Could Have Done Otherwise - What does it mean exactly?

  • Thread starter moving finger
  • Start date
  • Tags
    Mean
In summary: Thanks for your input!In summary, according to libertarians, "free will" means that, no matter what one does, if the act was a free will act then one could have done otherwise (CHDO). This seems to be the libertarian definition of free will.
  • #1
moving finger
1,689
1
The phrase “Could Have Done Otherwise” (CHDO) often comes up in philosophical discussions on the existence or non-existence of “free will”. Many libertarians would claim “free will” means that, no matter what one does, if the act was a free will act then one could have done otherwise (CHDO) than what one actually did. This seems to be the libertarian definition of free will.

Firstly - Would any libertarians out there disagree with this? If yes, I would welcome your comments as to what you believe is the “correct” libertarian definition of free will.

Now to my question. What, exactly, are we to understand from the expression CHDO?

Does it mean CHDO in the counterfactual sense? In other words, does it mean :

A : CHDO = “If circumstances had been different, then it follows that I could have wanted to do something different, and it also follows that I could have done something different”

Or does it mean CHDO in some non-counterfactual (ie factual) sense, in the sense of either B or C below :

B : CHDO = “If circumstances had been exactly the same, then I could still have wanted to do something different, and it follows that I could have done something different”

or even :

C : CHDO = “If circumstances had been exactly the same and I had not wanted to do something different, then I could still have done something different”

Or does it mean something else entirely?

My understanding of libertarian free will is that the libertarian would reject both (A) and (C) above as explanations of the meaning of the phrase CHDO. Why? Because (A) is completely compatible with a deterministic account (in other words, we do not need to posit libertarian free will in order to claim counterfactual ability to have done otherwise under determinism); and because (C) is irrational – what rational reason could I have for suggesting that I could have done something different if I had not wanted to do something different?

This seems to leave us with (B) above as the meaning of CHDO according to a libertarian.

I welcome comments or criticism at this stage (before developing the idea further)

Best Regards
 
Last edited:
Physics news on Phys.org
  • #2
Well it seems that you tie in the concept of free will which is really one of physics with politics.
 
  • #3
MeJennifer said:
Well it seems that you tie in the concept of free will which is really one of physics with politics.
politics?

There are two meanings of libertarian - one is philosophical and one is political. This is a philosophy thread... I wasn't referring to political libertarians (I guess I should have made that clear).

See here for one discussion of libertarian free will in the philosophical as opposed to political sense :

http://www.morris.umn.edu/academic/philosophy/saylesdefense.html

Best Regards
 
  • #4
Some people claim to have difficulty understanding Alternative Possibilities. Here is a formal definition.
Let S1 be the state of the universe at time T1, and L the laws of the universe. According to indeterminism, there is (in general) a set of more than one future states of the universe {S21, S22, S23...} , which are all possible in that they do not contradict the conjunction of S1 and L. <poss>(L AND S1 AND S21)

<poss>(L AND S1 AND S22)

<poss>(L AND S1 AND S23)

According to determinism , the set {S2..} has only one member ... there is only one state which is compatible with the conjunction of S1 and L (being the only possibility, it is also actual and ncessary). [nec](L AND S1 AND S21)

NOT(L AND S1 AND S22)

In standard single-world metaphysics, only one possibillity can actually occur. The other members of the set are "alternative possibilities". Thus, if S22 is the possibility that occurs, S21 and S23 are the AP's in that case.
 
  • #5
B : CHDO = “If circumstances had been exactly the same, then I could still have wanted to do something different, and it follows that I could have done something different”

A bit simplistic, but basically right. We often have competing desires
and aims, the weightings or evaluations we attach to them
are themselves subject to evolution over time.

C : CHDO = “If circumstances had been exactly the same and I had not wanted to do something different, then I could still have done something different”

Too simplisitc. It assumes you only have one "want" in mind at a time.
 
  • #6
Hi Tournesol

Tournesol said:
Let S1 be the state of the universe at time T1, and L the laws of the universe. According to indeterminism, there is (in general) a set of more than one future states of the universe {S21, S22, S23...} , which are all possible in that they do not contradict the conjunction of S1 and L.
Many thanks for this post – very interesting.

One small point : Do you consider the "laws of the universe" as you describe above to be prescriptive or descriptive laws?

All three interpretations (A, B and C) of CHDO that I have suggested would seem to be compatible with the above analysis of alternative possibilities – would you agree?

A is compatible with both determinism and indeterminism, whereas B and C are compatible only with some form of indeterminism – would you agree?

Do you have any views on which of A, B or C (if any) a (philosophical) libertarian would accept as a correct interpretation of CHDO?

Best Regards
 
  • #7
moving finger said:
B : CHDO = “If circumstances had been exactly the same, then I could still have wanted to do something different, and it follows that I could have done something different”
Tournesol said:
A bit simplistic, but basically right. We often have competing desires and aims, the weightings or evaluations we attach to them are themselves subject to evolution over time.
I would be happy for you to suggest a better description if you consider this too simplistic.

If you are suggesting that the “weightings or evaluations” change between the two circumstances in question, then this would not conform to the condition that circumstances are exactly the same in both cases.

“Circumstances exactly the same” means, in effect, that if in the original case we have state S1 and laws L (to use your terminology from post #4), then in the second case we must once again have state S1 and laws L – there is no possibility for “weightings or evaluations” to change between the two cases (since the weightings and evaluations are part of either S1 or L).

Interpretation B of CHDO entails that S1 and L are identical in both cases (the original case and the “could have done otherwise” case). As far as I can see, the only way that this could give rise to a different outcome (following your analysis of alternative possibilities) is if the outcome is a result of some form of indeterminism in the process which leads from antecedent states (the circumstances defined by S1 and L) to "what we want".

Best Regards
 
Last edited:
  • #8
A SFW seems to be an arbitrary way of deciding which reasons ultimately take precedence, similar to rolling dice. The odds may be calculated, but in the end it all comes down to luck, and nothing more.

A self-forming action isn't "just random" -- instead it is connected
to a persons character. However the connection works
forwards, not backwards; the action forms future character.

The argument that libertarianism must be false because
it cannot "explain" people's actions is only valid if there
is a realistic alternative. But there isn't -- and there
isn't for reasons of human psyhcology, not the way
the universe works. People just can't give totally
bulletproof chains of explanations for their
actions that go all the way to the BB. Sooner
or later the chain of explanations peters out in somethign
that is just assumed. something that might as well be indeterminsitic.
 
  • #9
moving finger said:
One small point : Do you consider the "laws of the universe" as you describe above to be prescriptive or descriptive laws?

I mean the "real" laws, or an ideal description thereof.

All three interpretations (A, B and C) of CHDO that I have suggested would seem to be compatible with the above analysis of alternative possibilities – would you agree?

No. I am assuming that some set of circumstances has
occurred and is fixed. The A interpretation amounts
to the claim that the history of the universe could have been
different. That is different from the claim that more than one thing could
have happened under a different set of circumstance.

of course, A can't be translated into a power psosessed by
an agent.
Do you have any views on which of A, B or C (if any) a (philosophical) libertarian would accept as a correct interpretation of CHDO?

B is the closest
 
Last edited:
  • #10
moving finger said:
If you are suggesting that the “weightings or evaluations” change between the two circumstances in question, then this would not conform to the condition that circumstances are exactly the same in both cases.

It could conform to the external circumstances being the same.

“Circumstances exactly the same” means, in effect, that if in the original case we have state S1 and laws L (to use your terminology from post #4), then in the second case we must once again have state S1 and laws L – there is no possibility for “weightings or evaluations” to change between the two cases (since the weightings and evaluations are part of either S1 or L).

Then talk about S1' which is the state of everything outside my CNS.

Interpretation B of CHDO entails that S1 and L are identical in both cases (the original case and the “could have done otherwise” case). As far as I can see, the only way that this could give rise to a different outcome (following your analysis of alternative possibilities) is if the outcome is a result of some form of indeterminism in the process which leads from antecedent states (the circumstances defined by S1 and L) to "what we want".

Naturalistic FW is based on indeterminism, but to avoid the Burridan's
Ass problem it needs to operate over some span of time.
 
  • #11
Suppose we have 5 events, A, B, C, D and E where each event can & will cause one of the other events once it is triggered (an event triggers another in a deterministic, non-random manner, albeit possibly through a complex process). It's a cycle. Once one of A, B, C, D and E is triggered this cycle follows a path of events, maybe A, B, C, A, D, A, E, B ... etc.
Clearly, this cycle is autonomous. Like an agent with free will the cycle alone determines its own behavior. On the other hand the path the cycle takes is completely deterministic. Is this enough to say the cycle has free will?
What if each of A, B, C, D and E were to randomly trigger each other? Then the cycle is both autonomous and non-deterministic. Does it have free will then?
 
  • #13
Hi MF. I have no comment on your OP directly. I'm curious though what you feel are the implicit assumptions you're making. Specifically, it seems to me you're only concerning yourself with weakly emergent phenomena and you've ruled out strongly emergent ones, if in fact anything like strong emergence can exist.

In short, the idea might be that strongly emergent phenomena which evolves within some mechanism or system, might causally affect a subset of that system, however that causal affect might not be deterministic. It seems to me this might more closely match the perception we have of "free will".

For a definition of "weak emergence" I'll quote the guy everyone else seems to be quoting, Bedau.
Weak emergence applies in contexts in which there is a system, call it S, composed out of "micro-level" parts; the number and identity of these parts might change over time. S has various "macro-level" states (macrostates) and various "micro-level" states (microstates). S's microstates are the intrinsic states of its parts and it's macrostates are structural properties constituted wholly out of microstates. Interesting macrostates typically average over microstates and so compresses microstate information. Further, there is a microdynamic, call it D, which governs the time evolution of S's microstates. Usually the microstate of a given part of a system at a given time is a result of the microstates of "nearby" parts of the system at preceding times; in this sense, D is "local".

And for "strong emergence" I'll quote http://consc.net/papers/emergence.pdf" .
We can say that a high-level phenomenon is strongly emergent with respect to a low-level domain when the high-level phenomenon arises from the low-level domain, but truths concerning that phenomenon are not deducible even in principal from truths in the low-level domain.



The best way of thinking of this sort of possibility is as involving a sort of downward causation. Downward causation means that higher-level phenomena are not only irreducible but also exert a causal efficacy of some sort. Such causation requires the formulation of basic principals which state that when certain high-level configurations occur, certain consequences will follow. (These are what McLaufhlin (1993) calls configurational laws.)



To be clear, one should distinguish strong downward causation from weak downward causation. With strong downward causation, the causal impact of a high-level phenomenon on low-level processes is not deducible even in principal* from initial conditions and low-level laws. With weak downward causation, the causal impact of high-level phenomenon is deductible in principal, but is nevertheless unexpected.
*where Chalmers states "not deducible even in principal" I take to mean "not onticaly determinable"

If we assume only weakly emergent phenomena exist, then we can model any of those phenomena as if there are local causal actions acting on each 'microstate'. Those causal actions might be deterministic or random. It seems to me, this is one of the implicit assumptions you're making in your OP.

On the other hand, if we assume there is such a thing as strong emergence and strong downward causation, I believe what's being suggested is that some emergent phenomena can have a causal affect on some set of microstates which is neither deterministic nor uncaused.

Let's look at computationalism for a second: IMHO, computationalism can only result in weakly emergent phenomena. The actions of any computational device can be determined from the "microdynamics" of the system as Bedau would put it. For example, a switch in a conventional digital computer acts in a completely deterministic way and is only affected by the local affects of electrical potentials that act on it. There is no downward causation which could possibly change the state of a switch, it is only affected by the local microdynamics.

If one suggests consciousness arises from the act of computation, and thus it is "strongly emergent" then I'd have to say that such a claim is a trivial one in the sense that it is not needed, even if it is real, to explain the action of the computer or system. Why invoke "strong emergence" when the actions of a computer can easily be explained by the actions of the switches or other parts of the device? Certainly those subjective experiences have no causal impact on any portion of any computational device. They can't. No part of the computer is susceptible to influences beyond the local ones acting on each portion of the device. The only benefit (if you can call it a benefit) gained by suggesting consciousness arises from the act of computation is an explanation for subjective feelings, but the system has no additional capability or features which arise from this allegedly strongly emergent phenomenon of subjective experience. Thus the need for 'downward causation' for any strongly emergent view of consciousness.

If downward causation is conceived of as being a feature of consciousness then, we might also suggest that there are influences created by the phenomena which are not deterministic, but they do result from a cause. Thus, for "CHDO", we might suggest that there is another definition which isn't covered by the OP because of the implicit assumptions made. We might suggest that CHDO arises from a strongly emergent phenomenon which causes a change in microstates through downward causation, but that this change in microstates, although caused, is not ontically determinable. I think this kind of "free will" is much closer to what we actually experience, and also provides for such things as "moral responsibility".

I think another alternative is to suggest downward causation will create only a single, very specific and ontically deterministic change in the system or mechanism. In other words, if there is such a thing as downward causation, one might suggest that given some state of a mechanism or system, that state results in a deterministic change of state to some portion of that system or mechanism. Such a definition of downward causation seems a bit awkward though as it almost seems like it isn't downward causation at that point since this says that a set of microstates will result in some deterministic microdynamic. I haven't thought about alternative enough though. At any rate, I wonder if any of this gives you some ideas about what the implicit assumptions are in your OP.
 
Last edited by a moderator:
  • #14
Tournesol said:
A self-forming action isn't "just random" -- instead it is connected to a persons character. However the connection works forwards, not backwards; the action forms future character.
(Note : Self Forming Action / Self Forming Will are terms borrowed from Kane’s treatment in The Oxford Handbook of Free Will)

This is inconsistent, or incomplete. You claim SFAs (self forming actions) are not random, but at the same time that SFAs form character and not the reverse. In other words (according to your account), character supervenes on SFAs, SFAs do not supervene on character. What, then, forms SFAs (if SFAs are not random)? In other words, SFAs are either regular (have regular relationships with some kind of antecedent states) or irregular (no relationships with any antecedent states). If irregular then SFAs would be random.

Tournesol said:
The argument that libertarianism must be false because it cannot "explain" people's actions is only valid if there is a realistic alternative. But there isn't -- and there isn't for reasons of human psyhcology, not the way the universe works. People just can't give totally bulletproof chains of explanations for their actions that go all the way to the BB. Sooner or later the chain of explanations peters out in somethign that is just assumed. something that might as well be indeterminsitic.
Nobody has claimed (so far) in this thread that libertarianism is false. I am simply trying to understand what exactly a libertarian means when he says he could have done otherwise.

If I were to argue against the concept of libertarian free will, then I would argue not that it is false, but that (if fully explicated, which not many people apart from your good self seems prepared to do) it either leads simply to arbitrary behaviour, or else the very notion is either inconsistent or incoherent.

You claim that there is no realistic alternative to libertarianism. I disagree. That “people just can't give totally bulletproof chains of explanations for their actions that go all the way to the BB” is a reflection of the limits of our epistemology (what we can know about the world), it is not necessarily a reflection of lack of determinism. Chaos (the mathematical kind) and the HUP already place limits on what we can know about the world (in other words, even if the world were totally deterministic we still would not be able to know everything because of the limits to our knowledge imposed by chaos and HUP). I agree with your comment that “sooner or later the chain of explanations peters out in somethign that is just assumed. something that might as well be indeterminsitic”, and indeed the Big Bang itself might have been a quantum-indeterministic event, and the world may operate according to quantum-stochastic (but not totally random) principles rather than deterministic principles. If the libertarian simply wishes to equate his concept of “free will” with stochastic behaviour as opposed to deterministic behaviour then I have no problem with this – but there is no more autonomy or “will” involved in stochasticity than there is involved in determinism, and stochasticity entails much less control over one’s actions than does determinism.

Tournesol said:
A can't be translated into a power psosessed by an agent.
Agreed.

Tournesol said:
B is the closest
Thank you.

Tournesol said:
It could conform to the external circumstances being the same.
Okay, I understand. You are taking the weightings and evaluations to be internal to the “will”, and not included in the external circumstances. The problem with this is that the weightings and evaluations are part of the antecedent states which “determine” the will (or the outcome of the will), hence must be considered part of the circumstances.

We need to compare apples with apples.

What determinism says is that given two perfectly identical worlds (identical not only in the physical, but also in all other senses, the “soul” (if such exists) and “will” sense) the subsequent evolution of these worlds will also be identical. In other words, under determinism the “will” has no power to choose that one world will be any different to the other – all is determined.

What the libertarian concept of CHDO seems to be saying is that, given these same two perfectly identical worlds, the subsequent evolution of these worlds will not necessarily be identical, that the “will” has some kind of power to “choose” that one world will be different to the other.

In other words, CHDO assumes that all circumstances (not only those external to the will) are identical in both cases. This includes any weightings or evaluations that the will may “use” in deciding what to do. In other words, anything which has any causal role in determining the outcome of the will must be included in the fixed antecedent “circumstances”.

Tournesol said:
Naturalistic FW is based on indeterminism, but to avoid the Burridan's Ass problem it needs to operate over some span of time.
How does this avoid the problem? Surely Burridan’s Ass had all the time in the world to decide which way to go?

-Job- said:
Suppose we have 5 events, A, B, C, D and E where each event can & will cause one of the other events once it is triggered (an event triggers another in a deterministic, non-random manner, albeit possibly through a complex process). It's a cycle. Once one of A, B, C, D and E is triggered this cycle follows a path of events, maybe A, B, C, A, D, A, E, B ... etc.
Clearly, this cycle is autonomous. Like an agent with free will the cycle alone determines its own behavior. On the other hand the path the cycle takes is completely deterministic. Is this enough to say the cycle has free will?
By “autonomous” here I assume you mean “acting with free will”.

As you said, the path is deterministic. In what sense therefore could you claim that this series was autonomous? It’s about as autonomous as a computer program.

-Job- said:
What if each of A, B, C, D and E were to randomly trigger each other? Then the cycle is both autonomous and non-deterministic. Does it have free will then?
Why should random behaviour result in autonomy? If my computer program is deterministic and has no autonomy, how does it suddenly become autonomous simply by virtue of introducing some random element into its calculations? Random elements will simply result (if they result in anything at all) in random or stochastic behaviour, there is not necessarily any autonomy involved.

The problem with the AP idea is that, to have any value at all, an AP must be under the control of the will – the will must be able to select which AP will occur. But if APs occur according to some kind of random process then the will has no more control (in fact it has even less control) over the eventual outcome than if the process had been entirely deterministic.

Best Regards
 
Last edited:
  • #15
Hi Q_Goest

Q_Goest said:
Hi MF. I have no comment on your OP directly. I'm curious though what you feel are the implicit assumptions you're making. Specifically, it seems to me you're only concerning yourself with weakly emergent phenomena and you've ruled out strongly emergent ones, if in fact anything like strong emergence can exist.
I’m not assuming anything specific about emergent phenomena, except that I am assuming that however the world operates, determinism is either true or false in that world.

Q_Goest said:
In short, the idea might be that strongly emergent phenomena which evolves within some mechanism or system, might causally affect a subset of that system, however that causal affect might not be deterministic. It seems to me this might more closely match the perception we have of "free will".
If not deterministic then it seems to me the evolution of the system must be either random (showing no regularities whatsoever) or stochastic (showing probabilistic regularities).

Q_Goest said:
The best way of thinking of this sort of possibility is as involving a sort of downward causation. Downward causation means that higher-level phenomena are not only irreducible but also exert a causal efficacy of some sort. Such causation requires the formulation of basic principals which state that when certain high-level configurations occur, certain consequences will follow. (These are what McLaufhlin (1993) calls configurational laws.)
Why can we not apply the same 3 alternative notions of CHDO to such downward causation? In such a case, the “circumstances” includes the higher-level configurations that you refer to. In other words, given a particular state of the world, including the higher-level configurations of any downward-causation phenomena in that world, are there any alternative possibilities available, or is the future determined?

Think of it in terms of comparing two parallel and identical worlds, where all the circumstances including higher-level configurations of any downward-causation phenomena, are precisely replicated in both worlds. Determinism would say that both worlds will necessarily evolve with time in an identical fashion, whereas indeterminism would say that the worlds would not necessarily evolve with time in an identical fashion. This would be true whether downward-causation phenomena were present or absent.

We need to be careful not to confuse determinable (what you refer to as deducible) with deterministic. A system can be deterministic without being determinable. Whether “causal impacts” are deducible (= determinable) in principle or not is not necessarily relevant to the question of whether the universe is deterministic or not. A 100% deterministic world may be not deducible (ie not predictable or not determinable) by virtue of (for example) Heisenberg’s Uncertainty Principle, and/or mathematical chaos, and/or self-referentiality.

In summary, I believe that in any given system, whether it includes downward-causation phenomena or not, determinism will be either true or false. If determinism is false, then we say that the system is indeterministic (where here we take indeterministic to include stochastic), and this gives rise to “alternative possibilities” or “could have done otherwise”. If we have two parallel and identical worlds, then if determinism is true these worlds will evolve identically in all respects (no alternative possibilities, and no could have done otherwise), whereas if determinism is false then these worlds will not necessarily evolve identically in all respects (alternative possibilities exist and so too does could have done otherwise).

Best Regards
 
Last edited:
  • #16
Hi Movingfinger. I think you're still clutching to a very specific model of the world that may or may not be correct. People often think in terms of any given phenomena being deterministic, stochastic or random, but this model may only be applicable to weak emergence or "upward causation".

From Campbell: In short, the difficulty arises from combining 'upward determiniation' with 'downward causation'.
Campbell seems to recognize the same problems you're having with CHDO. I'd suggest reading http://www.lehigh.edu/~mhb0/physicalemergence.pdf" for more on this. They also quote J. Kim "Mind in a Physical World" MIT Press, 1998 which it seems from reading Campbell, is applicable to what you're searching for. It seems their arguments follow the same path I've already provided above.

What I believe is being suggested is that downward causation, created by a mental property (M-property in the literature), can cause a change in a physical property (P-property) which can not be explained, nor predicted, even in principal. They would appear random, and in affect they would have the capability of producing two different, parallel worlds. But would you call such an event random or even stochastic if it is caused by some mental property? You could simply suggest it is random, but I'd argue that such a conclusion doesn't suffice because any change in state which is caused can't be random. On the other hand, it can't be deterministic either since there can be more than one possible outcome.

I realize this too is only a possible model of the world, just as there can be others and I don't know how it fits in with the "libertarian definition of free will". If you're only looking to understand the libertarian perspective, I think one can only recognize that such perspectives may contain implicit assumptions which may or may not be valid. However, strong emergence and downward causation has some very interesting aspects to it that may leave you head scratching. As I've mentioned previously, the concept has obviously already been considered. Suggesting strongly emergent phenomena arise from computationalism is "superfluous" and "causally impotent" which I believe is what Alexander is suggesting when he says:

It supposes something to exist in nature which has nothing to do, no purpose to serve, a species of noblesse which depends on the work of its inferiors, but is kept for show and might as well, and undoubtedly would in time, be abolished.

That seems like one of the most damning condemnations of computationalism I've ever heard.
 
Last edited by a moderator:
  • #17
moving finger said:
By “autonomous” here I assume you mean “acting with free will”.

Not at all, actually. By autonomous i meant that, from a strictly physical viewpoint, the cycle alone is responsible for it's behavior, which is the case. There's no notion of a choice in here, especially since the path the cycle will take is deterministic. That the cycle doesn't have a "choice", and yet it is autonomous is what i was proposing.

As you said, the path is deterministic. In what sense therefore could you claim that this series was autonomous? It’s about as autonomous as a computer program.

It would be as autonomous, possibly, as a human.

Why should random behaviour result in autonomy? If my computer program is deterministic and has no autonomy, how does it suddenly become autonomous simply by virtue of introducing some random element into its calculations?

I was calling it autonomous before introducing randomness into the calculations, you misunderstood that.
 
  • #18
Hi Q-Goest

Q_Goest said:
Campbell seems to recognize the same problems you're having with CHDO. I'd suggest reading http://www.lehigh.edu/~mhb0/physicalemergence.pdf" for more on this. They also quote J. Kim "Mind in a Physical World" MIT Press, 1998 which it seems from reading Campbell, is applicable to what you're searching for. It seems their arguments follow the same path I've already provided above.
Thank you for the reference - I shall read it with great interest.

Q_Goest said:
What I believe is being suggested is that downward causation, created by a mental property (M-property in the literature), can cause a change in a physical property (P-property) which can not be explained, nor predicted, even in principal.
It seems that you are perhaps confusing determinability (whether something can be explained/predicted, even in principle) with determinism (whether, given antecedent states of the world, there is one and only one possible outcome). A world may be 100% deterministic (ie determinism is true in that world) without being determinable (even in principle). Limits to determinability in our own world include chaos, Heisenberg's uncertainty principle, and the in-principle indeterminability of self-referential systems - but none of these in-principle limits to determinability implies that determinism is necessarily false in our world.

Q_Goest said:
They would appear random, and in affect they would have the capability of producing two different, parallel worlds.
"Appearing random" is an epistemic property, it does not follow from a random appearance that the world is (ontically) random, neither does it follow that more than one outcome is possible.

Q_Goest said:
But would you call such an event random or even stochastic if it is caused by some mental property?
It depends on whether the cause (the mental property) is itself caused or uncaused. Is the so-called mental event supposed itself to be caused, or uncaused?

If the mental event is itself uncaused (ie spontaneous) then presumably it's behaviour is either random or stochastic - if not, what else could it be?

If the mental event is itself caused, then move the question back to the cause of the mental event.

Q_Goest said:
You could simply suggest it is random, but I'd argue that such a conclusion doesn't suffice because any change in state which is caused can't be random.
Of course it can. If X is a random event, and X causes Y, then Y is also random (since it depends on X). Thus Y is caused but is nevertheless random (because the causal chain originates in a random event)

Q_Goest said:
On the other hand, it can't be deterministic either since there can be more than one possible outcome.
How do you know there can be more than one possible outcome? For this to be true, you must assume determinism is false in the first place.

Q_Goest said:
I realize this too is only a possible model of the world, just as there can be others and I don't know how it fits in with the "libertarian definition of free will". If you're only looking to understand the libertarian perspective, I think one can only recognize that such perspectives may contain implicit assumptions which may or may not be valid. However, strong emergence and downward causation has some very interesting aspects to it that may leave you head scratching.
I have already looked into downward cauation some time back, and found nothing significant which (imho) would have any real bearing on free will. But I will take a look at the Campbell & Bickhard paper and get back to you.

Best Regards
 
Last edited by a moderator:
  • #19
-Job- said:
Not at all, actually. By autonomous i meant that, from a strictly physical viewpoint, the cycle alone is responsible for it's behavior, which is the case. There's no notion of a choice in here, especially since the path the cycle will take is deterministic. That the cycle doesn't have a "choice", and yet it is autonomous is what i was proposing.
OK, I see what you mean. However I think that most libertarians would claim that any system which has no "choice" in the way that it acts is not autonomous. It comes down to the definition of autonomous, which a libertarian would claim entails ability to choose, whereas you do not. (Here I am also assuming that choice is defined in the libertarian sense of free will choice).

-Job- said:
It would be as autonomous, possibly, as a human.
I tend to agree - because I do not believe either a human or the system described is autonomous (in the libertarian sense of the word)

Best Regards
 
  • #20
MF said:
Tournesol said:
A self-forming action isn't "just random" -- instead it is connected to a persons character.
However the connection works forwards, not backwards; the action forms future character.
(Note : Self Forming Action / Self Forming Will are terms borrowed from Kane’s treatment in The Oxford Handbook of
Free Will)

This is inconsistent, or incomplete. You claim SFAs (self forming actions) are not random, but at the same time that
SFAs form character and not the reverse.

They are not random in the way that making decisions according to
a die is random. Such decisions have no causes and no consequences either.
However, is still true that SFA's have no *prior* causes.
There is more than one sense of "random" here.
In other words (according to your account), character supervenes on SFAs,
SFAs do not supervene on character.

It's not that black-and-white.
What, then, forms SFAs (if SFAs are not random)? In other words, SFAs are either
regular (have regular relationships with some kind of antecedent states) or irregular (no relationships with any
antecedent states). If irregular then SFAs would be random.

Not in the way a die-throw is random. They have a relationship
to character, but a forwad-firing one.
The argument that libertarianism must be false because it cannot "explain" people's actions is only valid if there is a
realistic alternative. But there isn't -- and there isn't for reasons of human psyhcology, not the way the universe works.
People just can't give totally bulletproof chains of explanations for their actions that go all the way to the BB.
Sooner or later the chain of explanations peters out in somethign that is just assumed. something that might as well be
indeterminsitic.

Nobody has claimed (so far) in this thread that libertarianism is false. I am simply trying to understand what exactly a
libertarian means when he says he could have done otherwise.

If I were to argue against the concept of libertarian free will, then I would argue not that it is false, but that
(if fully explicated, which not many people apart from your good self seems prepared to do) it either leads simply to
arbitrary behaviour, or else the very notion is either inconsistent or incoherent.

The "arbitrary behaviour" complaint is just the same point again. Libertarian FW is only
arntirary in the senses of not being able to give totally bulletproof chains of explanations for actions that go all
the way to the BB.
You claim that there is no realistic alternative to libertarianism. I disagree. That “people just can't give totally
bulletproof chains of explanations for their actions that go all the way to the BB” is a reflection of the limits of
our epistemology (what we can know about the world), it is not necessarily a reflection of lack of determinism.

No, but so what ? You seem to have confuse causes with reasons. The point of insisting that
actions must have causes is to ensure that they have reasons; rationality is the payoff,
causality is a means to the end. Crazy, irrational actions have casues as much as any other. Causality
is not part of a version of FW worth wanting, raionallity is.

Chaos (the mathematical kind) and the HUP already place limits on what we can know about the world (in other words,
even if the world were totally deterministic we still would not be able to know everything because of the limits to our
knowledge imposed by chaos and HUP). I agree with your comment that “sooner or later the chain of explanations
peters out in somethign that is just assumed. something that might as well be indeterminsitic”, and indeed the Big Bang
itself might have been a quantum-indeterministic event, and the world may operate according to quantum-stochastic
(but not totally random) principles rather than deterministic principles. If the libertarian simply wishes to equate
his concept of “free will” with stochastic behaviour as opposed to deterministic behaviour then I have no problem with
this – but there is no more autonomy or “will” involved in stochasticity than there is involved in determinism,
and stochasticity entails much less control over one’s actions than does determinism.

The control lies in the SIS's selection of proposal from the RIG.
It could conform to the external circumstances being the same.

Okay, I understand. You are taking the weightings and evaluations to be internal to the “will”, and not included in
the external circumstances. The problem with this is that the weightings and evaluations are part of the antecedent
states which “determine” the will (or the outcome of the will), hence must be considered part of the circumstances.
No. By "circumstances" I mean both what is temporally prior, and what is "outside the head".
What the libertarian concept of CHDO seems to be saying is that, given these same two perfectly identical worlds,
the subsequent evolution of these worlds will not necessarily be identical, that the “will” has some kind of
power to “choose” that one world will be different to the other.
In other words, CHDO assumes that all circumstances (not only those external to the will) are identical in both
cases. This includes any weightings or evaluations that the will may “use” in deciding what to do. In other words,
anything which has any causal role in determining the outcome of the will must be included in the fixed
antecedent “circumstances”.
No, libertaraian CHDO deals only with identical external circumstances.
The private thoughts in your head can vary.
Naturalistic FW is based on indeterminism, but to avoid the Burridan's Ass problem it needs to operate over some
span of time.
How does this avoid the problem? Surely Burridan’s Ass had all the time in the world to decide which way to go?
Its decision will be disconneted from its prior state of mind.

Why should random behaviour result in autonomy? If my computer program is deterministic and has no autonomy, how does
it suddenly become autonomous simply by virtue of introducing some random element into its calculations? Random elements
will simply result (if they result in anything at all) in random or stochastic behaviour, there is not necessarily
any autonomy involved.

If "autonomy" means that the output of a systems cannot be predicted from
a complete specification of its input , then random behaviour *does* mean
autonomy.

If autonomy involves rationallity (to a realistic extent and under appropriate circumstances)
then a SIS will be needed as well as a RIG!
The problem with the AP idea is that, to have any value at all, an AP must be under the control of the will – the will
must be able to select which AP will occur. But if APs occur according to some kind of random process then the will
has no more control (in fact it has even less control) over the eventual outcome than if the process had been
entirely deterministic.
The SIS can filter the output of the RIG.
 
  • #21
Tournesol said:
They are not random in the way that making decisions according to a die is random. Such decisions have no causes and no consequences either. However, is still true that SFA's have no *prior* causes. There is more than one sense of "random" here.
Could you explain what other sense of “random” there is?

Tournesol said:
It's not that black-and-white.
OK, then what is it?

Tournesol said:
Not in the way a die-throw is random. They have a relationship to character, but a forwad-firing one.
Can you explain what the relationship to character is?

Tournesol said:
The "arbitrary behaviour" complaint is just the same point again. Libertarian FW is only arntirary in the senses of not being able to give totally bulletproof chains of explanations for actions that go all the way to the BB.
No, it is arbitrary (in the so-called Darwinian model) in the sense that some causal chains originate in random events. Any causal chain that has an origination in a random event results either in arbitrary consequences, or in deterministic consequences. There is no “third way” that is neither arbitrary nor deterministic (where here we take arbitrary consequences to include stochastic consequences).

Tournesol said:
No, but so what ? You seem to have confuse causes with reasons.
I am not confusing anything – I am responding to your claim that the fact “people just can't give totally bulletproof chains of explanations for their actions that go all the way to the BB” allegedly shows there is no realistic alternative to libertariansim. It shows nothing of the kind.

Tournesol said:
The point of insisting that actions must have causes is to ensure that they have reasons; rationality is the payoff, causality is a means to the end. Crazy, irrational actions have casues as much as any other. Causality is not part of a version of FW worth wanting, raionallity is.
Causality is very much a part of any free will worth wanting, because without causality we have randomness. I want to determine my actions, I don’t want my actions to be random in any way.

Tournesol said:
The control lies in the SIS's selection of proposal from the RIG.
The SIS does not control which ideas are thrown up by the RIG – these are generated at random. The final outcome of the RIG plus SIS is either deterministic or random, or an unpredictable mixture of the two, depending upon just how the RIG and SIS are configured – you cannot generate free will (as understood by most libertarians) by simply mixing random processes with deterministic processes.

Tournesol said:
No, libertaraian CHDO deals only with identical external circumstances. The private thoughts in your head can vary.
If the private thoughts in my head can vary, and these thoughts are part of the antecedent causes of my behaviour, then of course I can do otherwise – even in a deterministic world. The whole point of CHDO is that the precise circumstances, including “private thoughts in my head” must be identical in both cases. If we allow that there can be differences between the two cases then we don’t need free will to make them different, do we?

Tournesol said:
If "autonomy" means that the output of a systems cannot be predicted from a complete specification of its input , then random behaviour *does* mean autonomy.
That “the output of a systems cannot be predicted from a complete specification of its input” is (epistemically) random behaviour. Are you defining autonomy to be synonymous with random behaviour?

Tournesol said:
If autonomy involves rationallity (to a realistic extent and under appropriate circumstances) then a SIS will be needed as well as a RIG!
This simply results in either deterministic or random behaviour, or an unpredictable mixture of the two.

Tournesol said:
The SIS can filter the output of the RIG.
Which makes the output either random or deterministic, or an unpredictable mixture of the two.

Best Regards
 
  • #22
They are not random in the way that making decisions according to a die is random. Such decisions have no causes and no
consequences either. However, is still true that SFA's have no *prior* causes. There is more than one sense of
"random" here.
Could you explain what other sense of “random” there is?

a) having no sufficient prior cause
b) having no prior cause, and no lasting effects.

Not in the way a die-throw is random. They have a relationship to character, but a forwad-firing one.
Can you explain what the relationship to character is?

self-forming.

The "arbitrary behaviour" complaint is just the same point again. Libertarian FW is only arntirary in the senses of not being able to give totally bulletproof chains of explanations for actions that go all the way to the BB.

No, it is arbitrary (in the so-called Darwinian model) in the sense that some causal chains originate in random events.

That it yet another re-statement of the same, already-answered, objection.

Any causal chain that has an origination in a random event results either in arbitrary consequences, or in
deterministic consequences. There is no “third way” that is neither arbitrary nor deterministic (where here we take
arbitrary consequences to include stochastic consequences).
You need to be talking about reasons, not causes. In what way does "arbitrariness"
impinge on FW ? Is strict deterministic causation any better ?
No, but so what ? You seem to have confuse causes with reasons.
I am not confusing anything – I am responding to your claim that the fact “people just can't give totally bulletproof chains of explanations for their actions that go all the way to the BB” allegedly shows there is no realistic alternative to libertariansim. It shows nothing of the kind.
Explain how strict determinism supports FW, then.

The point of insisting that actions must have causes is to ensure that they have reasons; rationality is the payoff,
causality is a means to the end. Crazy, irrational actions have casues as much as any other.
Causality is not part of a
version of FW worth wanting, raionallity is.

Causality is very much a part of any free will worth wanting, because without causality we have randomness.
The point is whether or not you would have rationality. You are confusing
causes with reasons.

I want to determine my actions, I don’t want my actions to be random in any way.

Then why don't you take yourslef along
to a hypnotist and get him to hypnotise you into
floowing a script of his own devising?

Or do you want to *freely* determine *your* actions, not just have
actions that *are* determined ?
The control lies in the SIS's selection of proposal from the RIG.
The SIS does not control which ideas are thrown up by the RIG
No. So what ? Are you saying you don't have FW becasue you
can't turn round to your RIG and demand an idea to
cure cancer ?

these are generated at random. The final outcome
of the RIG plus SIS is either deterministic or random, or an unpredictable mixture of the two, depending upon just
how the RIG and SIS are configured – you cannot generate free will (as understood by most libertarians) by simply
mixing random processes with deterministic processes.

That is exactly what libertarians who think about the
subject *do* think FW is. Any further option is just incomprehensible.
No, libertaraian CHDO deals only with identical external circumstances. The private thoughts in your head can vary.
If the private thoughts in my head can vary, and these thoughts are part of the antecedent causes of my behaviour,
then of course I can do otherwise – even in a deterministic world.

The private thoughts in your head can't vary in an indeterminsitic world. Or did you mean
external world ?

The whole point of CHDO is that the precise circumstances, including “private thoughts in my head” must be identical in both cases.

No. The defintion of FW only requires that your decisions are not
(entirely) caused externally. CHDO is not compatiblism -- it is not
an attempt to snatch freedom from the jaws of detemrinsm.

If we allow that there can be differences between the two cases then we don’t need free will to make them different, do we?

You have lapsed into the Basicness Assumption.

FW is *constituted* by indeterministic could-have-been-otherwise, in combination
with other factors. It is not *suposed* to eb something "extra".
If "autonomy" means that the output of a systems cannot be predicted from a complete specification of its input , then random behaviour *does* mean autonomy.
That “the output of a systems cannot be predicted from a complete specification of its input” is (epistemically) random behaviour. Are you defining autonomy to be synonymous with random behaviour?

It's not my schtick.

If autonomy involves rationallity (to a realistic extent and under appropriate circumstances) then a SIS will be needed as well as a RIG!
This simply results in either deterministic or random behaviour, or an unpredictable mixture of the two.

FW *is* a mixture of the two. Unless you make the Basicness Assumption.

The SIS can filter the output of the RIG.
Which makes the output either random or deterministic, or an unpredictable mixture of the two.

FW *is* a mixture of the two. Unless you make the Basicness Assumption.

Your original point was about control. I have answered that,
and now you have introduced a new point -- that
what I am describing is not FW because it is a higer-order
mixture of factors, and therefore not Basic.

The objection is arbitrary. Why should we regard FW that way ?
 
  • #23
MF said:
They are not random in the way that making decisions according to a die is random. Such decisions have no causes and no
consequences either. However, is still true that SFA's have no *prior* causes. There is more than one sense of
"random" here.
Could you explain what other sense of “random” there is?

a) having no sufficient prior cause
b) having no prior cause, and no lasting effects.

Not in the way a die-throw is random. They have a relationship to character, but a forwad-firing one.
Can you explain what the relationship to character is?

self-forming.

The "arbitrary behaviour" complaint is just the same point again. Libertarian FW is only arntirary in the senses of not being able to give totally bulletproof chains of explanations for actions that go all the way to the BB.

No, it is arbitrary (in the so-called Darwinian model) in the sense that some causal chains originate in random events.

That it yet another re-statement of the same, already-answered, objection.

Any causal chain that has an origination in a random event results either in arbitrary consequences, or in
deterministic consequences. There is no “third way” that is neither arbitrary nor deterministic (where here we take
arbitrary consequences to include stochastic consequences).
You need to be talking about reasons, not causes. In what way does "arbitrariness"
impinge on FW ? Is strict deterministic causation any better ?
No, but so what ? You seem to have confuse causes with reasons.
I am not confusing anything – I am responding to your claim that the fact “people just can't give totally bulletproof chains of explanations for their actions that go all the way to the BB” allegedly shows there is no realistic alternative to libertariansim. It shows nothing of the kind.
Explain how strict determinism supports FW, then.

The point of insisting that actions must have causes is to ensure that they have reasons; rationality is the payoff,
causality is a means to the end. Crazy, irrational actions have casues as much as any other.
Causality is not part of a
version of FW worth wanting, raionallity is.

Causality is very much a part of any free will worth wanting, because without causality we have randomness.
The point is whether or not you would have rationality. You are confusing
causes with reasons.

I want to determine my actions, I don’t want my actions to be random in any way.

Then why don't you take yourself along
to a hypnotist and get him to hypnotise you into
following a script of his own devising?

Or do you want to *freely* determine *your* actions, not just have
actions that *are* determined ?

It is for you to say how self-detemination
differs from outside determination ... in a deterministic universe.

The control lies in the SIS's selection of proposal from the RIG.
The SIS does not control which ideas are thrown up by the RIG
No. So what ? Are you saying you don't have FW becasue you
can't turn round to your RIG and demand an idea to
cure cancer ?

these are generated at random. The final outcome
of the RIG plus SIS is either deterministic or random, or an unpredictable mixture of the two, depending upon just
how the RIG and SIS are configured – you cannot generate free will (as understood by most libertarians) by simply
mixing random processes with deterministic processes.

That is exactly what libertarians who think about the
subject *do* think FW is. Any further option is just incomprehensible.
No, libertaraian CHDO deals only with identical external circumstances. The private thoughts in your head can vary.
If the private thoughts in my head can vary, and these thoughts are part of the antecedent causes of my behaviour,
then of course I can do otherwise – even in a deterministic world.

The private thoughts in your head can't vary in an indeterminsitic world. Or did you mean
external world ?

The whole point of CHDO is that the precise circumstances, including “private thoughts in my head” must be identical in both cases.

No. The defintion of FW only requires that your decisions are not
(entirely) caused externally. CHDO is not compatiblism -- it is not
an attempt to snatch freedom from the jaws of detemrinsm.

If we allow that there can be differences between the two cases then we don’t need free will to make them different, do we?

You have lapsed into the Basicness Assumption.

FW is *constituted* by indeterministic could-have-been-otherwise, in combination
with other factors. It is not *suposed* to eb something "extra".
If "autonomy" means that the output of a systems cannot be predicted from a complete specification of its input , then random behaviour *does* mean autonomy.
That “the output of a systems cannot be predicted from a complete specification of its input” is (epistemically) random behaviour. Are you defining autonomy to be synonymous with random behaviour?

It's not my schtick.

If autonomy involves rationallity (to a realistic extent and under appropriate circumstances) then a SIS will be needed as well as a RIG!
This simply results in either deterministic or random behaviour, or an unpredictable mixture of the two.

FW *is* a mixture of the two. Unless you make the Basicness Assumption.

The SIS can filter the output of the RIG.
Which makes the output either random or deterministic, or an unpredictable mixture of the two.

FW *is* a mixture of the two. Unless you make the Basicness Assumption.

Your original point was about control. I have answered that,
and now you have introduced a new point -- that
what I am describing is not FW because it is a higer-order
mixture of factors, and therefore not Basic.

The objection is arbitrary. Why should we regard FW that way ?
 
  • #24
Hi Q-Goest

Q_Goest said:
Campbell seems to recognize the same problems you're having with CHDO. I'd suggest reading http://www.lehigh.edu/~mhb0/physicalemergence.pdf" for more on this.
I've read the paper - very interesting, but most of the paper deals with physicalist attempts to explain emergence - there is only a very brief section on downward causation - and the focus seems to be largely arguments against reductionism, not arguments against determinism, and nothing in there that I can find seems to suggest any lack of determinism in emergent or downward-causation processes.

Perhaps you could explain how you think the paper is relevant to the issue of CHDO?

Best Regards
 
Last edited by a moderator:
  • #25
The Basicness Assumption.
A popular argument against free will has it that free will is incompatible with determinism (because of lack of Alternative Possibilities or "elbow room") and with indeterminism (because will is not "mere randomness"). This seems to neglect the alternative the free will is a judicious mixture of the determinism and indeterminism -- after all, we cannot infer "salt is not sodium chloride" from "salt is not sodium" and "salt is not chlorine", true as both those statements are.
Or perhaps there it is more than that ? Perhaps the determinist is making the assumption that free will is a basic ingredient to the universe, and using that as a reason to exclude the possibility that it is a composite, emergent phenomenon. This would certainly explains why Gordon Orloff says things like:- "if so, how and why doesn't everything in the universe -- atoms, cells, dogs, cars -- possesses this unnatural quality? [free will]" Why should they ? If free will is an outcome of the engineering of the brain, we would not expect to find it in the absence of any other mental faculties. We would hardly expect to find thought in the absence of memory, for instance. If free will is looked at as a psychological phenomenon, it depends on other psychological phenomena. we have specific reason to think it is dependent on other mental faculties, because, we need the ingredient of rationality to distinguish free will from "mere caprice". If we adopt the hypothesis that free will is indeed and outcome of a complex combination of determinism and indeterminism, we have further reason not to ascribe it to systems with the wrong engineering: systems like sticks and stones, or deterministic computers.

This does not add up to chauvinism, by the way. Non-humans could have the appropriate engineering, and appropriate equivalents of the accompanying mental faculties. Even a convincing artificial intelligence could have free will -- if it had genuine rationality and and genuine indeterminism. The basicness assumption seems to provide a justification for the supernatural assumption: determinism and indeterminism seem to be the only logical options for basic features of the world, so if free will is a third basic feature, it must be supernatural. But, I am contending, it is a third option which is natural but non-basic.
 
  • #26
Tournesol said:
The Basicness Assumption.

Tournesol - I have no problem with your suggestion that human decision-making may involve various mixtures of determinism and indeterminism - this may very well be true, and your Darwinian model may be a good approximation to how some parts of the brain work.

Having said this - the Darwinian model does not give rise to anything that can be distinguished from a simple mixture of determinism and random behaviour, and certainly the vast majority of libertarians (I believe) would reject the notion that free will can emerge simply by combining determinism with indeterminism - most libertarians seem to think that free will arises from some mystical realm which is beyond the simple ideas of determinism and indeterminism - it is neither, but they cannot say what it is.

Best Regards
 
  • #27
Hi MF. Thanks for going to the trouble of reading all the way through that paper. I'll try and get back to you in the next few days.
 
  • #28
MF

So my naturalistic theory is wrong becasue it is a *naturalistic* theory,
and only a supernatural theory is acceptable ? It would
have saved a lot of trouble if you had said that at the outset.
 
  • #29
After all, the point is whether FW can be defined
coherently, not whether it can be defined so as
to keep mystics happy.
 
  • #30
Tournesol said:
MF

So my naturalistic theory is wrong becasue it is a *naturalistic* theory,
and only a supernatural theory is acceptable ? It would
have saved a lot of trouble if you had said that at the outset.
Hi Tournesol

I never said your theory is wrong *because* it is naturalistic. I do not believe any theory of free will can be both coherent and complete, because I do not believe the concept of free will itself is a coherent notion.

I believe the main problem with your theory is that it is not in fact a theory of free will - it does not explain how free will differs from a simple mix of determinism and indeterminism.

Let me try to explain below :

Does Free Will entail indeterminism?

Most supporters of libertarian Free Will claim that Free Will is incompatible with determinism, on the grounds that Free Will entails “could have done otherwise”. In principle, indeterminism allows the ability to have done otherwise. The problem is that indeterminism alone simply gives rise to random (arbitrary) behaviour – and that is not a kind of Free Will that anyone really wants.

Undaunted, some supporters of libertarian Free Will claim that a suitable combination of indeterminism and determinism, linked and operating in just the right way in certain agents, can give rise to Free Will. One such example is the Darwinian model, explained in your paper.

Your model uses an indeterministic Random Idea Generator (RIG) coupled with a Sensible Idea Selector (SIS). The RIG is essentially a genuinely indeterministic random number generator, with each random output being linked to (generating) a different ”idea”, which is then forwarded to the SIS. The SIS operates either completely or mostly deterministically, evaluating the relative merits of different ideas according to some rational algorithm. In this way, random ideas can be presented for analysis by the SIS, and an unpredictable but rational or near-rational overall output from the Darwinian model is guaranteed.

Depending upon the parameters used in your model, we can generate deterministic, indeterministic, or an indeterminable combination of deterministic and indeterministic behaviour – the model can certainly appear under certain circumstances to act both rationally yet unpredictably.

But is all “rational yet unpredictable behaviour” necessarily indicative of Free Will? You clearly do not think so, but you provide no way of testing the output of your model to enable us to distinguish between genuine Free Will (whatever that might be) and any arbitrary combination of deterministic and indeterministic behaviour. Indeed, you do not even attempt to analyse what you think might be the necessary and jointly sufficient conditions for Free Will – therefore there is no way apart from “blind faith” that we can be sure your model is operating according to Free Will. The model simply seems to mimic some of the Free Will properties which libertarians presume some agents do possesses (ie humans) – viz the model can act under certain circumstances both rationally yet unpredictably. It seems that this behaviour is enough (at least for you) to safely lead to the conclusion that the model does indeed possesses Free Will.

But what happens if we replace your indeterministic RIG (iRIG) with a deterministic RIG (dRIG)?

A dRIG would be something similar to the random number generator (RNG) on your computer – for all practical purposes this RNG produces a sequence of “random” numbers which are (in principle) indistinguishable from the sequence of random numbers that would be produced by an indeterministic RNG. The one key difference is that if we reset the deterministic RNG, then it produces exactly the same series of “random” numbers all over again (this is why it is called deterministic). Apart from this, there is absolutely no in principle difference between the output of an iRIG and a dRIG.

Thus, we could remove the iRIG from your Darwinian model, and replace it with a similarly configured dRIG, and the model would operate in exactly the same way as it did before. The dRIG would produce deterministically “random” ideas which are evaluated and selected by the SIS. In operational terms, and in terms of the behaviour and output of the model, the model with the dRIG would be literally indistinguishable from the original model incorporating the iRIG.

But the new model, with the dRIG, is now a deterministic model. It acts both rationally and unpredictably, in exactly the same way as the original model, but if we reset the dRIG and start all over again, the new model will behave identically to the way that it behaved the first time. Apart from this, there is absolutely nothing anywhere in the operation, behaviour or output of the model which is in any way different to the original model with the iRIG. Aside from resetting the dRIG, neither the model itself (if it could express an opinion), nor anyone else external to the model, would be able to tell whether the model was operating deterministically or not.

If the original model (with the iRIG) possesses Free Will, as you claim, then on what rational basis can we claim that the new model (with the dRIG), which is internally and externally operationally indistinguishable from the original model, does not possesses Free Will?

The libertarian would doubtlessly claim that only the model with the iRIG has access to “alternate possibilities”, the model with the dRIG (being completely deterministic) “cannot do otherwise” than what it does, and it is this difference (according to the Libertarian) which ensures that the iRIG version is acting with Free Will, whereas the dRIG version is not. But the two models are completely indistinguishable, both to external observers and internally to the models themselves! If both models were operating side by side, and you did not know which was which, there is absolutely no test of their operation, behaviour or output that you could perform that would allow you to say “this one has Free Will, and this one does not”. Indeed, if the models were somehow able to report on their own “perceptions” of their internal decision-making processes, their reports would be indistinguishable. There would be no rational operational basis at all for discriminating between the two models. If one of the models possesses Free Will and the other does not, then it would seem that this Free Will is purely epiphenomenal, in other words it is totally ineffective in terms of its influence on the operation, behaviour or output of the model which possesses it.

If there is no rational operational basis for discriminating between the two models, then it seems clear that at least one of the following three statements must be true in the context of the kind of Free Will as defined by you :

1) Both models possesses Free Will, or
2) Neither model possesses Free Will, or
3) Free Will is purely an epiphenomenon, with absolutely no causal efficacy at all

Conclusion : Free Will (if it exists at all in the Darwinian model) is either ineffectual, or does not entail indeterminism (and is therefore compatible with determinism), or both.

Best Regards
 
Last edited:
  • #31
Hi MF. I posted a responce here, but decided I may have made some errors in interpretation of the paper. I'll have to get back to this later.
 
Last edited:
  • #32
moving finger said:
I never said your theory is wrong *because* it is naturalistic. I do not believe any theory of free will can be both coherent and complete, because I do not believe the concept of free will itself is a coherent notion.

The naturalistic concept of free will, or the supertnatural one ?

I believe the main problem with your theory is that it is not in fact a theory of free will - it does not explain how free will differs from a simple mix of determinism and indeterminism.


I don't see why I should have to. I am also not in the habit of explaining
why steel differs from a mixture of iron and carbon.

Your only reason for rejecting the "mixture" approach is the
supernaturalism -- which neither of us actually
believes in.



But is all “rational yet unpredictable behaviour” necessarily indicative of Free Will?

Yes. why not ?

You clearly do not think so, but you provide no way of testing the output of your model to enable us to distinguish between genuine Free Will (whatever that might be) and any arbitrary combination of deterministic and indeterministic behaviour.


I am using the following definition of free will

"the power or ability to rationally choose and consciously perform actions, at least some of which are not brought about necessarily and inevitably by external circumstances".

And everything in my model of FW fulfils the conditions implicit in that.

If "coherent" means anythig, it mean *internally* consistent.

Bringing in *external* defintiions of FW does *not* demosntrate
incoherence.

Indeed, you do not even attempt to analyse what you think might be the necessary and jointly sufficient conditions for Free Will

Yes I do: AP. UR, rationality.


The libertarian would doubtlessly claim that only the model with the iRIG has access to “alternate possibilities”, the model with the dRIG (being completely deterministic) “cannot do otherwise” than what it does, and it is this difference (according to the Libertarian) which ensures that the iRIG version is acting with Free Will, whereas the dRIG version is not. But the two models are completely indistinguishable, both to external observers and internally to the models themselves!

That doesn't follow


Real Randomness and Pseudo-randomness: objectivity
Some people claim it is impossible in principle to empirically detect the difference between real, intrinsic randomness and pseudo-randomness. Whilst initially plausible, this is in fact doubtful as sophisticated procedures like the Aspect experiment show. Even if it is true, the main thrust of the argument is that a free will is possible if determinism is possible, not that indeterminism-based free will is actually true. The possibillity of indeterminism-based free will is thus established even if the truth of indeterminsim based free-will is epistemically inaccessible. "it is not necessarily true" is no rebuttal to "it is possible".

Real Randomness and Pseudo-randomness: subjectivity
A variation on that argument has it that substituting pseudo-randomness for real randomness in the brain would make no subjectively detectable difference. It is difficult to see how anyone could be sure at the time of writing. There is considerable disagreement about how and to what extent subjective consciousnes relates to the physical. Whether a physical system is random or deterministic has a physical basis -- it is part of the total physical situation. Physicalism requires only that consciousness supervenes on the physical, not that it supervenes on any particular aspect of the physical, so it is physicalistically allowable for the difference between real- and pseudo-randomness to be subjectively detectable. As ever, it should be born in mind that the claim "naturalistic libertarian free will is possibly true" is not contradicted by scenarios the claim naturalistic libertarian free will is possibly false", only be the claim that it is actually false.

Real Randomness and Pseudo-randomness: necessity
Yet another variation on the same objection has it that real randomness is not actually necessary to solve the "engineering" problem -- that pseudo-randomness would have been just as good. As stated that is true, buit it is not very relevant. Nature might have evolved a pseudo-random-number generator in the brain, but that doesn't mean She did. It might have been "easier" to take afvantage of the thermal noise present in all systems. In any case, the usual response applies. The modality is wrong. To say that our thesis might not have been true does not mean it is actually false. And in any case, it is only a claim to the effect that something is possible
 
  • #33
Hi MF.

MF said: A world may be 100% deterministic (ie determinism is true in that world) without being determinable (even in principle). Limits to determinability in our own world include chaos, Heisenberg's uncertainty principle, and the in-principle indeterminability of self-referential systems - but none of these in-principle limits to determinability implies that determinism is necessarily false in our world.
Yes, I agree completely.

MF said: If the mental event is itself caused, then move the question back to the cause of the mental event.
You've said this and many things like this. I was trying to imagine some way that determinism and random causal actions might be seen as inapplicable to nature in some way. If that were the case, CHDO might take on a new meaning. It might be that CHDO is the mental state acting in some downward causal way on the physical state. Perhaps the physical state is ontically indeterminate, which I believe means it is genuinely random, not just unknowable. Feel free to correct my terminology there. So if the mental state or domain is the cause to the physical state, is there any reason to suspect these states are separate and distinct?

For some reason, I intuitively feel there must be some way around the issue of CHDO or 'free will' being based solely on deterministic and random causal actions. The paper by Campbell & Bickhard is really a critique of J. Kim's book and in so doing, it does a decent job of reviewing some of the arguments for and against downward causation and also ties in mental and physical domains. It seems as if these concepts would have something to do with rephrasing the determinism/random argument and 'free will' or CHDO but I can't yet put my finger on what it is.

One interesting issue the paper does raise, is if a mental state can equal a physical state. I don't think they're equal, though I don't think even a computationalist thinks they're equal. So this says, that any subjective experience, any thought, at any time, can not be determined by the physical state. I think this means that the mental and physical states are independent of each other to a very large degree. So if a mental state can equal an exceedingly large, potentially infinite number of physical states, and if a mental state can have some kind of causal influence over the physical state, it almost sounds as if a mental state is in some kind of domain, just like a physical state. In this case, I'm using the word "domain" to mean a set of dimensions. The physical dimensions are the three of length and one of time. If there were a true mental domain then it might be defined as existing in some other set of dimensions. These dimensions then might have some limited access to each other through strongly emergent phenomena.

The only strongly emergent phenomena that can be taken seriously in any way though must emerge at a molecular level. Being a reductionist I can't see any way complexity on a classical scale can give rise to anything that might be defined as 'strongly emergent'. The reason is not so clear unfortunately, but in short, any classical system is affected only by local causal actions. In engineering for example, I can't think of a single phenomena which can't be modeled that way, and of course they are modeled that way - as small, local chunks known as control volumes, finite elements, free body diagrams, and many other terms.

To get back to the point, if strongly emergent phenomena exist, the substrate they emerge from is a molecular level chunk of matter (ex: DNA). If a molecular chunk of matter is affected in any way by the phenomena it produces, and if we can say this affect is 'downward caused' (which seems more like 'upward caused given the strongly emergent phenomena is based at a molecular level) then it seems that if we want to maintain any kind of physicalism <not sure if that's the correct term> then we need to postulate additional dimensions that are in principal, not measurable from the physical domain and would appear random, and similarly from the mental domain, affects in that domain might appear random, but if one were to cause the other in some way, is that deterministic? I think at that point, determinism and random depend on what basis you are using. Both the physical and mental domains would have random elements from the perspective of their own domains, but when viewed from the opposite domain they may be deterministic. The problem with all this is one could still ask (as you have), is the entire system deterministic or random? Anyway, seems like a difficult question to try and think clearly about, perhaps because it is entirely counterintuitive to our own experience.
 
  • #34
An event cannot be proven random, to my knowledge, because to suggest something is random is to suggest it has no cause. If it has no cause, how to you prove it is random when it is occurring? Magic?

Premise 1: Random events occur.
Premise 2: Events are caused.

Result: If an event is occurring, it cannot be random because it must be caused.

No alternative explanation to causation has been given as I doubt there can be an alternative explanation. Perhaps one can refute causation, but I am lost when it comes to how.
 
  • #35
VI. Does Physical Indeterminism actually exist ?
The "Law of Large Numbers"
Quantum Amplification and Instrumentation
Chaos and Classicism
The Macroscopic Evidence for Indeterminism
The Myth of the "Heisenberg Cut"
The "Law of Large Numbers"
A common reaction to QM is that it doesn't matter since quantum randomness will never manifest itself at the macroscopic level -- that is, in the world of sticks and stones we can see with the naked eye. An appeal is usually made to the "law of large numbers", according to which random fluctuations at the atomic (or lower level) will cancel each other out in a macroscopic object, so that what is seen is an averaged-out behaviour that is fairly predictable.

Something like this must be happening in some cases, assuming QM is a correct description of the micro-world, or there would not even be an appearance of a deterministic macro-world. Since deterministic classical physics is partially correct, there must be a mechanism that makes the QM micro-world at least approximate to the classical description.

Quantum Amplification and Instrumentation
However, it it were the case that all macroscopic objects behaved in a 100% deterministic fashion, there would be no evidence for QM in the first place -- since all scientific apparatus is in the macro-world ! A geiger-counter is able to amplify the impact of a single particle into an audible click. Richard Feynman suggested that if that wasn't macroscopic enough, you could always amplify the signal further and use it to set off a stick of dynamite! It could be objected that these are artificial situations. This is rather desperate, however, because there is a well-known natural mechanism that could do the same job: classical chaos.

Chaos and Classicism
A classically chaotic system is by definition one that is critically sensitive to its initial conditions. "critically" sensitive means that any variation in initial conditions, no matter how slight, can bring about a change in the macroscopic behaviour of the system, no matter how large. Since there is no lower limit to critical sensitivity, it must extend all the way "down" to me microscopic world of quantum physics. Thus, hurricanes need not be started by butterfly wings, they can be started by electrons!

The term "classical" misleads some people. Chaos can be defined within the framework of classical physics, which is strictly deterministic. This is sometimes taken to mean any chaotic system encountered in nature (such as a weather system) is classical and deterministic. However, when we tall about ordinary, non-chaotic systems being classical, we mean they are *approximately* classical. Classical physics is not entirely wrong; it worked for 100's of years after all. But it is not entirely right either. "Classical" systems are quantum systems that approximate classical behaviour.

Thus any chaotic system that you can actually encounter, such as a weather system, is only approximately classical. It has no underlying determinism. At the most fundamental level it is a quantum system -- because everything is.

So we can have classical system that behave predictably (ordinary Newtonian physics), quantum systems that behave predictably on the macroscopic level (through the Law of Large Numbers), classical systems that behave unpredictably (through classical chaos) and quantum systems that unpredictably on the macroscopic as well as microscopic level (chaos and other "quantum amplifiers").

The macroscopic Evidence for Indeterminism
In fact, this is not just theoretical. Conventional big-bang theories generally require an input of quantum indeterminism to provide the large-scale structure of the universe. A singularity exploding according to classical laws would expand evenly in every direction, leading to a boring universe consisting of an evenly dispersed cloud of gas. So when you look at the night sky, you are seeing evidence for macroscopic randomness!

The Myth of the "Heisenberg Cut"
One last word: Heisenberg's uncertainty principle does include a constant "h", and it is very small. But is is not an upper limit that prevents uncertainty from leaking into the macroscopic world. In fact, the mathematical form of the Uncertainty principle:
delta_x . delta_p >= h_bar
is an inequality. It sets a lower limit on the amount of uncertainty but no upper limit.
 
Back
Top