• Welcome to the Internet Infidels Discussion Board.

According to Robert Sapolsky, human free will does not exist

I submit that the "mind" and the "will" are philosophical and theological concepts that have no place in any discussion of science, physics, math, biology, etc.
Unless, of course, we understand chaos theory, which is empirical, well established, and effectively eliminates your submission.
Some folks loosely / carelessly refer to the brain as the mind, but they are different things.
Indeed. Mind is what the brain does. The idea that the brain and the mind are the same is analogous to the idea that the car and the journey are the same thing.

It's a gross error.

As is "I don't understand this, therefkre nobody understand this"; Or "I don't understand this, therefore it is an eternally ineffable mystery that man should cease to attempt to resolve".

The same is true of the "will." Outside of the realm of philosophy or theology, it is better to speak about a "decision" or, better yet, "brain function."

Indeed. Lots of the words we need for these discussions have been (perhaps irreperably) tainted by theology and/or by wildly other erroneous philosophies. Which makes understanding hard, but not impossible.

How does chaos theory
Here is a nice bible story.

I didn’t really mean to sleep with your wife, but the Big Bang (AKA God) made me do it.

:rolleyes:

The personal slur aside... speaking in general, the events of the world, including your own nature and circumstances made you want to do it. Your will, your wants and needs made you want to do it, you felt a strong urge, so you act accordingly.

That is how compatibilists define free will, and it is wrong for the given reasons.

What personal slur? :unsure:

The 'your wife' bit. Referring to someone's wife in that way is personal and has implications. It's meant to be insulting.
 
Once the linguistic baggage is removed from the conversation, the question boils down to whether the brain and/or brain function operates in a manner that is not truly, entirely, and perfectly determined in advance by antecedent activity of the universe or is more akin to quantum activity that is, at best, probabilistic (and not because of any hidden variable) and capable of activity that is, at least in some small way, independent of antecedent activity.
Well, it does if we don't know about chaos theory, yes.

But we probably should know about the central theories that apply in a field, before making bold dichotomous claims about that field.

"Deterministic" is widely assumed to imply "predictable". But any meteorologist can tell you that ain't so - and a weather system is significantly less complex than a living mammalian brain.

How would chaos theory or indeterminism help with the brain/mind process of decision-making or the idea of free will?
 

The 'your wife' bit. Referring to someone's wife in that way is personal and has implications. It's meant to be insulting.

Um … no. It was an illustrative example. But I suppose the Big Bang coerced you to say what you did. :rolleyes:
 
Indeed. Mind is what the brain does. The idea that the brain and the mind are the same is analogous to the idea that the car and the journey are the same thing.

To say that "Mind is what the brain does" sounds like you are using the term as a verb, when it is, in fact, a noun (outside, of course, from secondary meanings that I don't understand to be in play here -- such as when the term is used in the following sentences: "Please mind the store until I get back" or "I don't mind if you come with me").

Whether I am right or wrong about this observation, however, I still believe it is best to avoid the term altogether and use the term "neurological activity." I suppose that "cognitive function" also could work, but that arguably would exclude autonomic function.

The problem with using the term "Mind" in this discourse is that it invites the Cartesian notion of a ghost in the machine that exists separate and apart from the body. I could be mistaken, but I don't believe that is intended in the discussion on this board.
 
you would simply acknowledge that the two of you are using these words in different senses
How have I not? I have literally told YOU, right I front of him, that when he says "determinism" he is NOT using the "compatibilist" definition, no matter how he says he is.

I am NOT the one who has a difficulty telling the difference, even insofar as I suggest a whole second set of words to attach to this distinct concept of "pre-determinism", and described the specific difference.

The problem is that he refuses. He absolutely refuses to acknowledge that the sense which I have described over a hundred times, and over a hundred times differentiated does not itself imply pre-determinism/Radical Fatalism.

I feel little obligation to tell him that the definition that compatibilists use for determinism does not comport to pre-determinism, when I have told him this so many times.

He needs to hear it from someone who isn't me.

Perfectly reliable causation does not imply radical fatalism.
 
Once the linguistic baggage is removed from the conversation, the question boils down to whether the brain and/or brain function operates in a manner that is not truly, entirely, and perfectly determined in advance by antecedent activity of the universe or is more akin to quantum activity that is, at best, probabilistic (and not because of any hidden variable) and capable of activity that is, at least in some small way, independent of antecedent activity.
Well, it does if we don't know about chaos theory, yes.

But we probably should know about the central theories that apply in a field, before making bold dichotomous claims about that field.

"Deterministic" is widely assumed to imply "predictable". But any meteorologist can tell you that ain't so - and a weather system is significantly less complex than a living mammalian brain.

How would chaos theory or indeterminism help with the brain/mind process of decision-making or the idea of free will?
I don't think indeterminism is even under discussion here, so lets leave that out altogether.

How would chaos theory help with the brain/mind process of decision-making

It wouldn't. It is just an observed attribute of the brain/mind process that it is chaotic, so we can conclude that the decisions made by the brain/mind process cannot be pre-determined in any way.

or the idea of free will?

Well, my idea of "free will" is that it entails the confering of basic desert responsibility. That is, to determine whether a decision is made "of my own free will", the criterion we look at is "am I personally responsible for that decision?".

If the only way to predict a future decision by me is to wait and see what I do, then the answer is "yes"; The decision was my responsibility.

If you push me off a cliff, it is easy to predict that I will hit the ground. You can even calculate what my speed will be when I do.

If you put me in a restaurant, it is impossible to predict what I will order. The choice is mine, and I am responsible for it. It is a freely willed choice.

You, I understand, have a different definition of "free will", carefully crafted to be impossible. Nobody should be shocked or surprised that free will as you define it is impossible, because that's baked in to the definition. However, your definition is tautological, and as such, uninteresting, despite being true. So I am talking about something more interesting - the question of whether I am responsible for my actions, or whether I can blame impersonal external forces going back to the beginning of time.
 
What is this "mind" you are referring to, that is not subject to exactly the same sort of descriptive predictability of a falling rock?
A mind is a set of complex and chaotic processes in a brain.

Even mindless processes lack the descriptive predictability of a falling rock, once they are complex enough to exhibit chaotic behaviour*.

There is a huge amount of evidence that a human brain (and hence its activity, part of which we call "mind") is chaotic, and therefore fundamentally unpredictable, not because it contains any non-deterministic parts, but because it is impossible in principle to know all of the inputs it uses (including its own 'initial state') to a sufficient degree of accuracy as to make more than a very vague forecast of its most plausible outputs.







* Chaotic behaviour is exhibited by a system when changes to the inputs too small to measure can lead to massive and obvious changes in outcome - the classic thought experiment here is the Butterfly Effect, which illustrates the chaotic nature of weather systems in the medium term.
This is telling part of the story, but not all of the story. Yes, Ian Malcolm is right about complex systems greatly compounding the problems of prediction. However, this is a question of scale. Can I predict with great accuracy exactly how much rain will fall into my yard gauge tomorrow? No. But I can model with considerable accuracy the chance of rain in a given hour within a larger test area. At very large scales, our predictions (it will rain somewhere on this continent tomorrow) will have an accuracy nearing (though never quite reaching) 100%. People often misinterpret chaos to mean that no prediction is possible, that if we do not have absolute certainty then random guesses, intuition, divine revelation, and predictions based on solid, scientifically derived models are all equally likely to be correct. This is not the case. We often have a pretty good idea of what percentage of doubt or range affects our predictions in a particular case, and in the situation of a well-developed, field tested theory set, that margin is apt to be small. This would not be the case if those complex phenomena were truly random. We have every reason to assume that uniformitarianism holds for weather, language, genetics, psychology, and so forth.
 
Once the linguistic baggage is removed from the conversation, the question boils down to whether the brain and/or brain function operates in a manner that is not truly, entirely, and perfectly determined in advance by antecedent activity of the universe or is more akin to quantum activity that is, at best, probabilistic (and not because of any hidden variable) and capable of activity that is, at least in some small way, independent of antecedent activity.
Well, it does if we don't know about chaos theory, yes.

But we probably should know about the central theories that apply in a field, before making bold dichotomous claims about that field.

"Deterministic" is widely assumed to imply "predictable". But any meteorologist can tell you that ain't so - and a weather system is significantly less complex than a living mammalian brain.

How would chaos theory or indeterminism help with the brain/mind process of decision-making or the idea of free will?

Exactly. Chaos Theory, Indeterminism, Quantum Theory and all other theories that are based on the perceived observation of a lack of perfect predictability miss the point that the historical usage of the term Determinism in philosophy describes an idealized paradigm that exists without respect to human ability to (i) understand the entire confluence of antecedent activity that produces current or future activity, or (ii) predict with certainty the activity that will be produced by the confluence of antecedent activity. It is a fallacy to say that Philosophical Determinism is falsified by the fact (if it were a fact) that the future cannot be predicted with any certainty and is therefore probabilistic.

The weather example is a good one. The fact that we are not able to predict the weather with certainty does not prove, ipso facto, that the weather is not perfectly deterministic. While it is true that weather not being perfectly deterministic would preclude our ability to perfectly predict the weather, another reason for our inability could simply be human limitations to detect and understand all of the existing variables that determine the weather. It is a fallacy to glom onto the former reason as a logical necessity when the latter reason is equally possible (from a logical perspective).
 
Indeed. Mind is what the brain does. The idea that the brain and the mind are the same is analogous to the idea that the car and the journey are the same thing.

To say that "Mind is what the brain does" sounds like you are using the term as a verb, when it is, in fact, a noun (outside, of course, from secondary meanings that I don't understand to be in play here -- such as when the term is used in the following sentences: "Please mind the store until I get back" or "I don't mind if you come with me").
Is "Journey" a verb? I think it's a noun.

We can say "<noun> is what a <noun> does" in English with perfect sense - the verb here is "does".
Whether I am right or wrong about this observation, however, I still believe it is best to avoid the term altogether and use the term "neurological activity." I suppose that "cognitive function" also could work, but that arguably would exclude autonomic function.
Fair enough.
The problem with using the term "Mind" in this discourse is that it invites the Cartesian notion of a ghost in the machine that exists separate and apart from the body. I could be mistaken, but I don't believe that is intended in the discussion on this board.
Yeah, I agree. "Mind" has overtones of the very dumb idea of substance dualism. I certainly don't want to imply that I subscribe to any such nonsense.

As I mentioned above, the theologists have tainted much of the language needed in these kinds of discussions; I don't see an easy way out, other than the verbosity of constantly clarifying that dualism is not being implied here.

Maybe "thought" is better than "mind", but that carries implications of conscious thought, while "mind" is the combination of both conscious and sub-conscious brain activity.
 
Someone who can’t tell the difference between humans and camels on one side and rocks on the other has seriously fucked up metaphysics. And yes, it’s all metaphysics, of which science is a subset.
I'm not the one espousing confusion as to whether rocks or camels or humans have minds. "Mind" is your concept here, defend it if you can. And if all you can come up with is "well isn't it obvious?", than what you have isn't even technically speaking a philosophy, as you've never thought it through systematically.
 
What is this "mind" you are referring to, that is not subject to exactly the same sort of descriptive predictability of a falling rock?
A mind is a set of complex and chaotic processes in a brain.

Even mindless processes lack the descriptive predictability of a falling rock, once they are complex enough to exhibit chaotic behaviour*.

There is a huge amount of evidence that a human brain (and hence its activity, part of which we call "mind") is chaotic, and therefore fundamentally unpredictable, not because it contains any non-deterministic parts, but because it is impossible in principle to know all of the inputs it uses (including its own 'initial state') to a sufficient degree of accuracy as to make more than a very vague forecast of its most plausible outputs.







* Chaotic behaviour is exhibited by a system when changes to the inputs too small to measure can lead to massive and obvious changes in outcome - the classic thought experiment here is the Butterfly Effect, which illustrates the chaotic nature of weather systems in the medium term.
This is telling part of the story, but not all of the story. Yes, Ian Malcolm is right about complex systems greatly compounding the problems of prediction. However, this is a question of scale. Can I predict with great accuracy exactly how much rain will fall into my yard gauge tomorrow? No. But I can model with considerable accuracy the chance of rain in a given hour within a larger test area. At very large scales, our predictions (it will rain somewhere on this continent tomorrow) will have an accuracy nearing (though never quite reaching) 100%. People often misinterpret chaos to mean that no prediction is possible, that if we do not have absolute certainty then random guesses, intuition, divine revelation, and predictions based on solid, scientifically derived models are all equally likely to be correct. This is not the case. We often have a pretty good idea of what percentage of doubt or range affects our predictions in a particular case, and in the situation of a well-developed, field tested theory set, that margin is apt to be small. This would not be the case if those complex phenomena were truly random. We have every reason to assume that uniformitarianism holds for weather, language, genetics, psychology, and so forth.
Sure. But that's also true in the context of human decision making. We can make good statistical predictions about the decisions of large numbers of people - for example in election forecasting, while predicting a single person's vote can be far more difficult - or not. Some individual votes (or other decisions) are very predictable indeed, just as some weather forecasts are very confident indeed.

Human behaviour isn't random; Far from it. But it can be unpredictable. Chaos isn't randomness. The Mandelbrot Set comes out exactly the same every single time you do the maths. But predicting the colour of a given infinitessimal pixel in that set may be impossible, because as your precision varies, so can the output.

There exist decisions for which no predictor other than the decision making system itself can have sufficient information about the starting conditions to determine the output.

At that point, it is unreasonable to assign responsibility for that decision to anything other than the system.
 
Last edited:
I will note that several of the people here have been discussing this topic for a long time.

The earlier threads about this topic go back for years.

Most of the compatibilists here have limited patience at this point, perhaps me moreso than others because I also see the people who come through Reddit about this topic making the same mistakes in waves.

As I have pointed out, the ability to do otherwise is present and real and observed in the left doing otherwise than the right; some segment of the universe did otherwise.

This is again because any time any statement references ability, it no longer references a single instance but a "type". It references some metaphysical concept divorced from any one place or time but true about 'physics'; and that any time someone figures it to apply to an individual specific thing at a unique time and place, they are wrong to do so, spouting nonsense. There is of physics the ability to make something otherwise than as is made here, and that IS by tautology 'the ability to do otherwise'. It's demonstrated in a simple act of looking around you and there being variance.
 
Maybe "thought" is better than "mind", but that carries implications of conscious thought, while "mind" is the combination of both conscious and sub-conscious brain activity.

Why not simply go with "brain function" "brain activity" "neurological function" or "neurological activity" -- all of which are neutral to conscious versus unconscious and avoid Cartesian dualism.

Otherwise, thanks for your considered response to my prior post.
 
What is this "mind" you are referring to, that is not subject to exactly the same sort of descriptive predictability of a falling rock?
A mind is a set of complex and chaotic processes in a brain.

Even mindless processes lack the descriptive predictability of a falling rock, once they are complex enough to exhibit chaotic behaviour*.

There is a huge amount of evidence that a human brain (and hence its activity, part of which we call "mind") is chaotic, and therefore fundamentally unpredictable, not because it contains any non-deterministic parts, but because it is impossible in principle to know all of the inputs it uses (including its own 'initial state') to a sufficient degree of accuracy as to make more than a very vague forecast of its most plausible outputs.







* Chaotic behaviour is exhibited by a system when changes to the inputs too small to measure can lead to massive and obvious changes in outcome - the classic thought experiment here is the Butterfly Effect, which illustrates the chaotic nature of weather systems in the medium term.
This is telling part of the story, but not all of the story. Yes, Ian Malcolm is right about complex systems greatly compounding the problems of prediction. However, this is a question of scale. Can I predict with great accuracy exactly how much rain will fall into my yard gauge tomorrow? No. But I can model with considerable accuracy the chance of rain in a given hour within a larger test area. At very large scales, our predictions (it will rain somewhere on this continent tomorrow) will have an accuracy nearing (though never quite reaching) 100%. People often misinterpret chaos to mean that no prediction is possible, that if we do not have absolute certainty then random guesses, intuition, divine revelation, and predictions based on solid, scientifically derived models are all equally likely to be correct. This is not the case. We often have a pretty good idea of what percentage of doubt or range affects our predictions in a particular case, and in the situation of a well-developed, field tested theory set, that margin is apt to be small. This would not be the case if those complex phenomena were truly random. We have every reason to assume that uniformitarianism holds for weather, language, genetics, psychology, and so forth.
Sure. But that's also true in the context of human decision making. We can make good statistical predictions about the decisions of large numbers of people - for example in election forecasting, while predicting a single person's vote can be far more difficult - or not. Some individual votes (or other decisions) are very predictable indeed, just as some weather forecasts are very confident indeed.

Human behaviour isn't random; Far from it. But it can be unpredictable.
Correct.

And it should not be as predictable as it is, if our choices are in fact unbounded by cause and effect, and the physical properties of the nervous system, nutrition, the systems that affect our environment, and so forth. To have some room for error is only prudent when trying to analyze human behaviors, which in any specific case are often barely studied, but it is not reasonable to conclude that scientifically derived models are useless, or that they would not continue to become more useful as they are informed by a greater volume of data and observations.
 
The weather example is a good one. The fact that we are not able to predict the weather with certainty does not prove, ipso facto, that the weather is not perfectly deterministic.
Which is rather irrelevant to the discussion of compatibilism, as compatibilists do not claim that anything is not perfectly deterministic - only that the fact that a system is deterministic does not preclude that system from basic desert responsibility for its decisions - ie from the freedom to make choices.
 
Maybe "thought" is better than "mind", but that carries implications of conscious thought, while "mind" is the combination of both conscious and sub-conscious brain activity.

Why not simply go with "brain function" "brain activity" "neurological function" or "neurological activity" -- all of which are neutral to conscious versus unconscious and avoid Cartesian dualism.

Otherwise, thanks for your considered response to my prior post.
I would caution against solely looking at neuroscience as a predictor of human behavior. It is an incredibly important factor in the consideration of any human behavior, but the nervous system is ultimately a screen and processing matrix for the body's response to inputs, it does nothing "on its own" as it were, so on some level making any accurate guess about what an individual will do is dependent on having a perfect knowledge of their conditions, which is one of those chaotic factors that bilby was referencing above. We know it matters whether the examiner looks the subject in the eye when they ask the question, but not whether they will blink at the wrong moment. We know the result will be different if the test subject knows the examiner, but it can be difficult to assess whether a particular relationship is close enough to bias the results. Etc.
 
Maybe "thought" is better than "mind", but that carries implications of conscious thought, while "mind" is the combination of both conscious and sub-conscious brain activity.

Why not simply go with "brain function" "brain activity" "neurological function" or "neurological activity" -- all of which are neutral to conscious versus unconscious and avoid Cartesian dualism.

Otherwise, thanks for your considered response to my prior post.
So the duality of mind is an interesting subject.

Inevitably, all the systems we observe that report some internal state report this internal state in an indirect way.

We can see some things happen, but it is fundamentally going to look different to our senses as this report is rendered than it appears "from the perspective of the interaction".

Indeed, phenomena simply occur where they are according to physics, while we observe these phenomena in a largely disconnected way.

To that end, I have here for years widely questioned the very assertion that there is such a thing as "non-consciousness"; that all phenomena are in some manner experiential, but that most of these experiences integrate uselessly or fail to integrate at all.

If this is the case, asking "where does consciousness come from" would be like asking "where does physics come from".

If we were to then instead ask "why are we conscious of some things and not all things if all phenomena implies consciousness", we might answer something like "because THESE phenomena feed into one another to do meaningful work and expose the presence of surrounding statistically useful information, and THOSE just feed into one another as a solid, homogenous mass of silica."

Of course one of the other things this does is reveal humans to be rather "not-special", and implies some very uncomfortable things about the technologies we are developing today which are capable of genuine ongoing autonomous self-modification and maintenance.

"Unconscious" here just implies that it's not reporting to you, or perhaps even that you aren't generating a report to yourself or at least you aren't recording it.

Lots of things disrupt the artifact of our awareness of some manner of activity in our brains, or outright place those interactions over horizons. But you being deaf does not mean someone else is not speaking and making sound. It just means that YOU aren't able to hear it. Maybe they're deaf too... But the air is still wiggling even if nobody there can feel it.
 
Last edited:
Back
Top Bottom