Free Will Compatibilism and Incompatibilism

I think we should just show that the idea of libertarian free will is empty, and that there is no empirical basis on it.
Yes. But the moral responsibility which is supposed to follow from it is the idea we can deserve what happens to us without any link to consequences. In an ideal world all sentient beings would have lives worth living. Of course we are not in an ideal world but the point is so important. We just get the experiences we are given. Nobody can deserve bad ones not as is supposed and it is shockingly awful to think people or any animal can. If we get real and say we need some people to suffer for consequential reasons it changes how we feel. I don't like it one bit. But yes I'm resigned to it. The thing is it makes one want to reduce the harm to as little as possible and yes ideally irradicate it altogether. We can get a tremendous positive change from getting this right.
Quoted from here].
I think if you read here] and here], you have a pretty good summary of my viewpoints. I let it to Stephen to explain why he does not agree with me.
Consider: 1. Antecedent conditions, coupled with the laws of nature, entail what I do. 2. No one can change the facts of the past, or the laws of nature. 3. Therefore, no one can do, other than what she/he does, because no one can change the facts of the past, present or future or the laws of nature.
I suppose this is what you see as the core of a determinism that excludes the possibility of free will. Still I think it does not suffice: in a compatibilist view there is still no problem. What people do is determined, but people also determine what happens: this 'human determining' runs via reflection, anticipating possible future paths, knowledge, motives, etc etc. It is only when you define free will as the possibility that with your brain given in a certain state, different paths can be taken and that these somehow correspond to your will, that above point rule out free will. I must note I have sever problems with phrases like 'changing the present/past/future': change from what to what. I can change the colour of a wall by painting it, but I have no idea what changing the present/past/future could even mean. Yes, exactly. I'm sure you've noted I've set the argument up, only to knock it down.
Response: 1. Laws of nature are descriptive, not prescriptive. They are a subclass of all the true propositions that describe the world. 2. Therefore, if at some time I did other, than what I in fact did, then the laws of nature (descriptive, not prescriptive) would have been different. In doing x or y, I generate "laws": Law x, or Law y -- bearing in mind that, like Gresham's Law, laws of nature are descriptive, not prescriptive, and are a subclass of all true propositions. 3. Given that the laws of nature are descriptive and not prescriptive, no natural "law" can compel me to do anything, because natural laws are not actually laws: descriptions are not prescriptions. 4. It is true that I cannot change any (descriptive) "laws" of nature, such as E=MC2. Nor can I change the past, future, or for that matter, the present. However, no account of free will requires that I be able to change the past, present or future. Nor does any such account require that I be able to change true propositions that describe the world.
Right. The idea that we are 'compelled' by the laws of nature to do something looses all its power if you see what the laws of nature really are: descriptions of how things happen. Laws of nature do not make facts true, facts make laws of nature true. Laws of nature force nothing, so they also do not force me to do anything. Facts about me make laws of nature true. However, you have not broken the complete spell of determinism with this idea. Determinism says that given the state of the universe at a certain time, that for every time afterwards there is only one possible state. This definition is not affected by changing the view of laws of nature from prescriptions to descriptions. Quantum physics shows that this is not correct. If we rewound the tape of history and everything up until the present moment unfolded exactly the same, the next step is probabilistic, not deterministic. So Laplacean or hard determinism is known to be false. On the other hand, there is the Many Worlds deterministic version of QM, but this opens a new can of worms.
And except your further use of 'change the present/past/future' I agree with your train of thought.
Well, of course, I'm pointing out that this notion of "change" is fallacious.
But you said you were a libertarian, so I am awaiting your followup...
Yes, more about that later. In the above, I wanted to defend a compatibilist account.
I think we should just show that the idea of libertarian free will is empty, and that there is no empirical basis on it.
Yes. But the moral responsibility which is supposed to follow from it is the idea we can deserve what happens to us without any link to consequences. Then we agree, don't we? LFW is nonsense, and so is the idea that we really deserve praise, blame and punishment. Praise, blame and punishment are reactions from the community to actions of responsible persons. And responsible persons are those who are able to evaluate the consequences of their actions, including how the community sees their actions and how it possibly will react.
Quantum physics shows that this is not correct. If we rewound the tape of history and everything up until the present moment unfolded exactly the same, the next step is probabilistic, not deterministic. So Laplacean or hard determinism is known to be false. On the other hand, there is the Many Worlds deterministic version of QM, but this opens a new can of worms.
I think we can leave the question if the world is really determined or not when discussing compatibilism. Compatibilism states that determinism is a necessary condition for free will to exist. That is a pure conceptual standpoint in the first place. It is only a followup question if there is 'enough determinism' in the world that we really can have compatibilist free will.
I think we should just show that the idea of libertarian free will is empty, and that there is no empirical basis on it.
Yes. But the moral responsibility which is supposed to follow from it is the idea we can deserve what happens to us without any link to consequences. Then we agree, don't we? LFW is nonsense, and so is the idea that we really deserve praise, blame and punishment. Praise, blame and punishment are reactions from the community to actions of responsible persons. And responsible persons are those who are able to evaluate the consequences of their actions, including how the community sees their actions and how it possibly will react. Yep we agree :-)

GdB
I think I go beyond nobody really deserves anything. It makes sense to me to think about what sentient beings would deserve in a just world.
We get born with our one possible future ahead of us. We are capable of happiness and suffering. I think we all deserve our one possible future to be a happy one.

We get born with our one possible future ahead of us. We are capable of happiness and suffering. I think we all deserve our one possible future to be a happy one.
So if speeding makes me happy, even if I regularly cause serious accidents, that is no problem? Do I 'deserve' to be punished? Do I 'deserve' that my driver's license is taken away? And do I 'earn' the money I get for my work as a database specialist?
We get born with our one possible future ahead of us. We are capable of happiness and suffering. I think we all deserve our one possible future to be a happy one.
So if speeding makes me happy, even if I regularly cause serious accidents, that is no problem? Do I 'deserve' to be punished? Do I 'deserve' that my driver's license is taken away What I said was in a just world we'd all deserve to be happy. In this world we need to be deterred from speeding and if someone is unlucky enough for speeding to be their one possible future they need to be required to pay a penalty. But in a just world either they could speed without doing any harm or they'd be happy not speeding so wouldn't do it.

In a just world we’d all be happily determined to do what we should.
A just world can’t have what we should do being other than our determined path. Or happiness being other than our determined path.

Your ‘just world’ is a little bit too much of a fairy tale to be taken seriously, Stephen.

Quantum physics shows that this is not correct. If we rewound the tape of history and everything up until the present moment unfolded exactly the same, the next step is probabilistic, not deterministic. So Laplacean or hard determinism is known to be false. On the other hand, there is the Many Worlds deterministic version of QM, but this opens a new can of worms.
I think we can leave the question if the world is really determined or not when discussing compatibilism. Compatibilism states that determinism is a necessary condition for free will to exist. That is a pure conceptual standpoint in the first place. It is only a followup question if there is 'enough determinism' in the world that we really can have compatibilist free will. I think the compatibilist position can be summed up: We can do what we will, but we cannot will what we will. But, in order to have a will in the first place, we need causal determinism. This sort of free will is not enough for those who claim we need a more robust sense of free agency divorced from determinism. So people indeed operate with different definitions of free will. But what is causal determinism? QM clearly denies that the state of the universe at time T2 is uniquely determined by conditions at T1. We now know that this is not true. So surely QM must enter into the discussion somewhere, if not the discussion of free will, then the discussion of determinism by itself. Even putting aside QM, what would it actually mean to say that the state of the universe at T2 is uniquely determined by conditions at T1? Or to say that this state is determined at all, uniquely or probablistically, by conditions at T1? If we were to accept the Minkowski 4D spacetime model as more than a model, but actually telling us what the ontology of spacetime really is (not everyone accepts this, of course) then it seems we must accept that successor state are not "caused by" former states. Spacetime just is, and what looks like "causation" is really just a spacetime structure that can be modeled mathematically such that, if successor states did not look as if they had been caused by former states, then the actual world would be so chaotic that it is unlikely we would be here. So-called causal determinism, then, might be an anthropic effect. What this entails about so-called free will is open to debate. Probably you're familiar with the Libet exerpiments of the 80s, which demonstrates that our choices are generated in the mind prior to our consciously becoming aware of them. This seems like a big blow to most conceptions of free will -- certainly the libertarian conception -- but I think not. The experiments also showed that the conscious individual can veto the pre-conscious choice -- some have called this "free won't." The upshot, though, is that I don't think we have sufficient understanding of the mind and consciousness to rule out libertarian free will. The hard problem of consciousness remains unresolved. Until it is resolved, there remains room, perhaps a lot of room, for the idea that consciousness is qualitatively different from other, so-called causal, chains, and the Libet experiment shows that perhaps consciousness can uniquely override so-called causal chains. If this is so the possibility remains open that mind can be uniquely originative of actions. Certainly the Strong Free Will Theorem hints at this very idea, as far as I understand it.
I think the compatibilist position can be summed up: We can do what we will, but we cannot will what we will. But, in order to have a will in the first place, we need causal determinism.
Yep.
This sort of free will is not enough for those who claim we need a more robust sense of free agency divorced from determinism. So people indeed operate with different definitions of free will..
Of course there are different definitions around, we should be clear about it. To show compatibilism is correct, one must do several things: - show that we do not have the experience of libertarian free will, and explain why we have the illusion that we have it - show that the concept of libertarian free will is incoherent - show that the compatibilist definition of free will suffices as basis for our daily assigning of free will and responsibility - and several more.. Determinsm not being exactly the case is not a problem of for compatibilism: we just need 'enough determinism' (adequate determinism). So the whole QM stuff is of no importance, except when it would turn out that our actions are caused by purely random quantum events in the brain. Minkovski space is of no importance at all. There is a one to one relationship between world lines and events in our normal 3D + Time space. Laws of nature in our real world translate one to one to possible paths in Minkovski space. At most Minkovski space is a strong picture for imagining what it means that everything is determined. I couldn't care less about the Libet experiments. The experimental setting is as far off from real life settings as it can be: flex your hand for no reason at all. If I go to work in my car, i.e. I drive and have reasons for it, and my brain is attached to some highly sophisticated brain measure device, and neurologists can already say minutes in advance if I will take the curve to the right or to the left, I really cannot care. This is an action of free will, and you can read it from my brain: so what? It doesn't touch the compatibilist definition of free will at all. Libet's experiments show there is no invisible homunculus in the brain, that what I do has a causal 'foreplay'. What else would I expect? Libet's experiment is like an empirical proof that square circles do not exist.
The upshot, though, is that I don't think we have sufficient understanding of the mind and consciousness to rule out libertarian free will. The hard problem of consciousness remains unresolved. Until it is resolved, there remains room, perhaps a lot of room, for the idea that consciousness is qualitatively different from other, so-called causal, chains, and the Libet experiment shows that perhaps consciousness can uniquely override so-called causal chains. If this is so the possibility remains open that mind can be uniquely originative of actions. Certainly the Strong Free Will Theorem hints at this very idea, as far as I understand it.
We don't need neurological knowledge to know if we have libertarian free will: we know we haven't. On physical level everything is determined or Q-random. Both do not support LFW. Consciousness is not a soul that somehow steers (parts of) the brain.
We don't need neurological knowledge to know if we have libertarian free will: we know we haven't.
"We" "know" nothing of the kind, but "you" bloviate quite impressively.
We don't need neurological knowledge to know if we have libertarian free will: we know we haven't.
"We" "know" nothing of the kind, but "you" bloviate quite impressively. Here ]is a mirror. You wrote:
The upshot, though, is that I don't think we have sufficient understanding of the mind and consciousness to rule out libertarian free will.
My point is that we do not need more understanding of the mind and consciousness to rule out libertarian free will. Give me an example of how neurology could discover we have libertarian free will. What can a neurologist discover that reveals that: - some processes in the brain have no causal history, - are not in conflict with laws of nature, like energy conservation, - but correlate with my reasons for acting, and are consistent with, well, let's say my character Do we need more understanding of the mind and consciousness to see that this is impossible?

The problem with a necessary link between past and present can be expressed as a problem of luck. If there is a rule I shouldn’t break, in order not to break it I need the distant past to be a certain way so the past, in conjunction with the law of nature, leads to me obeying the rule. I didn’t choose the (distant) past nor do I have any control over it whatsoever.

  1. Circumstances not of my choosing would have had to have been different for me to have done otherwise.
  2. If circumstances not of my choosing had been appropriately different I would have done otherwise.
    Once that’s understood it’s obvious that bringing more luck into it in the form of indeterminism won’t help.
    And that’s it, people just dig their heels in because they don’t want to believe it.

This is a reaction on Pec’s post here]. I hate it to dig up posts that are covered with pages of useless photons…

Swartz is talking about counterfactual worlds, or non-actual possible worlds. If it's true today that tomorrow there will be a sea battle, the truth of the statement is grounded by what actually happens in the future. If there is a sea battle tomorrow, then today it's true that tomorrow there will be one; but if there isn't, then today it's true that tomorrow there won't be a sea battle.
Hmm, yes, but laws of nature are not timeless facts about individual occurrences: laws of nature are about types, not tokens. As Swartz says, in his own words:
Laws of nature are: 1. are factual truths, not logical ones 2. are true for every time and every place in the universe 3. contain no proper names 4. are universal or statistical claims 5. are conditional claims, not categorical ones
Point 3 and 4 are essential here, and the see battle in the year 200 BC does not fit to these points. So if in another possible world, where the same laws of nature hold, under exactly the same starting conditions the same will happen. When it doesn't, then the objects/events are not of the same types, or the laws of nature are not timelessly true. Therefore I think that on replay (as ever excluding QM) the same would happen. I believe I quoted this before, but let me do so again. This begins on Page 157 of Chapter 11 of Swartz's book "The Concept of Physical Law." (the entire book, along with two other excellent books by him, is downloadable for free. Just Google "Norman Swartz" and "books.")
Clearly, what is common in both Laplace’s and Mill’s versions of the Principle of Uniformity is the notion: “Same history, same future." It is implicit in Mill’s version and explicit in Laplace’s that, were the world to be ‘backed up’, it would have to have the same future as the one it in fact has had, that there was but one possible future for the world, given its initial state. Applying this notion to the present case, one might argue: Whatever led up to one’s choosing B could not have been followed by C. Were the world to be brought back to its state just prior to one’s choosing B, one would choose B again (and again, and again, for however many times the world were to be backed up). In this world, the state prior to B could only be followed by B, never by C. We cannot, of course, conduct the experiment. We can’t back up or reverse this world and approach a decision point for a second time to see whether the same outcome will occur the second time through. The best we can do is to postulate another possible world identical in its history to this world up to some time t and ask whether the Principle of Determinism really requires that that world have the same future, subsequent to t, that the actual world has. I claim that the Principle of Determinism requires no such conclusion. [Bold face by me.] The maximal amount of similarity possible, for the case under consideration, is another world, W’, identical in its history to the actual world Wa up to the moment t but in which at t I choose C. Both are worlds in which the Principle of Determinism is true: Their respective events are members of sequences that fall under physical laws. World W’, could have been our world, if only I had chosen C at t rather than B. In short, the Principle of Determinism does not entail “same history, same future," does not entail that a given history can have one and only one possible outcome. A given history cannot, of course, have more than one outcome; but which outcome it will have is not something predetermined by that history. True enough, the future of that world is logically implied by the description of its history together with that world’s physical laws. But, to repeat, there is no predestination in any of this provided those laws take their truth from what has happened, is happening, and is going to happen, and not from some other source. Provided the laws are not autonomously true, that is, are not logically prior to their instances, there is no predestination: Nothing ‘compels’ or ‘requires’ events to occur. In a categorical sense, within empirically discovered limits, we are free to choose from among a set of genuinely alternative actions. How much we can choose, and what is outside our sphere of choosing, is something we discover about this world; it is a contingent matter knowable only a posteriori. But the crucial point is that there are some, at least, genuine choices open to us. The Principle of Determinism and the Principle of the Uniformity of Nature are different principles; the latter is a stronger principle in that it implies the former. The Principle of Determinism does not imply the Principle of the Uniformity of Nature (no more so than that it implies forecastability in principle). (Note that the Principle of the Uniformity of Nature is often stated counterfactually. Recall, for example, Laplace’s statement quoted a moment ago. The Principle of Determinism states merely that events fall under universal laws. It does not state that events can have only one possible outcome.) A further difference between the Principle of Determinism and the Principles of Uniformity lies in the fact that being determined and being uniform are different concepts. This can be seen by noting that the two fall under quite different determinables. Being determined allows of only two different values: Either a situation (event, etc.) is determined, or it is not. But being uniform admits of degrees: One situation may be more, or less, uniform than another.
I think the difference lies in the fact that Swartz has a slightly other description of determinism. I think most people would define determinism as 'same conditions, same outcome'. However, Swartz defines determinism as: Every event (/state) is a second member of a sequence (/pair) that falls under a universal (i.e., deterministic) physical law But if he says:
The maximal amount of similarity possible, for the case under consideration, is another world, W’, identical in its history to the actual world Wa up to the moment t but in which at t I choose C. Both are worlds in which the Principle of Determinism is true: Their respective events are members of sequences that fall under physical laws. World W’, could have been our world, if only I had chosen C at t rather than B. In short, the Principle of Determinism does not entail “same history, same future," does not entail that a given history can have one and only one possible outcome.
Then I conclude that W and W' do not have exactly the same physical laws. I think his next remark confirms this:
True enough, the future of that world is logically implied by the description of its history together with that world’s physical laws.
But then I think this becomes an epistemic problem: say W and W' have developed exactly the same until moment t. That means until t no physical law has been show differently between W and W'. That means e.g. that scientists who (geniuses as they are...) know all universal and timelessly true statements about W until t, cannot predict what one will do, B or C. Now when in W one does B and in W' one does C, then in these worlds the scientists will discover different timelessly true statements involved in one's choice of B or C. So both worlds are determined (it is not that equally B or C could be chosen), but we do not know how. That makes sense when you see physical laws as descriptions, and of course the events that are described have ontological precedence over descriptions. But if we know all laws involved in one's choosing B under the given conditions, then one will choose B. But of course there is nothing to force one to choose B. So, yes, I stick to my opinion: if we could replay history, starting with exactly the same conditions, and where all (known and unknown) physical laws are the same, then exactly the same events would follow. So I think my difference with Swartz is only that I put in '... and unknown'. This fits to his article] in the 'Internet Encyclopedia of Philosophy':
For example, if you were to choose(!) to raise your arm, then there would be a timelessly true universal description (let's call it “D4729") of what you have done. If, however, you were to choose not to raise your arm, then there would be a (different) timelessly true universal description (we can call it “D5322") of what you did (and D4729 would be timelessly false). Contrary to the Necessitarians' claim – that the laws of nature are not of our choosing – Regularists argue that a very great many laws of nature are of our choosing. But it's not that you reflect on choosing the laws. You don't wake up in the morning and ask yourself "Which laws of nature will I create today?" No, it's rather that you ask yourself, "What will I do today?", and in choosing to do some things rather than others, your actions – that is, your choices – make certain propositions (including some universal statements containing no proper names) true and other propositions false.
Accentuation by me.

GdB, I’ll write a fuller response to this a bit later, maybe not until tomorrow. I can’t recall offhand whether I actually linked the two chapters from Swartz’s book, The Concept of Physical Law, from which I am quoting. Both chapters bear reading. They are:
Free Will and Determinism]
and
Predictability and Uniformity]
Swartz posits a philosopher from Mars who visits earth to study. When confronted with the determinism/free will problem, he is puzzled by it, for while Martians have concepts of both determinism and free will, they have no concept – no need of such concepts – as libertarian free will or compatibilist free will. Told that his views on the issues sound something like compatibilism, there is this passage:

But other problems proved well-nigh incomprehensible to him. In particular, he was deeply puzzled why anyone would think that determinism posed a threat to free will and to one’s being accountable for one’s actions (some of them anyway). In his view of physical necessitation, according to his theory, that is, of what a physical law is, there was no tension, no conflict, between someone’s being free and at the same time having her actions subsumable under one or more physical laws. From time to time, persons with whom he discussed the problem characterized his position as “compatibilism." But this only deepened the Martian’s frustration. He could perfectly well see why, for example, theories of certain Christian apologists might be called “compatibilist." To be sure, there is a prima facie tension in positing an all-good Creator and in admitting that this world is beset with evil. Reconciling those two views really did seem to him to be aptly described as “compatibilism." But the present case was different. It struck him as odd to call his way of looking at things “compatibilism." After all, one did not stick that particular label on his belief, for example, that doubts and itches should coexist.
Swartz (through his Martian counterpart) proposes that the so-called "problem" is misconceived. More later, but for now I want to comment on this:
So, yes, I stick to my opinion: if we could replay history, starting with exactly the same conditions, and where all (known and unknown) physical laws are the same, then exactly the same events would follow. So I think my difference with Swartz is only that I put in '... and unknown'.
Yes, but Swartz's point is that these "unknown" laws are unknowable in principle, because they do not exist, until a first instance of them is realized. But more on this later.
The problem with a necessary link between past and present can be expressed as a problem of luck. If there is a rule I shouldn't break, in order not to break it I need the distant past to be a certain way so the past, in conjunction with the law of nature, leads to me obeying the rule. I didn't choose the (distant) past nor do I have any control over it whatsoever. 1) Circumstances not of my choosing would have had to have been different for me to have done otherwise. 2) If circumstances not of my choosing had been appropriately different I would have done otherwise. Once that's understood it's obvious that bringing more luck into it in the form of indeterminism won't help. And that's it, people just dig their heels in because they don't want to believe it.
Well, some people dig in their heels, not because they don't want to believe, but because they deny it's true. ;-)

Answer on Stephen posting here].

I thought the beans conversation was interesting. At the moment we disagree over the fairness of withholding the medical treatment from the man who chooses the beans. I don't see any fairness, though I might if we delve deeper (I dunno).
First, to make my example not too extreme I only said that the insurance company does not pay, not that A can't get a treatment. Withholding medical treatment might be a death sentence, and such a 'punishment' for a stupidity is not proportionate. So I take it A can pay his treatment, even if it is expensive. So, now you are asking if it is fair, that somebody does not bear the consequences of his action (eating the beans in this case) where he knows all the ins and outs of his action: he knows he has a huge chance that he gets an allergic shock, and that the insurance company has excluded payment for this. Sorry, that makes no sense to me. If he is so stupid he must bear the consequences. The question if he CHDO can be asked, but not in the metaphysical sense you have in mind. E.g. when he was forced to eat beans (some stupid student's party or so), then the insurance company should pay (and take regress on those who forced him of course...). Only when he was forced to eat the beans he can justifiedly say I CHNDO. But without such coercion he was free not to eat the beans, but he still did it. This has nothing to do with:
If causal antecedents had been slightly and appropriately different he would have chosen not to eat them.
Yes, but Swartz's point is that these "unknown" laws are unknowable in principle, because they do not exist, until a first instance of them is realized. But more on this later.
Hmmm... A physical law is supposed to be timelessly true. For me that means 'being true' but 'not known' at the same time. One can say that a physical law does not exist as long it is not described, but I think that the regularities that could be described by physical laws already exist before a description of it is found. But I'll wait for your longer reaction.