Science and free will

[yy2bggggs]

Human behavior is clearly complex and difficult to manage. There are no certainties, which is why I find determinism untenable and I don't find randomness incompable with free will.

A logical switch, on the other hand, is one of the simplest things I can possibly think of. One particular logical switch that is interesting in this regard is a NOT gate--in the absence of signal it produces an output, in the presence of a signal it produce no output.

This is deterministic and simple to the extreme. We should be able to build a computer that predicts the output, because it's deterministic. It's very simple--all the computer has to do is say that the output is a signal if there is no input, and the output is no signal if there is.

But computers have to say this, and to say this, they have to produce an output. A computer's means of producing an output is to introduce a signal.

If that signal is connected to the input of the NOT gate, then suddenly the simple, deterministic NOT gate isn't so simple to predict.

All of this in a perfectly deterministic hypothetical universe--one of the simplest of devices is unpredictable.

All of this to say that the implication: "Not predictable implies not deterministic", is horribly wrong for interacting entities.

 

[Paul C. Anagnostopoulos]

I went back and gathered up the relevant snippets of conversation to help us both recall the original context.

You said: You don't think any aspect of my decision was determined by precursors and my current state of affairs? Then you are saying my decision was entirely random. I don't think the world would look like it does if all decisions were random.

I responded: In what sense are you using the term 'random' here? Because I think you are now meaning random in a sense other than 'not determined'. How do you think the world would be different if all decisions were random rather than some being pre-determined?

You responded: The world would be nothing but a random quantum foam, if it existed at all.
I asked: Why would having all decisions be not determined in advance result in a random quantum foam?

you responded: Wait a minute, what do you mean by "in advance"? I'm saying that if there was no determinism at all, if everything was random, then the world would just be a random quantum foam with no order whatsoever.

At which point I repeated the question, which brings us to this post. When I originally asked the question, I was thinking of human decisions.

And I was thinking of every "decision," because why would just human decisions be random? Anyway, if everything was the way it is, except that decisions made by humans were random, then the world would be indescribably bizarre. Actually, I take that back. Humans would be extinct.

Post 473 : I think it is possible to have decisions that are not deterministic but probabilistic in ways that allow the individual to control, or skew if you prefer, the probabilities. I think you are terming that as 'random', which is fine, but I also think such controlled randomness is sufficiently different from uncontrolled randomness to merit a third category.

Every single word you use to describe the third category is one related to determinism or randomness. Deterministically skewing random probabilities does not escape from determinism and randomness.

Post 435: Consider this example. I am overweight. I wish to lose weight to improve my health and looks. I try dieting and fail. I join weight watchers with my sister, who also wishes to lose weight and we both succeed. The decision to lose weight is what I think UE is referring to as 'will'. It is a choice I make. By choosing to join weight watchers, I purposefully alter the probability that I will succeed in losing weight. But it isn't a deterministic alteration. I might or might not succeed with the program nor do I know, in advance, how much of a difference it will make in the probability of my success.

The purposeful alteration of your habits is deterministic. The outcome of the weight program is deterministic with some random factors. There is nothing here that escapes determinism and nondeterminism.

Everything physical is composed of sub-atomic particles. We describe the behavior of those particles with statistical distributions rather than deterministically. Doesn't that fit the definition of random? What am I missing here? Seems to me that everything is random and yet, we do not fly apart in a shower of subatomic particles.

At the quantum level, things may or may not be random. But there are physical laws that result in a macro level that is not random. These physical laws are largely deterministic. If you want to experiment with total randomness, you must eliminate the physical laws.

We have no argument. I am unable to distinguish between compatibilist and libertarian free will.

Compatibilist free will is compatible with determinism. Libertarian free will is explicitly incompatible with determinism because it postulates a method of making decisions that is neither deterministic nor random. From Wiki:

It entails the belief that human beings possess free will, that free will is incompatible with determinism, and that determinism is false.

I use random to mean predictable with a statistical distribution - i.e. not determined.

If all physical processes were stochastic with a uniform statistical distribution, the world would be a random cloud of particles.

~~ Paul

 

[Beth]

And I was thinking of every "decision," because why would just human decisions be random? Anyway, if everything was the way it is, except that decisions made by humans were random, then the world would be indescribably bizarre. Actually, I take that back. Humans would be extinct.

After reading your post through, I think you are assuming that random implies a uniform distribution. That is not an assumption I am making. Random does not imply uniform distribution. Now, without making that assumption, how would the world be different?

Every single word you use to describe the third category is one related to determinism or randomness. Deterministically skewing random probabilities does not escape from determinism and randomness.

The purposeful alteration of your habits is deterministic. The outcome of the weight program is deterministic with some random factors. There is nothing here that escapes determinism and randomness.

First of all, it's not a deterministic skewing of random probabilities. It's a directed but still random skewing of random probabilities. Second, you asked how I set up different categories of randomness. Why would such a classification be expected to "escape" determinism and non-determinism?

At the quantum level, things may or may not be random.

At the quantum level, things may not be random? What are you basing that speculation on? I thought the hidden variables hypothesis had been thoroughly discredited. What other hypothesis exists that implies it might NOT be random?

But there are physical laws that result in a macro level that is not random. These physical laws are largely deterministic. If you want to experiment with total randomness, you must eliminate the physical laws.

How are you distinguishing between 'total randomness' and 'largely deterministic'. I consider 'largely deterministic' to be a category of random. It has a probability distribution highly skewed towards one particular outcome.

Compatibilist free will is compatible with determinism. Libertarian free will is explicitly incompatible with determinism because it postulates a method of making decisions that is neither deterministic nor random.

So I am told. But that definition doesn't help me tell them apart in practice.

If all physical processes were stochastic with a uniform statistical distribution, the world would be a random cloud of particles.

~~ Paul

Ah....your answer makes much more sense now. That's why I kept asking you what definition of random you were using. You kept saying random = not determined but answering as if random = uniform distribution.

 

[Paul C. Anagnostopoulos]

After reading your post through, I think you are assuming that random implies a uniform distribution. That is not an assumption I am making. Random does not imply uniform distribution. Now, without making that assumption, how would the world be different?

I think we would still be extinct, or never have evolved at all. Even though our decisions would follow various probability distributions, they would be unrelated to the environment. Natural selection would crush us.

First of all, it's not a deterministic skewing of random probabilities. It's a directed but still random skewing of random probabilities.

Now you must explain what it means for something to be directed but not deterministically so. You're doing an awful lot of dancing here.

Second, you asked how I set up different categories of randomness. Why would such a classification be expected to "escape" determinism and non-determinism?

I don't expect that you can escape determinism and randomness. Apparently neither do you. So I'm left not understanding what you are arguing for.

At the quantum level, things may not be random? What are you basing that speculation on? I thought the hidden variables hypothesis had been thoroughly discredited. What other hypothesis exists that implies it might NOT be random?

There may be causes we do not understand. But it doesn't matter, we can assume that things are random at the quantum level.

How are you distinguishing between 'total randomness' and 'largely deterministic'. I consider 'largely deterministic' to be a category of random. It has a probability distribution highly skewed towards one particular outcome.

Yes, that's fine. I still have no idea what you're arguing for.

So I am told. But that definition doesn't help me tell them apart in practice.

That is because libertarian free will is nonsensical, and so there is nothing to tell apart.

Ah....your answer makes much more sense now. That's why I kept asking you what definition of random you were using. You kept saying random = not determined but answering as if random = uniform distribution.

When you asked for a definition, it did not occur to me that you were asking whether the random processes were uniform or nonuniform. They are both still random.

I think the confusion may lie in the question of where the nonuniformity in nonuniform probability distributions comes from. If it just is, then the random process has no deterministic element at all. But if the nonuniformity was imposed on the process, then it is still random but with a deterministic probability skewing. Either way, you've got nothing but determinism and randomness. There does not appear to be any room for a libertarian sort of will here.

~~ Paul

 

[Beth]

After reading your post through, I think you are assuming that random implies a uniform distribution. That is not an assumption I am making. Random does not imply uniform distribution. Now, without making that assumption, how would the world be different?

I think we would still be extinct, or never have evolved at all. Even though our decisions would follow various probability distributions, they would be unrelated to the environment. Natural selection would crush us.

Why are you assuming that the various probability distributions would be unrelated to the environment? Just because the environment can affect the probability distribution of outcomes does not make the resulting decision determined rather than random.

First of all, it's not a deterministic skewing of random probabilities. It's a directed but still random skewing of random probabilities.

Now you must explain what it means for something to be directed but not deterministically so. You're doing an awful lot of dancing here.

In circles in would seem. . I'll refer you again to a previous post:
Post 435: Consider this example. I am overweight. I wish to lose weight to improve my health and looks. I try dieting and fail. I join weight watchers with my sister, who also wishes to lose weight and we both succeed. The decision to lose weight is what I think UE is referring to as 'will'. It is a choice I make. By choosing to join weight watchers, I purposefully alter the probability that I will succeed in losing weight. But it isn't a deterministic alteration. I might or might not succeed with the program nor do I know, in advance, how much of a difference it will make in the probability of my success. That is an example of a directed but still random change in the probability distribution of possible outcomes.

Second, you asked how I set up different categories of randomness. Why would such a classification be expected to "escape" determinism and non-determinism?

I don't expect that you can escape determinism and randomness. Apparently neither do you. So I'm left not understanding what you are arguing for.

I find it helpful in furthering my own understanding. What are you arguing for?

How are you distinguishing between 'total randomness' and 'largely deterministic'. I consider 'largely deterministic' to be a category of random. It has a probability distribution highly skewed towards one particular outcome.

Yes, that's fine. I still have no idea what you're arguing for.

I'm not arguing for anything here. I'm just trying to understand your points. Could you answer the question and explain how you are distinguishing between 'total randomness' and 'largely deterministic'? Does total randomness imply a uniform distribution?

Ah....your answer makes much more sense now. That's why I kept asking you what definition of random you were using. You kept saying random = not determined but answering as if random = uniform distribution.

When you asked for a definition, it did not occur to me that you were asking whether the random processes were uniform or nonuniform. They are both still random.

Correct. I've no disagreement that they are both still random. I've been confused because many of your statements only make sense to me if random is assumed to imply the uniform distribution.

I think the confusion may lie in the question of where the nonuniformity in nonuniform probability distributions comes from. If it just is, then the random process has no deterministic element at all. But if the nonuniformity was imposed on the process, then it is still random but with a deterministic probability skewing. Either way, you've got nothing but determinism and randomness. There does not appear to be any room for a libertarian sort of will here.

~~ Paul

When we consciously impose a different distribution on a set of potential outcomes, that is what I consider intent being a causal factor, otherwise known as exerting free will. Whether it fits with LFW, I cannot say nor do I particularly care one way or the other.

eta: Why assume that anything other than a uniform probability distribution was imposed. That assumes that the unform distribution is the default one. There are some good reasons for making this assumption from a computational point of view, but I don't know that there is any philosophical justification for it.

 

[Paul C. Anagnostopoulos]

Why are you assuming that the various probability distributions would be unrelated to the environment? Just because the environment can affect the probability distribution of outcomes does not make the resulting decision determined rather than random.

We have to stop meeting like this. I simply do not understand what you are asking. You said "Let me repeat the question: How do you think the world would be different if all decisions were random rather than some being pre-determined?" Now you have to explain to me what you mean by random, because as far as I can tell you actually mean determined.

You appear to be playing games with the word random by allowing a process to be arbitrarily determined yet still calling it random. Technically, yes, any amount of randomness in a process allows us to call it stochastic, but there is clearly a difference between a largely random process and a largely deterministic one. So please rephrase your question and make it clear exactly how much randomness you're talking about.

n circles in would seem. . I'll refer you again to a previous post:
Post 435: Consider this example. I am overweight. I wish to lose weight to improve my health and looks. I try dieting and fail. I join weight watchers with my sister, who also wishes to lose weight and we both succeed. The decision to lose weight is what I think UE is referring to as 'will'. It is a choice I make. By choosing to join weight watchers, I purposefully alter the probability that I will succeed in losing weight. But it isn't a deterministic alteration. I might or might not succeed with the program nor do I know, in advance, how much of a difference it will make in the probability of my success. That is an example of a directed but still random change in the probability distribution of possible outcomes.

Your error is in "But it isn't a deterministic alteration." Yes, it is. The outcome isn't necessarily determined, but the alterations were.

If you rig a die so it's most likely to come up a 4, then the rigging was deterministic even though the results of a throw are still stochastic.

I find it helpful in furthering my own understanding. What are you arguing for?

That libertarian free will is incoherent.

I'm not arguing for anything here. I'm just trying to understand your points. Could you answer the question and explain how you are distinguishing between 'total randomness' and 'largely deterministic'? Does total randomness imply a uniform distribution?

I don't think so. I think you have to say "uniformly random" if that's what you mean. A random process can have various probability distributions. It is still random as long as it lacks correlation with other events.

When we consciously impose a different distribution on a set of potential outcomes, that is what I consider intent being a causal factor, otherwise known as exerting free will. Whether it fits with LFW, I cannot say nor do I particularly care one way or the other.

That isn't libertarian free will, but it is a form of compatibilist free will.

eta: Why assume that anything other than a uniform probability distribution was imposed. That assumes that the unform distribution is the default one. There are some good reasons for making this assumption from a computational point of view, but I don't know that there is any philosophical justification for it.

Was imposed by what?

~~ Paul

 

[Beth]

Why are you assuming that the various probability distributions would be unrelated to the environment? Just because the environment can affect the probability distribution of outcomes does not make the resulting decision determined rather than random.

We have to stop meeting like this. I simply do not understand what you are asking. You said "Let me repeat the question: How do you think the world would be different if all decisions were random rather than some being pre-determined?" Now you have to explain to me what you mean by random, because as far as I can tell you actually mean determined.

You appear to be playing games with the word random by allowing a process to be arbitrarily determined yet still calling it random. Technically, yes, any amount of randomness in a process allows us to call it stochastic, but there is clearly a difference between a largely random process and a largely deterministic one. So please rephrase your question and make it clear exactly how much randomness you're talking about.

I'm talking about ANY amount of randomness. Forgive me for sticking to the technical definition, but I am a professional statistician. Can you tell me how much randomness you would allow and still consider the process to be a deterministic one?

n circles in would seem. . I'll refer you again to a previous post:
Post 435: Consider this example. I am overweight. I wish to lose weight to improve my health and looks. I try dieting and fail. I join weight watchers with my sister, who also wishes to lose weight and we both succeed. The decision to lose weight is what I think UE is referring to as 'will'. It is a choice I make. By choosing to join weight watchers, I purposefully alter the probability that I will succeed in losing weight. But it isn't a deterministic alteration. I might or might not succeed with the program nor do I know, in advance, how much of a difference it will make in the probability of my success. That is an example of a directed but still random change in the probability distribution of possible outcomes.

Your error is in "But it isn't a deterministic alteration." Yes, it is. The outcome isn't necessarily determined, but the alterations were.

er, no they weren't. That was my point. Only direction of the alteration was determined, not the actual amount of change. Since there is randomness in the alternation, that makes it random not determined.

If you rig a die so it's most likely to come up a 4, then the rigging was deterministic even though the results of a throw are still stochastic.

I think this is where we are running into confusion. Why do you consider the rigging to be deterministic? What makes it deterministic as opposed to random? That fact that the probability was skewed towards a specific number?

I'm not arguing for anything here. I'm just trying to understand your points. Could you answer the question and explain how you are distinguishing between 'total randomness' and 'largely deterministic'? Does total randomness imply a uniform distribution?

I don't think so. I think you have to say "uniformly random" if that's what you mean. A random process can have various probability distributions. It is still random as long as it lacks correlation with other events.

So how are you are distinguishing between 'total randomness' and 'largely deterministic'? BTW, even it is correlated with other events, that doesn't mean it's not random.

eta: Why assume that anything other than a uniform probability distribution was imposed. That assumes that the unform distribution is the default one. There are some good reasons for making this assumption from a computational point of view, but I don't know that there is any philosophical justification for it.

Was imposed by what?

~~ Paul

You misunderstand my question. Why assume that if no probability distribution was imposed, that a uniform distribution is what will occur?

 

[cyborg]

Can you tell me how much randomness you would allow and still consider the process to be a deterministic one?

"Quantities" of randomness are irrelevant.

It's not the way that you do it: it's what you do. That's what gets results.

 

[Paul C. Anagnostopoulos]

I'm talking about ANY amount of randomness. Forgive me for sticking to the technical definition, but I am a professional statistician. Can you tell me how much randomness you would allow and still consider the process to be a deterministic one?

To be strictly deterministic, a process can have no randomness at all, except possibly the occasional random error. For example, a computer is considered deterministic even though memory errors occur occasionally.

er, no they weren't. That was my point. Only direction of the alteration was determined, not the actual amount of change. Since there is randomness in the alternation, that makes it random not determined.

Okay, so you deterministally decided to join weight watchers, but the level of your determination to lose weight is random. That's fine. All of these combinations are just fine. I have no argument with compatibilist free will.

I think this is where we are running into confusion. Why do you consider the rigging to be deterministic? What makes it deterministic as opposed to random? That fact that the probability was skewed towards a specific number?

There is really no point in continuing this, since we both agree that we're talking about compatibilist free will. And I have no problem with calling a process stochastic unless it is completely deterministic. And I agree that it is difficult to know whether a process is completely deterministic. However, if I program a computer to skew the die toward 4 with a specific probability, then I think it is fair to say that the skewing is deterministic. The roll is not, of course.

So how are you are distinguishing between 'total randomness' and 'largely deterministic'? BTW, even it is correlated with other events, that doesn't mean it's not random.

It's not a dichotomy, it's a spectrum. A conventional computer is completely deterministic. Radioactive decay appears to be entirely random. Other processes fall on the spectrum in various places. I would say a process is entirely random if it does, indeed, lack any correlation with other events. In that case, it may or may not be uniformly random.

You misunderstand my question. Why assume that if no probability distribution was imposed, that a uniform distribution is what will occur?

I don't assume that, as I just said above.

I think we are in massive agreement here. I'd love to here your opinion about libertarian free will.

~~ Paul

 

[Beth]

To be strictly deterministic, a process can have no randomness at all, except possibly the occasional random error. For example, a computer is considered deterministic even though memory errors occur occasionally.

Right. So why do you think that if all decisions were random, humans would be extinct? What definition of random were you using?

I think we are in massive agreement here. I'd love to here your opinion about libertarian free will.

I think I don't understand it and have pretty much given up trying to. The last few posts UE made it about only confused me more.

Beth

 

[Belz...]

Right. So why do you think that if all decisions were random, humans would be extinct?

If everything were random, there would be no structure and no physical laws, so I think it's safe to say that nothing we know would exist. Of course, if you consider the the laws today are the result of local, temporary order in the underlying chaos due to some symmetry break, sure.

 

[Paul C. Anagnostopoulos]

Right. So why do you think that if all decisions were random, humans would be extinct? What definition of random were you using?

A definition that places it opposite "deterministic," which means far over on the random side of the spectrum. The sort of random that is largely uncorrelated with anything. I apologize for not making this clear all along.

I think I don't understand it and have pretty much given up trying to. The last few posts UE made it about only confused me more.

Join the crowd.

~~ Paul