whims, desires and thoughts cross our mind at the time.
Or some such fairly broad language.
But good luck in unpacking that.

I do tend to include the phrase "the appearance of..." because without that you are effectively asserting that what you see is how it is.
That if it appears to be free then it is.
And I think if you start from there you will never establish whether it is truly free or not.
But then if you also define "free" merely by its appearance....

 

[NotEinstein]

If that is what I have done it is inadvertent.
I see "must" and "free" effectively as opposites.
One can define "must" as obligated, lacking ability to do something else, required to, etc.

That makes sense.

I'm not sure I'd agree.
I don't see a random system as necessarily being any freer than a deterministic one.
At least not in the sense I am using the term here.

I could be wrong, but didn't you introduce "randomness" as an example of a system that doesn't do as it "must"? (And thus considered "free"?)

Imagine a prisoner that gets let outside to exercise one hour a day.
In a deterministic world it would be the same hour each and every day.
In a random world it would be a different hour each day, with the hour decided either by some unknown or random component.
Is the prisoner ever "free" on the matter of which hour they are let out?

No, indeed not. But can you then give an example of a "free" system? Because if there can be no such system, then my point about "free" being a useless definition comes up again.

No, I am not saying that, although with what you have taken to be "free" above I can see how you might say it.
With (probabilistic) random outcomes I still see it not being free, even though it could theoretically end up in a number of outcomes (per the example above).
But I'm finding it difficult to put into words what is clearly an additional element to the notion, such that it rules out random selection as being "free".

Yes, I suspect that's the same difficulty I announced when I entered this thread. For some reason, defining "free" seems to be a tough nut to crack.

I think we all have an intuitive idea of what the concept entails, do we not?

Absolutely, which makes the inability to define it satisfactorily all the more frustrating.

Something like the ability to make conscious decisions, to choose from between multiple possible outcomes according to whatever whims, desires and thoughts cross our mind at the time.
Or some such fairly broad language.
But good luck in unpacking that.

I sense a "what is consciousness" coming up.

I do tend to include the phrase "the appearance of..." because without that you are effectively asserting that what you see is how it is.
That if it appears to be free then it is.
And I think if you start from there you will never establish whether it is truly free or not.
But then if you also define "free" merely by its appearance....

The problem with appearances is that they are fundamentally subjective, so people will end up with different interpretations of the definition of "free". But then again, that's no worse than the current situation.

 

[iceaura]

I see "must" and "free" effectively as opposites.

Engineers don't.
There is another intuition available - it derives from noticing that "must", on the level at which the will exists and effectively acts, has all the properties of a significant degree of intuitive freedom except the supernatural ability to contravene natural law.

"I think we all have an intuitive idea of what the concept entails, do we not?"
Absolutely, which makes the inability to define it satisfactorily all the more frustrating.

We are stuck on a fundamental difference in intuition, between those who require freedom to be the ability to contravene natural law and those who require it to embody an ability to make decisions and act in accordance with a person's decision criteria, moods, memories, dreams, immediate sources of information, and so forth - fully recognizing that whatever happens will be a physical event (rewindable in Newtonian terms, determined by physical law and cause/effect sequences, and so forth).