Posts Tagged ‘free will’

Thoughts on “The Illusion of Conscious Will”

June 11, 2021

So as part of reading some books on free will, I had bought and read “The Illusion of Conscious Will” by Daniel M. Wegner.  I’m not going to go through this in detail since I don’t think the thesis ends up being all that interesting, and instead I’m going to give some general comments on the idea and the thesis and maybe one or two points that interest me.  So there also aren’t going to be a lot of quotes here.

Anyway, the main overall thesis, as I understand it, is that we have a sense or feeling of the cases where we are consciously willing an action and where we aren’t consciously willing an action and it is happening due to an external force or some automatic habit.  We use this to drive our sense that we have free will and are making choices:  we work things out in our consciousness and then feel that our actions follow from that conscious decision.  What Wegner wants to argue is that our conscious will doesn’t always align with our actual actions, and so spends a lot of time providing instances where they come apart, and where we think the action wasn’t conscious when it was and where we think the action was conscious when it wasn’t.  And I agree that these cases happen and that our conscious sense of what we do and what the full cause of our actions are can be out-of-sync.  I noted that this suggests that what we have are two different modules or things that talk to each other and can interpret the outcomes in different ways.  The problem is that in order to have this conclusion hold for the cases where we do think it obvious, Wegner needs to show that he can generalize from these cases to the general case, and he both fails to do so and his examples are not ones that strongly imply that sort of connection.

The main issue is that everyone — even the most staunch libertarian about free will — accepts that we have automatic responses and that there are cases where we can be fooled into thinking that those responses weren’t kicking in and weren’t influencing our decisions.  We can be fooled, so what he needs are cases where we wouldn’t think that was a case where we are fooled, and yet we are.  And yet most of his cases are not those sorts of cases.  He references the Libet experiments, but those aren’t examples of full-on decision-making like we think are the paradigmatic cases of conscious will (and I find them suspect as well).  Other than that, most of his cases are cases where people are manipulated or tricked, or else cases where he’s tracking influences as opposed to outright determinants.  He relies a lot on cases like hypnosis and even later on “spirit possession”, but most people will agree that these are cases that might interfere with conscious will.  So we aren’t at all surprised.  He also talks a lot about a lot of experiments where the participants are tricked in some way, but again in situations where it is deliberately vague where the true agency lies it won’t surprise anyone at all that sometimes we get it incorrect.  This does apply to most of his cases, which means that his case for this being a general feature of us is a bit weak.

I will concede this:  I think from the evidence he has shown it is clear that we don’t have a simple intuitive sense of when our conscious will is responsible for our actions and when it isn’t.  So this suggests, at a minimum, that the consciousness part of our brains isn’t getting something like a return code from the action saying that it was done in response to it, but is indeed trying to infer from what it observes what happened and if it was in line with what it suggested, and then concluding that things are working fine or if there’s a problem from that (we can all cite cases where our conscious will seemed to want to do something and we did something different, and were puzzled by that, and looked for an explanation).  So, yes, the two can come apart.  The basic problem with this thesis, for me, is that while they can come apart they usually don’t.  For the most part, the actions we take are indeed in line with our conscious will, and while Wegner gives some examples — mostly cases of hypnosis — where we seem to rationalize our actions after the fact, for most people for the most part we don’t do that.  We don’t need to.  So it’s difficult to imagine that our conscious will could really be so completely separate from our actions and yet stay aligned so much of the time.  Surely we would notice this far more than we do?

This also gets into the issue of:  if conscious will doesn’t cause our actions in any strong way, then what does it do?  Wegner tries to address this in his last chapter, but runs into the overarching issue with this sort of analysis:  in order for the experience of conscious will to have a purpose, it must have some sort of causal impact on what our actions ultimately are.  It may not need to be before the actions are taken, but at a minimum our analysis of that would have to make a change in the structure of the things that really determine the action so that on subsequent occasions we would act differently.  This, then, requires that there be some kind of causal connection between our experiences and our actions.  But if that is going to be necessary, then why can’t that be before the actions are taken?  If we are going to trust Libet, we may not be willing to accept direct causation, but we’d have to accept that conscious deliberation could very well set up the same sort of structures that can be activated when it is time to take the action.  And then for immediate choices, it could do so immediately before the action is taken and so indeed be responsible for the action that happens, even if some of the potentials are activated beforehand.  So, then, it wouldn’t really be an illusion of conscious will.  Our conscious deliberations would be responsible for our actions most of the time.  Wegner’s evidence might eliminate direct interaction, but unless he wants to declare conscious will useless then he needs some sort of causal connection that can be used both after we take actions and before preserving conscious will.  And conscious will is so prominent and takes up so many resources that it is difficult to imagine that it could do absolutely nothing for us.

So I concede that Wegner makes a good case for separating conscious will from action.  So, at a minimum, those are probably in different “modules” in the brain.  But I don’t think he can get from there to insisting that it’s all an illusion, and don’t think his examples or arguments get us there.  The examples are not of the typical cases of conscious will, and his discussions of how the experience of conscious will could be useful requires a causal connection that we could use to actually implement conscious will, making it no longer an illusion.  Thus, I don’t think his book gets him to where he seems to want to go.

Final Points from “Four Views on Free Will”

May 14, 2021

So, I finished off this book by reading the last forty or so pages where each person responded to the others.  I’m not going to get into this in detail because for the most part they didn’t really address issues that I personally was bothered or concerned about, and since obviously I’m not really on board with any of the views I didn’t find the defenses compelling, but neither did I find the defenses of the defenses compelling.  So instead of doing that, and since I’m going to be doing more with free will over the next little while, let me pull out a couple of points of interest that were talked about in that section and then let this book go.

Kane, for obvious reasons, takes on the Frankfurt cases directly, with an interesting argument.  Let me remind everyone of what these cases are:  they are cases where in general there is some kind of trigger set so that if an agent would make “the right choice” on their own, they are allowed to make it, but if they are going to make “the wrong choice” the trigger is hit and things are manipulated so that they instead make “the right choice”.  These are aimed at showing that someone can have moral responsibility even if they really could not do otherwise, and so that there really aren’t alternate possibilities.  There is really only one possible outcome given this, but nevertheless if they choose the action freely we would still hold them morally responsible for their actions.  This, then, is meant to show that in order to have the right kind of choice for free will — the kind of choice that preserves our moral responsibility — we don’t need to actually have real alternative possibilities.

Kane replies that the issue here is that all of these examples rely on there being some kind of phenomena that the trigger can use to determine that the “wrong” decision is being made before the decision is actually made so that the trigger can instead manipulate the decision-making process to produce the “right” decision.  However, this actually assumes a deterministic universe, because that point has to be one where the decision is completely determined but also has to be before the decision is actually made.  Libertarians about free will and anyone who thinks that free will requires alternate possibilities will point out that if this is true then there really aren’t any alternate possibilities after that point.  For the examples to work, the trigger must work 100% of the time and by necessity.  Otherwise, it’s indeterminant and based on probability, which means that there indeed could be an alternate possibility.  And it can’t wait to trigger until after the decision is made because that would be an obvious example of someone’s free will being overridden.  So Kane argues that this is a bit of a problem for those sorts of examples.

While I do think that’s true, I don’t think it really refutes the point.  The idea here is more that even if there was only one possible outcome we would intuitively still consider the person to be morally responsible if they happened to choose using their normal decision-making processes as opposed to some kind of external decision-making process or strong external determining factor.  And I think that’s true, and it’s important for compatibilists, because in general those theories rely on arguing that the decision is free if it was done by our decision-making processes working properly and without undue external influence (ie they aren’t being overridden by some external process or cause, like we see in the Frankfurt examples).  So these examples show that if the person made the decision themselves but there was a case where their decision would be overridden to that decision anyway so they’d always only make a specific decision we’d still want to differentiate the case where the person’s internal processes made the decision instead of the external cause.  So, as I noted earlier in this examination, I think the alternate possibilities model is being overinterpreted to insist that there be a “real” alternate possibility or else the decision isn’t free, when really that’s meant in a far more colloquial or folk way where we think that it must be a real possibility that the decision — as per the normal decision-making processes — could come out with a different decision than the one it did.  Kane is right that Frankfurt cases can’t actually produce that case, and so it must be a different notion of alternate possibilities that we’re after.  So I do side with the idea that what we’re really after is that the person themselves must be responsible for the decision in an important way, which I also then agree with Kane that that implies the sort of alternate possibilities that I outlined above:  in order for the decision-making processes to be properly responsible for the outcome it must be reasonably possible that they could have ended up with a different decision.  And from here we can ask if that’s possible if determinism is true.

The other thing that was raised by a couple of them is the idea that we don’t hold people morally responsible for their actions if, say, someone puts a gun to their head and threatens to kill them if they don’t.  As I noted earlier, this is a very common example used to show that someone in that case really doesn’t have free will.  But speaking as someone who is at least Stoic-leaning, it’s difficult for me to accept that because the Stoics would hold them morally responsible for their actions, and the fact that they do that is in fact one of the more appealing things of Stoic moral philosophy for me, because it follows from the idea that someone is morally responsible for what they do, not what other people do.  So if they choose to commit that immoral action because of that threat, they did choose to do that and so are responsible for it.  So the question, then, is whether or not they should be considered morally responsible for it.

I think it’s educational to look at how most people intuitively react to the Stoic view.  They don’t react to it as if it’s contradictory, but instead seem to react to it as if it’s heartless.  They don’t tend to respond to that with an argument that no, the person isn’t responsible and really couldn’t have chosen otherwise, but instead that it’s heartless and cold to expect someone to do that.  This carries on to cases where someone, say, is stealing for themselves or especially for their children.  It’s not that they couldn’t choose to not steal that food or that money or whatever, but that it’s difficult for us to expect them to not steal that under pain of accusing them of immorality.  So it looks like this is less an example where they lose responsibility, but instead a case where they lose morality, or at least that morality should not judge them for that action.

We can carry on from this to look at how this works wrt kleptomania vs these cases.  The kleptomaniac is clearly not responsible for their choice to steal things, as they have a compulsion that they cannot resist.  They literally cannot choose otherwise.  But in the other cases when pressed everyone has to admit that the person could choose otherwise.  They are not under an irresistible compulsion to steal that thing.  But it seems like asking too much to demand they do that.  So I’ve captured the distinction as I see it with the statement that we understand why they made that choice or even felt that they had no other choice, while understanding that they do indeed have another choice.

I was recently reminded of another case that might capture this distinction, from the show “Beast Wars”.  In “Code of Hero”, the character of Dinobot has been struggling with the idea that if a set of disks that tell the future unerringly capture that future, then everything he does has already been determined and so he has no free will.  However, through various means he discovers that, yes, the future can indeed be changed.  Heavily damaged, he discovers that his enemies are going to attack and possibly eliminate the creatures that will become the humans and thus ensure that the Decepticons win the Great War.  His allies will arrive too late, but if he tries to engage his enemies he will almost certainly die.  And he says this:

The question which once haunted my being has been answered. The future is not fixed, and my choices are my own… and yet, how ironic! For I now find, I have no choice at all!

He didn’t literally have no choice.  He could have stood down and not disabled the internal process that would have preserved his life but left him unable to fight.  But based on who he was and what the situation was, he was going to choose to defend them.  He was, in this sense, perhaps morally compelled to do what he did.  And we’d obviously consider him morally responsible for doing the moral thing even though, for him, it was as irresistible a compulsion as the gun to the head or the starving children.  So if we’d consider someone morally responsible for this when the compulsion is moral, surely we have to do the same when the compulsion is immoral.  In both cases, the person is facing a strong influence, not a completely irresistible compulsion.

So, for me, the focus on moral responsibility in the modern discussions, at least, always focuses far too much on the moral part and not enough on the responsibility part.  For free will, the issue was that we felt that determinism would remove responsibility, because we couldn’t be held responsible for those actions if it wasn’t us who decided them.  In these other cases, it’s morality that’s being challenged, where we don’t think that a proper morality can reasonably demand such resistance from its adherents.  So despite everyone always raising those as examples where we don’t have free will or don’t make free choices, I now insist that we, in fact, do.  And thinking that we don’t is responsible for confusing the issue of just how free will and determinism are claimed to be incompatible.

Anyway, that’s all I want to talk about wrt this book.  I actually did enjoy it, even if I didn’t agree with it all that much, and do think it a decent resource for someone who wants to get a basic idea of what this free will thing is about (some parts might be too technical, but you also get advocates for each view all talking about it in one place, which I think makes up for that).  I will be moving on to more incompatibilist views in the near future, including a book-length one by Pereboom.

Manuel Vargas on “Revisionism”

May 7, 2021

The last of the four views in “Four Views on Free Will” is Revisionism by Manuel Vargas.  Now, this was a view that I had never heard of before, so as I said last time I was looking forward to finding out what it was.  After reading the essay, I don’t find it a particularly strong position and, in fact, wonder if it really counts as a distinct view and category at all.

First, let me describe what it is.  As best I understand it, it’s the idea that we might be wrong about our intuitive view of free will, like we were wrong about our intuitive views about things like water and a host of other things that science managed to clarify.  So, in order to solve the issues we’re having with reconciling our conception of free will with the determinism that science says is the nature of our mental operations, what we need to do is come up with a scientifically informed definition of free will that respects and captures our intuitions while clarifying the concept enough so that the problems go away.  So the definition needs to be close enough to the intuitive definition so that we can recognize it, and yet different enough that we can see that once we understand the real definition of free will we can see why our problematic definitions are no longer relevant.

I always thought that there was another position that described that:  compatibilism.  Pretty much every compatibilist I’ve ever read has in some way pointed out the errors in the folk or intuitive view of free will that cause the seeming contradiction between free will and determinism.  For the most part, the compatibilist project is to capture what is most vital about the folk view of free will and explain why a) that’s all that’s important free will and b) that core is compatible with free will.  Dan Dennett even explicitly says that his position captures the “varieties of free will worth wanting”.  So it has always seemed to me that even if it isn’t explicitly part of the definition of compatibilism, practically they are going to have to significantly change the concept of free will as we understand it to resolve the seeming contradiction.

Now, Vargas argues that it can’t just be compatibilism because we can have people who are incompatibilist revisionists and people who are compatibilist revisionists.  But I think that this actually gives the game away.  For a philosophical problem this intransigent, it’s pretty much a given that we are going to have to clarify and adjust our concept of free will to resolve the apparent contradiction.  So pretty much everyone who doesn’t want to abandon the concept entirely is going to argue that it will have to be significantly altered.  So Vargas is going to have to argue that libertarians and compatibilists want to mostly preserve the concept, hard determinists/incompatibilists are going to want to abandon it, and revisionists will have to want to change it more than libertarians and compatibilists would accept but still maintain it as a useful concept that meaningfully maps to our intuitive view to avoid being a de facto hard determinist/incompatibilist but simply try to preserve the name for some reason.  That’s … a pretty narrow line he’s going to have to tread.  And worst of all, once he manages to establish that new concept, he’s going to have to answer the question of whether it is compatible with a deterministic world or not, and so he’s going to have to address the other question and end up holding one of the other positions anyway.

So that’s why I don’t think it a particularly useful new view/position/category in the free will debate.  Everyone who wants to preserve free will is going to agree that the concept needs to be altered, and won’t disagree with the analogy that it might need to be changed as radically as the concept of water was altered, so revisionism is not really a position that they’d disagree with.  And once he’s done revising, he’s still going to have to answer the main question that the other three positions pretty much exhaust all the possibilities for.  So it’s not distinct enough for its own category, and ultimately has to take one of the other three positions anyway.

Still, let me talk about his proposed revision.  I think that he falls into the trap of worrying too much about morality and not enough about responsibility, because his revised concept/definition is very strongly moral.  He follows on from the definition we’ve talked about before relating it to praise and blameworthiness, and clarifies it to say that a choice is free if it is responsive to praise and blame relating to its moral status, and thus making it so that further decisions are more likely to take those moral considerations into account.  So if the praise or blame alters how they make similar decisions in the future, then we have free will (as best I understand it).  The main issue here is that while that ties it tightly to morality, it doesn’t tie it tightly to responsibility, and as I’ve argued responsibility is the main stumbling block here.  To illustrate this, we can ask this question:  what happens if they are praised or blamed and it doesn’t make them more responsive to moral reasons?  Are they responsible for that outcome, or not?  If they aren’t, then they wouldn’t have free will by any reasonable definition.  But Vargas’ view doesn’t tell us whether they are responsible for that process or not, and so it doesn’t really give us a concept that we can use to settle the question.  And solving that issue requires adopting a stance that would fall into one of the other views.

Another reason why I think he focuses too much on morality is that he says that one of the main objections to his view is that it is “merely consequentialist”, which he immediately tries to rebut by pointing out that the moral position of consequentialism might, in fact, turn out to be the right one.  But I’m not convinced that that’s what’s meant by that criticism.  A more reasonable criticism there is that he is trying to define free will not by what it is, but instead by what the results or consequences of free will are.  It is reasonable to say that someone becoming more responsive to moral praise or blame or in fact even being responsive to moral praise or blame is in fact what free will does.  But that wouldn’t explain what free will is.  To illustrate this, ask what would happen if we found that something else was also able to do that (such as a computer algorithm).  Would we have to argue that the computer algorithm also is free will, even if it works completely differently and so in a way that it is clearly not the same sort of thing as what we originally called free will?  This has actually been a problem for the philosophy of mind view of functionalism, since it says that anything that performs the functions of consciousness counts as consciousness, but qualia and actual experience is not generally captured by the functions functionalism talks about, so does that mean that things that have the functions that functionalism captures but don’t have qualia are, nevertheless, still conscious?  And if they are, do we need to invent another term to capture consciousness with qualia since it’s clearly distinct?

So that might be the objection here.  Vargas’ concept might tell us what free will does but doesn’t tell us what free will is.  And even in his example of water the new concept clearly told us what water was and derived from that what water does and why it does what it does.  The concept Vargas advances doesn’t seem to do any of that, and so doesn’t seem like it would be a satisfactory definition even by his standards.

So those are the four views.  The last forty or so pages are each philosopher’s responses to the others, that I was quite careful not to read before writing about all of the views.  So I’ll do a short summary of their responses in the next post before likely taking a break from talking about free will to read the other books I have.

Derk Pereboom and Hard Determinism/Incompatibilism

April 30, 2021

The next view in “Four Views on Free Will” is Derk Pereboom’s view on what he calls Hard Incompatibilism and I call Hard Determinism.  I’ll get into that terminological difference in a minute.  First I want to note that I actually have one of his books on my reading queue, and so it is quite possible that when I read that his view will be fleshed out and I’ll have a different view on it.  However, for now, all I have to go on is what he says here.

So, let me start by looking at why he wants to call the view “Hard Incompatibilism”.  As I understand it — again, it’s from an actual book and this is a simple blog post so quoting things in detail isn’t going to be happening — the main reason is that he is taking the position here as being more opposed to the Libertarian and Compatibilist view in the sense that it holds that free will doesn’t exist.  Thus, he thinks both that free will wouldn’t exist if determinism is true but even if determinism isn’t true there is at least one case — if the source of the indeterminism is not the right sort of source — where we don’t have free will either.  The problem I have with this renaming is that when it comes to the actual form of the debate, it ignores that both Libertarians and what I’d call Hard Determinists share what should be call the Incompatibilist position that free will is incompatible with determinism.  Libertarians accept that and say “So much the worse for determinism” and Hard Determinists accept that and say “So much the worse for free will”.  Pereboom’s definition makes the Libertarian and Compatibilist views seem far more similar than they really are.  Pereboom is more anti-free will in general, and calling that position Hard Incompatibilism confuses what the traditional debate was about.  I don’t disagree that perhaps a classification is required for people who think free will as traditionally understood is just plain impossible, but don’t really think it is beneficial to insert it into the classic positions by redefining those who at least used to be called Hard Determinists into a new one.  Pereboom would need, then, to create a new one to insert himself into, but I’m not sure it’s a significant difference for us to have the two categories of Hard Determinist and Hard Incompatibilist, but even to Pereboom the two positions are not the same.

Moving on from that, though, Pereboom also wants to clarify what sort of thing we need to have free will.  As with the other two, he doesn’t like the “alternate possibilities” idea, instead want to focus on the “agent is ultimately responsible” idea, which I have no issue with.  However, he wants to talk about that responsibility being moral responsibility, which always raises alarm bells for me.  A lot of Hard Determinists like to make the split between responsibility and moral responsibility because they want or need to claim that the agent is in some way responsible for their actions but aren’t morally responsible, and often attempt that — Jerry Coyne is a really good example here — by trying to eliminate morality from the picture:  we aren’t morally responsible for our actions if determinism is true because morality becomes meaningless and so there is nothing moral that we can use to make any kind of responsibility moral responsibility.  This ignores that the main reason we think that you can’t have moral responsibility if determinism is true has always been that we don’t think that the person can be properly or meaningfully responsible for their actions.  So we are arguing that it’s the responsibility part of moral responsibility that’s lacking, not the moral part.  After all, it’s pretty easy to see that even if morality is true we could describe a process as maximizing utility or treating other agents as ends as well as means, so the specific moralities wouldn’t go away, but since morality is based on oughts and ought has to imply can, if they couldn’t do anything else than what they do then they can’t be held morally responsible for it (this is what drives the “alternate possibilities” classification).  So we always have to be aware that proper responsibility is the challenge here, not morality.

Pereboom is actually better at this than most I’ve read, as he doesn’t rely as much on us having real and meaningful responsibility (although he does talk a lot about what we should do if we properly understand Hard Incompatibilism, which always implies that we are responsible for what we do and don’t do).  But his definition of moral responsibility, I feel, isn’t all that helpful.  He argues for adopting the “blameworthy/praiseworthy” definition, where we are morally responsible for our action if we could properly be blamed or praised for it.  The problem with this is that it’s still pretty vague.  It sounds good for thought experiments where we can talk about whether the person should be blamed or praised, but it opens up the potential for all sorts of confusions when we try to look at the cases to see just why someone should be blamed or praised for their action, and confusions right around the precise cases that those debating free will will be disagreeing about.  We probably need a more robust notion of “responsible” in order to make any headway on this issue.

Case in point:  Pereboom gives four cases that he thinks forms a progression that shows that in most common cases a person is not morally responsible for a specific action (he seems to be using an example from Clue and so talks about Professor Plum and Mrs. White).  As I understand the cases, they are this:

1) Neuroscientists can deliberately manipulate Professor Plum’s reasoning process to make him have desires, at least, that are more rationally egoistic than moral, even though sometimes — I guess either when they don’t manipulate him or when the desires that are there at the time happen to work out that way — he can act morally.

2) Instead of directly manipulating his reasoning/desire-formation, they instead build in a set of desires that strongly bias him towards rationally egoistic choices, although he can overcome them with his other decision-making processes.

3) Instead of those desires being implanted by the neuroscientists, he gets them from training from his culture and upbringing.

4) This is all determined by physicalist determinism.

Pereboom is aiming this at compatibilism, and he argues that if we have to follow the chain down in all of these cases it means that intuitively we think that these are cases where the person is not morally responsible, in particular by the definition that compatibilist cases say that our decisions should follow from the person’s character and in all of these cases the rationally egoistic cases the decisions are following their character but are more or less determined outside of them.  The big problem I have with these cases is that I think that proper moral responsibility comes in at Case 2), and so the rest of his chain fails, but as a Libertarian I also think that there is no moral responsibility in Case 4).  The reason for this is that I see both Cases 2) and 3) and essentially defining tendencies for Professor Plum, but his normal decision-making processes can overcome those tendencies (because in the cases Pereboom specifically says that he can).  Since he can overcome those tendencies and if he’s observing any of his actions at all he could come to know that he has those tendencies, he is indeed properly responsible in those cases where he doesn’t overcome those tendencies.  In fact, while I may be misremembering — and I didn’t talk about it — I think that for Robert Kane making free decisions is entirely about interrupting the causal path through an act of will, whatever that means.  So we can indeed claim that someone is praise or blameworthy for their decision to follow a tendency and not override it.  And even for the character argument, when one overcomes tendencies and when one doesn’t is a reflection of their overall character, so I don’t even think that the compatibilists that Pereboom aims this at will be refuted by these cases.

For Case 1), I think that Professor Plum isn’t morally responsible because no matter how you interpret the case it’s more than overcoming tendencies.  My first blush interpretation was that the neuroscientists go through the entire reasoning process for him and come to that conclusion, even though sometimes they come up with the more moral option.  In this case, the entire reasoning process has been subverted and if he isn’t responsible for his reasoning then he isn’t responsible for his decisions.  I also think that the more common — at least in my experience — view of compatibilism would agree, since it would insist that the decision-making process must be functioning properly and in that case it isn’t.  If they are merely determining all of his desires, then I would again argue that he isn’t morally responsible because even as a Libertarian we should reason based on our actual desires, and his desires are completely determined by others.  That’s more than his having a simple tendency.  It is impossible for him to ever want anything else than what they determine he wants, and naturally he will try to act based on the desires he has.  In Cases 2) and 3), he can form new desires and potentially remove old ones, which is not the case here.  And for the compatibilism I’ve talked about above, implanting desires clearly invalidly manipulates the reasoning process, and so the decision-making process isn’t valid, and so it’s not a free choice either.

For 4), compatibilists, of course, will argue that once we suss out how this system will all work, we will see that in the physically determined case the decision-making process is working properly.  For libertarians like me, what we see is that there’s no agent-causality at all and so no real decision-making process, as everything is determined by things outside of the agent.  So Pereboom’s fourth case does capture the clash between compatibilists and incompatibilists, but his chain doesn’t really show that compatibilists are forced to accept that Case 4) is the same and is similarly problematic to Case 1).

The last thing I want to talk about is Pereboom’s attempts to deal with the argument from phenomenology, which is that it really, really feels like we really make decisions.  He uses at least twice Spinoza’s argument that maybe once we understand the causal process, we would then see how the phenomenology is illusory or at least misleading.  The problem is that the phenomenology of something is what would get preference unless we have a very good explanation for why it shouldn’t get preference.  If I stick a stick into water and it looks like it bends, I’m perfectly justified in thinking that it really does bend until I get contradictory phenomenology or a good explanation for why it looks like it bends but really doesn’t or at least get a good argument for why the stick case is similar to other cases with light and so it would only look like it bends.  What Hard Determinists are trying to do here is appeal to the latter case, arguing that the components are all determined and so the process itself should be, even though the phenomenology strongly suggests otherwise.  This, though, is challenged by how consciousness itself does seem to be pretty special and so not standard, and so opponents can suggest that those material, determined things are correlations, not causes, and so don’t trump phenomenology in that way.  In short, Spinoza’s argument will work once Hard Determinists have provided sufficient evidence to think that the phenomenology is wrong, but Libertarians and Compatibilists will deny that they have (Libertarians because they feel Hard Determinists invalidly assume that mental things are deterministic, and Compatibilists because they feel that Hard Determinists invalidly assume that determined processes can’t work in a way that is consistent with the phenomenology we have).  So I felt his dismissal of the phenomenology to be too quick, and in fact it really seems like an attempt to assume that his position is correct and then demand that his opponents demonstrate that he’s wrong, which isn’t really a fair demand.

The last one is a new one to me, Revisionism.  I look forward to discovering what the heck it actually is …

John Martin Fischer and Compatibilism

April 23, 2021

So the next view in “Four Views on Free Will” is John Martin Fischer’s view on Compatibilism, which includes his unique view of compatibilism called semi-compatibilism.  What I’m going to do here is talk a bit about why compatibilism is attractive (since he does that at the beginning of the chapter) and then talk a bit about semi-compatibilism since I don’t think it really achieves what he wants it to achieve.

First, Fischer focuses on arguing that it seems obvious that we make choices, but also notes that determinism could be true, so a compatibilism that allows us to retain decision-making abilities even if it turns out that determinism is true.  Unfortunately, this way of talking about it seems to bias him towards simply making it possible for us to retain “free will” if determinism happens to turn out to be true, which isn’t going to satisfy either of the other two main sides — yes, there is a third one lurking out there since it’s the last segment of the book — because hard determinists think that determinism is true and libertarians are going to reply that there definitely seem to be types of determinisms that would eliminate all free will, and so finding that there might be types of determinisms that wouldn’t isn’t all that impressive, especially since the hard determinists are talking about determinisms that really do seem like they would eliminate all free will.  So compatibilists need to do more than simply carve out an exception in some determinisms for free will, but instead need to show why the determinisms we are likely to have don’t cause any issues for free will.  And I think that this impacts his semi-compatibilism, which I’ll get to in a moment.

Before that, though, I’d like to note that he is somewhat right about the appeal of compatibilism, with it being a view that bridges the two positions.  But I think that the main reason it is appealing to many — although many find it implausible rather than plausible — is a basic philosophical methodology.  Our intuitions strongly support the idea that we have some sort of meaningful free will and that we actually make meaningful choices.  However, a lot of the intuitions from science and even every day life suggest that for everything else things are pretty much entirely determined causally (we can ignore quantum phenomena because they don’t seem to have much if any impact at the macro level and we don’t observe quantum phenomena in our everyday lives).  So it raises the obvious question that if everything else in our world is determined, why would we be any different.  So what we have are two strong and pretty plausible and intuitive arguments for either side.  And when we have such a clash of arguments where we can make strong arguments for either side — think Kant’s Antimonies — and can’t see where either side’s argument goes wrong there is a natural philosophical tendency to ask whether maybe the problem isn’t in the arguments for each side, but instead in the arguments that we are making to say that they can’t both be true.  Maybe we’re wrong about that and they can both be true.  And all compatibilisms at their heart say exactly that:  both of those intuitive and strong arguments are, indeed, actually true.

Which means, of course, that the debates against compatibilism really should focus on whether it is indeed the case that both can be true, and so on whether we can have a meaningful free will even if things are totally determined.

Which leads to his semi-compatibilism.  The problem with it is that as I understand it he relies on a distinction between regulative control and guidance control, and argues that while determinism would kill regulative control (which requires access to alternative possibilities that determinism would eliminate) from guidance control which he argues would not or would not need to be impacted.  But those terms can be a bit confusing, so I’d like to focus on his example of the difference (from page 58) which is a Frankfurt-type example (I’ll be summarizing it as I understand it and not quoting it because I have an actual book and that would be a real pain for me):

Imagine that someone is going to vote, but hasn’t decided who to vote for yet.  Unbeknownst to them — but knownst to us — someone else has implanted a chip in their brain that will notify them if they decide to vote for Party B so that they can instantaneously flip a switch and change the decision so that they will instead vote for Party A.  So if they go through their process and decide to vote for Party A, then they would clearly have made a free choice because the chip in their brain is irrelevant to that process.  However, it’s also clear that they were going to vote for Party A no matter what happened.  So the fact that there was only one possible outcome doesn’t mean in this case that they had no free will.  I think Fischer at least analogizes regulative and guidance control to these two examples, making the case where their deliberation process chooses Party A the guidance case and so in a case where we have only that sort of control it looks like we would still be making a free choice even if we really “couldn’t choose otherwise”, since no matter what happened the person was always only going to choose to vote for Party A.

The problem I see here is that I don’t think that any real kind of determinism is going to allow for even that kind of control, and so semi-compatibilism can’t get off the ground.  Let’s alter the example to add that someone else inserts another chip that completely controls the person’s decision-making process to walk them through a decision-making path that ends with them deciding to vote for Party B.  They walk through that process and decide to vote for Party B, and then the chip is triggered and they end up having that decision changed to vote for Party A.  So they didn’t freely choose to vote for Party A at all, I think we can all agree (even Fischer).  However, let’s then change it so that the person who inserted the second chip instead sets it so that their decision-making process gets them to vote for Party A.  As before, the first chip is irrelevant to the outcome, but now we have to ask if the person made a free choice to vote for Party A.  And it looks like they didn’t, because their entire decision-making process was determined by an agent outside of themselves to without fail come to that specific conclusion.  It’s not only the case that they couldn’t have done otherwise, but also that they couldn’t have decided otherwise.  And if they couldn’t even have decided otherwise, there seems no room for any kind of control, whether regulative, guidance, or anything else.

What this shows is that it isn’t sufficient to show that a decision was the result of a deliberation and our decision-making processes to show that we should think that it’s free.  It’s entirely possible that under determinism our entire decision-making process was determined before it was even engaged, and so that it’s just “going through the motions” to produce a predetermined decision, and so has no causal impact on the outcome at all.  And if nothing in our decision making processes has a real and meaningful causal impact on the outcome, then our decisions don’t seem to be free at all, not matter what type of control we talk about.

So Fischer would need, at a minimum, to show how we can have a specific type of control that still remains even if determinism is true.  And he would need to address it in the forms of determinism that are more likely to be true, and hard determinists and libertarians both agree that that’s the very strict causal sense that would determine the entirety of our decision-making processes and so would look a lot like the example of the chip that I added to his Frankfurt example.  I did not see how he could escape that case in his chapter, and so see his semi-compatbilism as having as serious issues as other compatibilisms and so that it doesn’t really give us any advantage here.  It’s true that he can concede that we would lose one type of control and so not have to argue over that anymore, but he doesn’t seem to have shown that the same mechanism that eliminates regulative control doesn’t also eliminate guidance control and every other sort of control that we could have as well.

Next is the position that I have the least sympathy for, which is hard determinism or, as they — and many others, I suppose — call it, hard incompatibilism, which I don’t like as a term because libertarians are hard incompatibilists as well, which is precisely why they are libertarians.

Robert Kane and Determinism

April 16, 2021

I still intend to write about Mark D. White’s discussion of Batman, but several months ago there was a post at Tippling Philosopher I believe about free will where someone included a bunch of references to read on free will, which I recently decided to pick up and start reading, and I want to comment about the first chapter that I’ve read.  It’s from a book called “Four Views on Free Will”, and contains entries from Robert Kane, John Martin Fischer, Derk Pereboom, and Manuel Vargas.  I’ve read Robert Kane’s chapter, as should be obvious from the title, and want to talk about it.

Now, you might wonder what I’d have to talk about here, because Kane is a libertarian about free will and so am I.  But his defense of free will raises some interesting points and I think fails in a very interesting way, so that’s what I want to focus on.

To start with, I think he is completely right that libertarians should focus less on the “alternative possibilities” angle where we thinks that things could have happened differently and more on where the ultimate responsibility for the action lies.  If the ultimate responsibility for the action lies outside of the agent, then we haven’t really made a choice at all and pretty much everything that we want any kind of free will for is lost.  This also aligns better for a debate with compatibilists, because it gives them something to care about that they cannot simply assume into existence.  Focusing the debate this way forces the compatibilist to find a way to say that the agent is still ultimately responsible for their actions in a determined world, while the hard determinist has to accept that it isn’t merely moral responsibility that we might lose, but responsibility entirely, and so are forced to reject agent responsibility or else accept what is essentially a compatibilist position.  This, then, would focus the debate more on the specific issues that each side must address and where they actually do disagree (as some hard determinists, like Jerry Coyne, end up blurring the lines to avoid the consequences of their positions).

Another reason to do that is that the alternative possibilities angle was in general more of a heuristic than a position statement, but oftentimes people have taken that as if it was a position statement and so tried to refute the idea of free will simply by trying to split hairs over what it would mean to be able to choose otherwise.  This goes all the way back to at least John Locke with the question over whether someone who decides to stay in a room but unbeknownst to them the door is locked and so they actually can’t leave is staying in the room of their own free will (which would be absurd from both the standard free will perspective and the perspective of where the ultimate responsibility arises).  It also leads to the heuristic of libertarians wanting the world and the entire past to stay the same and yet for there to be a different outcome, which doesn’t work either.  It presents libertarians as insisting on decisions being indeterminate, but libertarians importantly want to make sure that decisions happen because of reasons.  They’re perfectly willing to accept that in order to make a different decision, our reasons would have to change.  They just don’t want something completely outside of us to change our reasons without us having to make an act of will to at least allow the change.

So Kane is also correct, I think, to accept that at the time of the decision our choices might indeed be determined by our reasons (meaning our beliefs and desires) but that the exertion of our will is required to form, at least, many of these reasons.  Kane calls these Self-Forming Actions and might imply that they don’t happen all that frequently, but I think at the very least we can both agree that we can take SFAs to either block the formation/acceptance of a reason or change one that’s existing.  However, if this doesn’t happen before we make a decision we will indeed act according to the relevant reasons that we have that we considered at the time while doing deliberations.  This, then, means that even libertarians will concede that if you know a person well enough you will be able to predict what someone will do, because they will ideally act according to the reasons they have.  For example, if the cafeteria is serving a choice between Sloppy Joes and poutine — to reuse an example from a Philosophy of Science class I took long ago — anyone who knows me will be able to predict with almost perfect accuracy that I will choose the poutine.  First of all, I really like poutine.  Second of all, I don’t really care for Sloppy Joes at the best of times.  Thirdly, even if I was tempted to go for the Sloppy Joes, it would be risky for me to try that at a cafeteria since I wouldn’t be sure how spicy it would be or if it would contain fried onions which make me sick.  So if someone knew my reasons, they’d know what I was going to choose.  But they would indeed need to know my reasons.

Now, Kane tries to defend libertarianism against the charge that it requires or posits something outside of the normal mechanisms by positing one that works inside our normal mechanisms, which is where I think he goes wrong.  He bases it on the idea of parallel processing in the brain, and as I understand it posits that what happens in decisions is that we have two or more different threads processing at the same time, and when that parallel processing is resolved we’ve made a decision.  He spends a lot of time trying to show that this is indeterminate in the right way without being just by chance, but I think the big problem with it is that it’s actually far too deterministic.  I think that both compatibilists and hard determinists could easily accept that mechanism as in fact proving their case, and not Kane’s.  Compatibilists could argue that this is indeed how our decision-making processes work, as we have multiple parallel processes that work through the various options and a resolution mechanism that settles on the ultimate decision, but because these are all brain functions and all brain functions are determined as per science.  The hard determinist could use this method to argue that all we have is an illusion of decision-making, caused by parallel processes happening and so producing the experiences of making a decision, but ultimately all of that is determined and so our decision-making processes don’t really do anything, and so choice is an illusion.  It doesn’t seem like a good move to try to preserve libertarian free will with a method that easily supports the other two positions — and, potentially, supports those positions better than libertarian free will.

The problem is with trying to come up with a method that fits with the “normal” mechanisms, because any attempt to do so will feed right back to the argument that all the normal mechanisms are deterministic.  Libertarians need there to be some other type of mechanism in play to avoid the counter that the mechanism they are trying to use is either deterministic — and so can’t justify libertarian free will by their own definition — or probabilistic and so ultimately by chance, which doesn’t work for libertarian free will either.  However, I think that libertarians can make a dent by insisting that all they want is an intentional mechanism, one that makes selections on the basis of ultimate meaning and not just on symbolic processing.  Why this works, I think, is it properly ties the debate over libertarian free will to the broader philosophical debate over how humans and other things can reason based on the semantics of statements as opposed to simply “reasoning” on the basis of their syntax, which ties it directly into pretty much everything humans do and everything that makes them different from other things.  Yes, the very same people denying the need for semantics in making decisions will deny we need it for everything else, but at this point libertarians can point out that they aren’t trying to invent something just for free will, but something that covers off and solves all of those specific problems.  If we can generate real understanding with a mechanism, then that mechanism can also have causal powers and so can also be the mechanism that makes our decisions for us.  So just as Kane moves towards centering the free will debate around ultimate responsibility which gets him alternate possibilities for free, I think libertarians can move the discussion towards having to be able to make decisions on the basis of the real meanings of our reasons and beliefs and desires and get free will from that (mostly) for free.

This would also fit in with the usual criticism I make of determinism — whether compatibilist or hard determinist — that by all of those mechanisms the actual reasons we think we have for making the decision may not be the actual ones that have the causal impact.  The inner perceptions that the brain gives us for our deliberations may indeed come apart from what the underlying brain mechanisms actually “represent”.  So hard determinists and compatibilists need to show how their mechanisms can guarantee that the reasons we work through while reasoning are indeed causally effective and are doing the work.  They need some kind of reason-based mechanism to make it all work … and if they invent one, then it’s actually them who risk inventing a mechanism that supports libertarian free will better than their own positions, since all libertarians want is for us to have unique mechanisms that are reason-based rather than the alternatives.

So while I think Kane’s way of framing the debate is a good one, I think his solution works better for his opponents than for him.  I intend to work through the other four views and will comment on them in the future.

Do we not have free will?

October 26, 2020

I came across this from a post by Jerry Coyne (it was a few weeks ago so I’m not going to track it down again), as it is fits neatly in his own preconceptions and arguments over whether or not we have free will.   It’s a video by Sabine Hossenfelder, who is a theoretical physicist who nevertheless felt the desire to weigh in on the free will debate.  She takes aim at some philosophical positions, too, but seems rather blissfully unaware of what they actually entail while being totally dismissive of them.  More on that later, so let me examine this scientific examination of free will.

She starts by outlining why she thinks we don’t have free will:

Last week, I explained what differential equations are, and that all laws of nature which we currently know work with those differential equations. These laws have the common property that if you have an initial condition at one moment in time, for example the exact details of the particles in your brain and all your brain’s inputs, then you can calculate what happens at any other moment in time from those initial conditions. This means in a nutshell that the whole story of the universe in every single detail was determined already at the big bang. We are just watching it play out.

These deterministic laws of nature apply to you and your brain because you are made of particles, and what happens with you is a consequence of what happens with those particles.

There are a couple of problems with this.  The first one relates to a criticism that Edward Feser makes of at least some scientific positions in “Scholastic Metaphysics:  A Contemporary Introduction”, where he points out that science reduces all of its phenomena to things like differential equations, and then insists that those things describe all of reality without remainder.  But how does science know this?  If all it ever looks for and incorporates into itself are these things, and it leaves out anything that doesn’t conform, how does it know that it is really capturing all of reality.  Feser uses the analogy of the drunk looking for his keys under the street light rather than where he dropped them because the light is better there and, in the case of science, that he’s had such great luck finding things under street lights so surely he will find his keys there as well.  For Hossenfelder, things are actually even more serious, because the move she actually makes is to argue that we can represent the natural laws with differential equations which is controversial in and of itself and then attempts to argue that a feature of differential equations is reflective of reality:  that given an initial term you can always calculate what will happen at a future point of time.

The first problem with doing this is that we don’t know that everything in reality can be described with differential equations of that sort.  While it has roughly worked that way, some of those equations have been a bit messier and for the most part we’ve worked them out as models, not as definitions or identities.  Most importantly here, we don’t know that we can describe conscious decisions using them because, right now, we don’t have equations describing how those work, and especially not ones where we can plunk in an initial value and know what will happen at any future point in time.  I’m not saying that we can’t do that, but it is entirely possible that conscious decisions are something that really cannot be described in that way and using differential equations.  So she’d be jumping the gun to use that to argue that we can’t have free will.

The worse problem, though, is that as noted above she’s not actually making an empirical observation or talking about the observed behaviour of, well, pretty much anything in existence, even the ones for which it works.  No, what she’s doing is taking a property of differential equations, the things we are using to describe reality, and mapping it directly onto reality itself.  So it isn’t just an argument that those equations describe reality, but instead that they define reality.  Reality, by her argument, really is differential equations, so much so that the properties of differential equations can be considered to be properties of reality without any argument, experimentation or observation required.  Her argument for determinism is essentially that we describe reality with differential equations and differential equations by their nature would be deterministic, therefore reality is deterministic.  Even Coyne wouldn’t make such a blatantly scientistic argument, defining reality precisely by what science is currently using to model it.

The leads into the second problem:  quantum physics.  She does attempt to address an appeal to quantum mechanics later:

What about quantum mechanics? In quantum mechanics some events are truly random and cannot be predicted. Does this mean that quantum mechanics is where you can find free will? Sorry, but no, this makes no sense. These random events in quantum mechanics are not influenced by you, regardless of exactly what you mean by “you”, because they are not influenced by anything. That’s the whole point of saying they are fundamentally random. Nothing determines their outcome. There is no “will” in this. Not yours and not anybody else’s.

It is a fair reply to say that quantum mechanics, being random, doesn’t allow for the sort of free will that we want, which is “free” and yet influenced by external factors and so “sensible” given conditions.  Most Hard Determinists can use this to escape most challenges from those who use quantum mechanics to argue against determinism.  But not me, because the argument I make is that most Hard Determinists rely more or less on an argument that science has proven that all things are determined, and therefore our conscious decisions have to be as well, but even they have to admit that quantum mechanics is not determined, and so those arguments fail.  Not only has science not shown that everything is determined, it has in fact proven that statement false.  Science knows that lots of things aren’t determined at all, and so any argument based on the universality of determinism is doomed by science itself.

Now, most of the time most Hard Determinists can escape this by limiting their scope, and so for example by saying that determinism applies at the macro level and conscious decisions and brain operations are at the macro level.  But to make this argument actually even more devastating for Hossenfelder, she can’t actually go there.  First, because she repeatedly makes the mistake of talking about particles instead of things like neurons, and particles exist at the quantum level, and so she’d have to be talking about them.  But if we are charitable and grant that her phrasing there might be a little loose, she still has the problem that she’s argued that all things are describable by differential equations which has the consequence that they are deterministic, and since quantum mechanics are not deterministic then either they can’t be described by differential equations or else the initial condition property doesn’t hold.  Since her only argument is this universality based on differential equations, this would be an issue for her, and attempts to escape it by appealing to the macro level fail because we don’t have those equations for conscious decisions yet.  So science has proven that some things cannot be described by differential equations in the way she insists produces determinism, and she has no way to get to conscious decisions being things that are described that way because we haven’t done it yet.  So not only is science not supporting her case, it’s actually working against it.

So, she turns to discussions of philosophy:

A lot of people seem to think this is a philosophical position. They call it “materialism” or “reductionism” and think that giving it a name that ends on –ism is an excuse to not believe it.

This comment clearly shows that Hossenfelder does not understand philosophy.  At all.

First, materialism isn’t a relevant position here.  Well, maybe eliminative materialism is, but that’s in the same vein as reductionism and there are a host of non-reductionist,  non-eliminativist materialisms out there.  Materialism is basically the idea that all that exists is matter.  It doesn’t take a strong stance on whether all material things are deterministic.  And good thing, too, for the aforementioned quantum mechanics is generally considered to be completely material and yet not deterministic.

Second, labeling things with an “ism” is never considered an excuse to not believe it, at least not in philosophical terms.  Pretty much any consistent position in philosophy is an “ism”.  If anyone is using that as an argument against her, then they would seem to understand philosophy about as little as she does.  I suspect that no one actually is doing that — hence the bit of snark here that’s probably unworthy of me — and that instead they are classifying the views into those positions to point out the known issues with those positions, and her response here is her either not getting that or dodging those criticisms.

Third, materialism and reductionism are, in philosophy, possibly the majority positions.  So anyone assigning her position to those labels as an attempt to get them dismissed out of hand should not be replied to with a comment that the move is invalid implying that we should not call her views such, but should be replied to with the fact that those positions are far from being ones that philosophy dismisses out of hand.

Fourth, her view is, at least, reductionist.  As noted above, she thinks that what is important about reality can be completely captured in differential equations, so much so that she thinks she can use the properties of differential equations themselves to determine what reality is and must be like.  The criticisms of reductionism are all about whether we can captured everything interesting and important about reality in things like that, or in her case it’s probably more accurate to ask whether all phenomena can be reduced without remainder to differential equations of the sort she relies on.  And as noted above, we can’t … and that’s before we start talking about the really complicated things like qualia and therefore consciousness.  So rather than being an insult or an attempt to dismiss her out of hand, it is instead an accurate classification of her position and her position is indeed actually particularly vulnerable to the objections raised against reductionism.  Her response here can only indicate that she has no idea what those positions are, and doesn’t want to be bothered to figure it out, and so wants to dismiss then out of hand with an almost certainly strawman reply of them wanting to use them to dismiss her views out of hand.

Well, of course you can insist to just not believe reductionism is correct. But this is denying scientific evidence.

The scientific evidence does not demonstrate reductionism, as anyone who actually knew the positions would know.  And just to demonstrate further that she doesn’t understand them, her justification for that statement is this:

We do not guess, we know that brains are made of particles. And we do not guess, we know, that we can derive from the laws for the constituents what the whole object does.

The first statement is irrelevant:  you can believe that brains are made of particles but that not all things that are important about the brain and what it does — meaning consciousness here — is reducible to physics (see emergentism, another evil “ism”, for example).  And what the challenges to reductionism are all about is questioning whether you can do that for every relevant phenomena, and there are good reasons to think that even with the more common ones — from, say, biology to physics — that it can’t be done.   Writing laws to reduce all animal behaviour, for example, to physics tends to leave us with gaps and rather useless physical laws.  So she first doesn’t want her view to be reduced to an “ism”, then insists that the “ism” is proven by science, but then never addresses the actual philosophical objections to her position, many of which use science.  And her specific view isn’t actually scientifically valid.  This makes this statement unintentionally hilarious:

If you make a claim to the contrary, you are contradicting well-established science. I can’t prevent you from denying scientific evidence, but I can tell you that this way you will never understand how the universe really works.

I feel confident in saying that you will never understand how the universe really works by reducing it to differential equations and then using the mathematical properties of differential equations to dictate to the universe how it really works.  You’re going to have to at least do some observing to figure that out.  And, remember, this is coming from a philosopher.  If the scientist is ignoring observations and the philosopher is saying that maybe we should do some, something has gone seriously wrong.

She then takes a quick stab at a philosophical argument:

You see, that thing you call “free will” should in some sense allow you to choose what you want. But then it’s either determined by what you want, in which case it’s not free, or it’s not determined, in which case it’s not a will.

Of course, this argument equivocates on “determined”.  As will be important later, the big issue with free will is that we want our conscious decisions to be the result of our conscious decision-making processes.  So wrt wants, we want it to be the case that given the wants I have and how I believe the world works, that my decision will be what rationally follows from them as best as my capacities can do them.  So we want our choices to be “determined” by my wants in that they rationally follow from them, and if we make a poor decision we can trace back the error we made — either in considering wants or considering the world — and note where our decision-making process failed.  So it’s independent of neither our wants nor our decision-making abilities.  But the operations of those processes are what determines it.  It has to be the case that the decision is not fixed until those processes complete.  Determinism in Hossenfelder’s sense breaks this by insisting that the actual outcomes are determined well in advance, at the Big Bang in fact.  What this means is that the details of that decision-making process and our wants might not matter.  We could very well have that deterministic process produce different conscious ruminations than what is “actually” used under the hood.  If that’s possible, then how do we know that we make decisions at all?  In fact, it’s more reasonable to say that we never make decisions at all.

Yes, our wants need to be in some sense free as well, so there are issues.  But it’s hardly the case that free will simply never made sense, as she asserts.  It’s only her equivocating her notion of determined with how wants would determine decisions that allow her to even make an argument that at first glance might appear to be one that we might need to consider for more than a second.

In summary, the idea that we have a free will which gives us the possibility to select among different futures is both incompatible with the laws of nature and logically incoherent. I should add here that it’s not like I am saying something new. Look at the writing of any philosopher who understand physics, and they will acknowledge this.

I suspect that she defines “any philosopher who understands physics” as those who would agree with her.  Yes, she isn’t saying something new.  Philosophers have raised issues with free will and its relation to physics for centuries.  But for centuries other philosophers — yes, even some who understand physics — have raised issues with those issues and with Hard Determinism, and she seems blissfully unaware of any of them.

But some philosophers insist they want to have something they can call free will, and have therefore tried to redefine it. For example, you may speak of free will if no one was in practice able to predict what you would do. This is certainly presently the case, that most human behavior is unpredictable, though I can predict that some people who didn’t actually watch this video will leave a comment saying they had no other choice than leaving their comment and think they are terribly original.

Okay, she has a point that the last comment isn’t original.  However, one of the issues with that is that there is a potential contradiction here, with Hard Determinists in various ways exhorting people to change their views or positions and take responsibility for their views in a way that isn’t compatible with Hard Determinism.  By her own views, Hossenfelder can talk and talk and talk at me about there being no free will but if the Big Bang did not deign to permit me to be convinced by her words then I won’t be, so I can’t bear any responsibility for my views.  And any attempts to move the relevant processes into the decision-making processes themselves and so retain that sort of responsibility lead one to some form of compatiblism, which is the view she is denigrating with the comment on redefining the term (they see it as coming to a proper understanding of it, not redefining it).  There’s another issue where many Hard Determinists talk as if they make choices and bear responsibility — usually positive — for their ideas and the like, when they are no more free than anyone else.

This leads to the underlying issue around those responses to Hard Determinists:  their views, if treated consistently and taken to their logical conclusion, so thoroughly contradict our actual experiences that it is almost impossible for us for actually talk and act as if Hard Determinism is true.  That it’s so foreign to our experience is prima facie reason to at least be skeptical of the position, and most Hard Determinists only have “Science says all things are determined” as an argument, which as we’ve seen is quite inadequate.

Others have tried to argue that free will means some of your decisions are dominated by processes internal to your brain and not by external influences. But of course your decision was still determined or random, regardless of whether it was dominated by internal or external influences. I find it silly to speak of “free will” in these cases.

That she finds it silly doesn’t make it, well, actually silly or not useful.  As noted above, we want our actions to be determined by our internal processes.  Given that, if they were able to show that the internal processes mattered more then that would be important and would flatly contradict her own stated position at the beginning of the post.  She might be able to quibble over whether we should call it “free will” but her view would still be wrong.  That’s hardly inconsequential.

What is really going on if you are making a decision is that your brain is running a calculation, and while it is doing that, you do not know what the outcome of the calculation will be. Because if you did, you wouldn’t have to do the calculation. So, the impression of free will comes from our self-awareness, that we think about what to do, combined with our inability to predict the result of that thinking before we’re done.

If the outcome of the calculation is already determined before it starts, then what is the calculation itself specifically doing?  Is it even doing the calculation that it purports to be doing?  The issue is one that we see with neural nets.  The actual calculation, at the hardware level, isn’t content-aware.  You can use that same neural net for a different purpose and given its input it will spit out an answer.  I remarked once that you could use a neural net trained to solve differential equations to play chess and it would work, which is probably a bit facetious, but perhaps what will drive the point home is that I could train a neural net to solve differential equations and then use it to play chess and it is possible that it would be better at playing chess than it would be at solving differential equations.  At the level of the hardware, content does not matter.

Which leads to that self-awareness.  See, the self-awareness isn’t us looking at the brain doing the calculation and thinking that it isn’t done yet so I don’t know what the answer will be.  It is us working through the calculation consciously.  When I am revamping my schedule while going for my daily constitutional, it’s not that my brain is calculating what will be best on which day and just not telling me what the answer is but that my conscious reasoning is walking through the options and ultimately deciding what works best on each day.  If we cut those processes out and place the ultimate responsibility on the Big Bang then all of that might be epiphenomenal.  Those considerations need not be what the brain itself is actually considering.  But if they aren’t, then are my decisions rational?  Even if they come to the conclusion that a rational analysis hints should be the decision?  What is the point of conscious deliberation if it doesn’t do anything?  But what could it possibly do under Hossenfelder’s view?

Suppose you have a computer that evaluates whether an equation has a real-valued root. The answer is yes or no. You can predict the answer. But now you can change the algorithm so that if you input the correct answer, the code will output the exact opposite answer, ie “yes” if you predicted “no” and “no” if you predicted “yes”. As a consequence, your prediction will never be correct. Clearly, this has nothing to do with free will but with the fact that the system you make a prediction for gets input which the prediction didn’t account for. There’s nothing interesting going on in this argument.

This highlights the issue here, as she proposes messing with the decision-making processes to produce wrong answers to make it unpredictable, and then says that that wouldn’t be free will.  All of those who advocate free will — even those who make the predictability argument — accept this.  What they argue is that we can’t predict the answer until the decision-making processes do their job because the answer is determined by those processes.  Hossenfelder thinks that’s not the case, so there’s still a difference in argument there that she doesn’t acknowledge.

Another objection that I’ve heard is that I should not say free will does not exist because that would erode people’s moral behavior. The concern is, you see, that if people knew free will does not exist, then they would think it doesn’t matter what they do. This is of course nonsense. If you act in ways that harm other people, then these other people will take steps to prevent that from happening again. This has nothing to do with free will. We are all just running software that is trying to optimize our well-being. If you caused harm, you are responsible, not because you had “free will” but because you embody the problem and locking you up will solve it.

Well, a number of Hard Determinsts, like Jerry Coyne, argue that understanding that we have free will should change how we treat people since they aren’t responsible for their actions.  So we should treat them more like we treat people who right now we consider are clearly not responsible for their actions, by trying to cure them and not considering them morally responsible for their actions.  So surely saying that people aren’t morally responsible for their actions and having us treat them as such will erode anything that depended solely on morality, as it would eliminate morality as such.  I also find it interesting that she rather blythly talks about locking people up on the basis of maximizing well-being, since morality tends to preclude us doing that.  Essentially, she rejects the idea of Coyne that we should cure or rehabilitate offenders since they have no choice and only solve the problem by locking them up if there is no way to do that and accepts the idea of locking people up because they are “bad”, in that they challenge her well-being.

Why am I telling you this? Because I think that people who do not understand that free will is an illusion underestimate how much their decisions are influenced by the information they are exposed to. After watching this video, I hope, some of you will realize that to make the best of your thinking apparatus, you need to understand how it works, and pay more attention to cognitive biases and logical fallacies.

Even Libertarians about free will believe that our decisions are influenced by the information we have, and exhort us to pay attention to errors in our decision-making processes.  If her Hard Determinism is true, though, our decision-making processes will just do what they do and there is no reason for us to attempt to correct them, especially since we will likely be unable to.  And so, at the end, we see Hossenfelder creep back towards compatibilism while trying to convince us that only Hard Determinism is the right answer, as so often happens in these discussions.

Coyne On Sapolsky on Free Will

June 1, 2020

Jerry Coyne is deeply involved in discussions of free will. He is, in fact, a hard determinist, and as seems usual for him as time has gone on he seems to have ended up disliking compatibilists even more than libertarians. Or, at least, he seems to gripe more about them than he does libertarians, as we can see in this post where he talks about a take on free will by Robert Sapolsky that aligns with Coyne’s view:

(He doesn’t address compatibilism in his entire interview, which I suspect reflects his view—and mine—that compatibilism is just a semantic game that distracts from the real issue: the hegemony of determinism.)

The thing is, though, Coyne doesn’t actually understand the compatibilist view, as has been demonstrated on many occasions. He has accused them of not accepting determinism — in the sense that all events, including intentional actions, are determined by impersonal causes — when in fact compatibilists have to accept that by definition. He’s accused them of playing semantics with notions of free will and choice despite them generally believing that removing all notions of choice cause greater issues. He’s jumped on a comment from Dan Dennett saying that if people accepted Coyne’s notion of free will it would leave to them acting in bad ways to insist that that’s all Dennett cares about, despite there being much evidence that Dennett is also concerned with creating a notion of choice because, well, he thinks that there still is one.

The point of compatibilism is this: Compatiblists accept that everything is determined, but still think that if we throw out all notion of choice then we will run into many problems, from it changing our behaviour in improper ways to it causing us to be unable to explain many common behaviours that we can readily observe. If we try to introduce a different term to cover these things, it will act so much like the old concept of choice that it will hardly seem to be worth the bother, and will reintroduce a number of the issues that hard determinists are having. They also end up convinced that they can do that, and introduce a notion of choice that is compatible with determinism and yet is still meaningful and captures what we seem to need the concept of choice for. It’s not a semantic game.

What Coyne has never really grasped are all of the cases where his own view would either need a notion of choice or imports it in order to seem reasonable at all. His big concern is prison reform, and he chides compatibilists for ignoring that:

One can see Sapolsky’s whole interview as the logical consequence of his determinism, which leads to an immediate consideration of how screwed up our penal system is. You don’t see many compatibilists, even though they’re determinists, worrying much about penal reform. But such reform is far more important and consequential than trying to redefine “free will” so people won’t get freaked out if they don’t think they have it.

Yes, I know some readers say that you can still favor penal reform if you’re a compatibilist, and that’s true. But then why do you see hard determinists like me, Sapolsky, and others being the determinists most concerned with penal reform, compared to compatibilist/determinist philosophers, who argue semantics ad infinitum and claim, falsely that their efforts aren’t directed toward keeping the Little People convinced that they have free will?

It is, of course, absolutely inconceivable why a group of philosophers in a philosophical discussion would focus far more on the overall philosophical issues rather than one specific potential consequence of that discussion. Especially since many of them might indeed feel that penal reform actually follows from and is justified by far better arguments than determinism, and so they are separate issues. After all, even for Coyne one of the strongest reasons to ditch a retributive approach is that it simply doesn’t work. Even if we all accepted Coyne’s version of hard determinism, if taking retribution actually reduced criminal behaviour then Coyne, to be consistent with his other views (like that we can use punishment for deterrence, which he mentions earlier in the post) then under this view determinism would justify retribution. Those compatibilists that Coyne derides might, thus, feel it better to go after retribution not from the argument that criminals cannot be responsible — or morally responsible, as Coyne likes to split the proverbial hair — for their actions and so retribution is not justified, but on the grounds that retribution itself is morally wrong, even if they can be held responsible. They can easily say that the only proper purposes for the penal system are rehabilitation and the protection of society, and while punishment might fit into that if it does work as deterrence, retribution clearly doesn’t. And many of them argue just that.

The argument that Coyne consistently misses is that many of the concepts that we would need to do rehabilitation and determine how to rehabilitate people rely on consider whether someone made a choice or made a decision to take the action or not. We can see this when we expand on his own notion of responsibility:

I would disagree in the sense that someone who commits a crime is the responsible person—the person who has to be reformed, treated, or “punished”. All that means is “this is the person who did the act.”

Coyne tends to argue, to align with hard determinism, that criminals have as much choice in the matter as things like rocks. So to say “this is the person who did the act” clearly can’t mean “this is the person who chose to do the act”. So we can argue that a rock that hits a window and breaks it is the thing that did the act of breaking the window. And so, from that, we’d want to figure out how to prevent that from happening again. But there’s a huge difference between the window breaking because a windstorm picked up a loose rock and pushed it through the window versus an angry neighbour picking it up and throwing it through the window. While both may be prevented by ensuring that there aren’t any loose rocks in the backyard, in the latter case we would definitely want to prevent that neighbour from choosing to break the window. And in so doing, we’d want to get at the intentions behind that neighbour doing that in the first place. This leads to another key difference, between the case where the neighbour was just trying to clear loose rocks and tossed it in a way where a freak ricochet hit the window and the case where the neighbour deliberately aimed that throw at the window. Which leads to another difference between the accidental ricochet case and a case where the neighbour deliberately aims at the window and misses. We’d still want to address the latter case as the same sort of event as where the neighbour throws it and hits it, but then all we have are the intentions to appeal to … things which aren’t relevant in Coyne’s strongly hard deterministic viewpoint, since intentions are internal, and only act through choices, which are the things that Coyne says we don’t have.

I’ve also traced in the past a chain of there being a difference in how a kleptomaniac should be treated vs someone who steals to survive vs someone who steals because they find it fun. If we are going to rehabilitate them, we need to take their differences into account, or else we will fail. But the only thing that is different between them is the degree of choice that they have and the reason they made the choices they made. The kleptomaniac clearly has no choice, as they will often steal despite the fact that if they had been given the choice in those circumstances but without the kleptomania they wouldn’t have done it. The person who needs to steal to survive has unfortunately circumstances but their choice-determining faculties are properly engaged and working. And the person who steals for fun is making a deliberate but dangerous choice, where it is difficult to see how merely changing the environment will help.

This is why many compatiblists are clear that their notion of choice involves engaging the choice-producing faculties of our brains. This allows them to preserve determinism while still accepting that part of our mental make-up involves things that can be reasonably called choices. Coyne can rail against this view all he wants, but it is difficult for him to refute it while still maintaining the idea of rehabilitation, because working out how to rehabilitate people will involve talking about those choice producing mechanisms and noting when they are working properly and when they aren’t.

I also think it possible that Coyne rails so much and so often against that minor point about people acting as if they have no responsibility if they became convinced that we have no free will is because it’s another point that he can’t actually refute. Coyne needs it to be the case that what we hear and are told can impact our behaviour, in order to push for changed to the penal system and for argument. But then under that Dennett’s point that telling people we have no free will will lead them to act in ways we don’t want them to does count. Coyne can try to argue that we should never tell people things that aren’t true, but it is clear that Dennett and most compatibilists believe that it isn’t actually true that we don’t have free will, but that we have free will in a certain sense, and as per Dennett that would be “the only sense that matters”. So they’d be debating between risking people misinterpreting free will to be the libertarian version vs people misinterpreting having no free will to meaning that there is no notion of choice at all anymore, and it’s perfectly valid, then, for them to claim that since both can be misinterpreted maintain the term “free will” is the least dangerous misinterpretation. And even if that worked Coyne would then have to ensure that he himself never “lies” to people to get them to do the right thing, despite it being entirely consistent with his own view. So he needs to completely dismiss the argument because it is one argument that he, himself, can never safely refute.

Anyway, let me turn my attention to things from Saplosky in the post, noting that Coyne is summarizing them which can actually be misleading:

Most of Sapolsky’s eloquent take on free will relates to the concept of humans as “broken cars.” When your car is broken, you don’t say it deserves to be punished. You either “rehabilitate” it by taking it to a mechanic, or if it’s a car that won’t ever work well again, you “sequester” it by putting it in the junkyard or letting it rust in your front yard. And since Sapolsky, like me, sees criminals as “broken humans” who simply reflect in their criminality the influence of their genes and environments, he sees no sense in putting people away because they made the “wrong choice” or “deserve” retribution. (Some readers here think these two ideas are not part of our penal system, but I would disagree strongly.)

The problem is this: those that we would actually consider to be “broken humans” are those that we have the most sympathy for, and the ones that are less “broken” are the ones that we most want to punish (whether we should or not). Someone who steals because they are a kleptomaniac is quite broken, but we generally want to rehabilitate them or “fix” them because we know that they are really not choosing that at all; their compulsion drives them even as their choice faculties scream at them to stop. The person who steals because they can is arguably not as broken, but is much less sympathetic. It’s hard not to see Saplosky’s view here as treating the kleptomaniac like the junk car and the unrepentant thief as someone to take to the mechanic. He can argue that we can fix the kleptomaniac — at least limit them so that they aren’t a threat — but this still places them in the same category which seems absurd. Sure, Saplosky and Coyne will argue that we really should do that and that them seeming absurd is exactly why we need to drop the notion of free will, but this runs up against the problem that sometimes people really can properly be said to have “made the wrong choice”. If they made a mistake of reasoning that leads to an action that causes harm, all we’d want to do to “fix” them is point out the mistake in reasoning. So, yes, they just made the wrong choice and have to be taught why it’s the wrong choice so that they don’t do it again. How do you do that without having a notion of choice?

Coyne includes a specific quote from Saplosky on this:

“I am of the stance that the entire criminal justice system, top to bottom, makes no sense whatsoever because it is predicated on 200-year-old biology. We have no control, ultimately, over anything we do. When we say ‘I’ve changed my mind’ about doing this or that, we are in fact saying ‘circumstances have changed my mind.’ We have no agency, and the criminal justice system does not make any sense at all.”

The problem is that often those “circumstances” really are “my choice-producing mechanisms have changed that decision”. For example, if someone is intoxicated we know that their choice-producing mechanisms are impaired (alcohol reduces inhibitions). Someone many make a choice while intoxicated but not be able to act on it, but then “change their mind” the next day when sober. But the only circumstances that have changed are that the choice-producing mechanism is no longer impaired. How does Saplosky propose to explain that with his notion of “circumstances have changed my mind”?

Another analogy used by Sapolsky is epilepsy. In the Middle Ages and even later, epileptics were thought to be possessed by demons, and were often punished or burned at the stake. But now we’ve discovered that epilepsy is a disease that one has no control over: a screw-up in the brain’s potassium channels. Now that epilepsy is medicalized, we don’t punish people for being epileptics, but instead try to alleviate their disease. And so should we do with criminals.

Did we ever think that people possessed by demons were acting under their own volition? My immediate thought is that we killed them or punished them because we thought that the original person was gone and that we were doing it to the demons (and don’t that just seem sad and wrong). So hardly the same sort of situation here, as it was predicated on them not being responsible and there not being a cure. In short, in that case wasn’t that us simply sending them to “the junkyard”?

At the end, Sapolsky answers two of Mirsky’s questions. First, does he think that neuroscientists will drive philosophers out of business? That is, will empirical studies of volition make philosophical lucubrations about free will obsolete? Sapolsky says “no”, that we should simply “force dead white male neurobiologists and philosophers to talk to each other more.” I mostly agree, for we need philosophers to clarify the concepts of “will” and “agency.”

They’ve been talking to each other for decades now (I studied neuroscientific theories in my undergrad Philosophy degree and in my Cognitive Science degree, where philosophy is a participating discipline). I think him overly optimistic about the impact his stuff will have on it. Philosophers are already paying attention those those fields; if they have anything interesting to say, philosophers will be interested in it.

Likewise, how we view “rewarding” people will change, he argues. Praising someone for their beautiful cheekbones, he says, is ludicrous: they have no “responsibility” for their cheekbones.

And yet, we do, so it seems that praise isn’t based on our notions of responsibility. As Coyne notes, doing so can impact behaviour, which is probably why we do it in the first place.

Anyway, that’s my discussion of free will for today. Remember, as a staunch libertarian I can indeed decide whether or not to talk about it, but as a hard determinist Coyne simply cannot help himself when it comes to talking about free will.

Is an Evil God as Plausible as a Good God?

October 19, 2016

Jerry Coyne has made a post linking to a video by Stephen Law that argues that the standard free will defense for God allowing us to act in evil ways would also work to justify an evil God. He posits that there might be an evil God who wants us to suffer and that might be the God who created everything. He then notes that the first counter to that would be that we’d imagine that the world would contain a lot more suffering than it actually does if that way the case, at which point he uses the “free will” defense to say that just as we can posit that a good God allows us to choose evil as an unfortunate but necessary consequence of us having free will, the evil God can allow us to choose good as an unfortunate but necessary consequence of us having free will. Thus, the free will defense allows us to justify both an evil and a good God, and so cannot be used to choose between them.

The first issue here is that it ignores the idea that allowing humans to have free will is, in fact, good in and of itself. If it is better for humans to have free will than it is for us to be nothing more than the willing puppets of God with the illusion of actually making free choices, then we can easily see why a good God would allow us to have free will and even why a good God would accept us acting evil out of that free will, but it’s difficult to see why an evil God would give us that great good in the service of willing evil acts. So, if we stick strictly to good and evil here, a good God would give us free will because giving us free will is in and of itself good, but it seems inconsistent for an evil God to give us such a great good. To oppose this, you either need to show that free will is not a good, or that it facilitates in some way some great evil. Both of those would be daunting tasks, to say the least.

However, as I’ve already discussed, talking about “good” and “evil” is almost always too vague. We need to work out what we mean by “good” and “evil” to really work this out. Even Law assumes that his evil God really wants to increase our suffering. So let’s compare, then, a moral God who wants us to act morally to a sadistic God who wants us to suffer. Since being moral requires us to have free will and the free will to choose to act morally — an argument that Coyne, at least, can’t oppose since he thinks that if we don’t have free will then the concept of morality is meaningless — we have a good reason for a moral God to want to give us free will. This doesn’t seem to work for the sadistic God. After all, if the sadistic God really wanted us to suffer, he could just turn us into self-aware puppets with a strong innate moral fiber, with the ability to understand that what we are doing is wrong, but being unable to stop ourselves from doing it. This world is not like that. Additionally, if we are given the free will to stop hurting each other, then we might, in fact, all just go ahead and do that, leaving a world with a lot less suffering. This doesn’t in any way promote the sadistic God’s plans. One might argue that allowing us to act immorally might mean that all of us might choose to always act immorally and so contradict the moral God’s plans just as much, but it is clear that the moral God needs to risk that in order for us to act morally at all. The sadistic God does not need to give us free will in order to make us suffer.

But wait, I can hear some of you cry! Aren’t you pulling a bait and switch here? Why are you comparing a moral God to a sadistic God? Why aren’t you comparing a moral God to an immoral God, a God who wants us all to act immorally all the time? After all, the immoral God requires us to have free will just as much as the moral God does. So, let’s take a look at that case.

In order to assess this, we first need to decide what it means to act morally. Let’s first consider the case where what it means to act morally is to act according to what God says is moral. In this case, then, moral God wants us to act according to the moral rules that he has defined, while immoral God wants us to violate those rules, well, pretty much all of the time. At this point, immoral God seems to be drifting towards “insane God”, setting up rules that he deliberately doesn’t want us to follow. The best we could say about this is that it might be a God that is encouraging us to think for ourselves … but since that might mean that we act according to the rules or not according to the rules as we see fit we no longer have an “immoral” God, but instead a God that allows us to make our own choices, and thus a “free will” God. But a free will God has little reason to make moral rules in the first place. Thus, an immoral God under this model seems rather implausible, as either it has to be advocating against follow its own rules, or ends up as a God that doesn’t really want us to act immorally at all.

Next, let’s look at moral relativism. Here, the moral God would want us to act according to and consistently with our own ideas of morality, while the immoral God would want us to, in fact, act against those assessments. But giving us the ability to assess the situation morally flies in the face of that. Why care about morality at all if we aren’t supposed to follow it? It’s difficult to imagine a reason why an immoral God would, again, want us to have any moral capabilities at all in that case. So we’d have to retreat to God wants us to suffer or God wants us to make choices. Either way, the immoral God seems out of the picture.

Finally, let’s look at there being an objective morality that is independent of what God thinks is moral (ie it’s not just defined by what God says is moral). A moral God would want us to follow it, while an immoral God would want us to violate it. But we have to ask why, in fact, each of them would want us to do that. A moral God can appeal to either the idea that as moral agents — created as such by God — it is our natural duty to try to act morally, or the idea that it is better for us to act morally. The immoral God can’t appeal to the idea that it is our duty to act immorally if we are created as moral agents, and if we are not created as moral agents we are amoral, not immoral. Again, in that case it would be better if they hadn’t created us as moral agents at all, and if immoral God doesn’t create us as moral agents, then it doesn’t need to care about morality — or free will — at all. However, immoral God can appeal to the idea that it is better for us to reject that objective morality and so that’s why we should act immorally, as well as why we should have an idea of what is moral, so that we know what to avoid. But then we can sub back into the original question and note that that’s not an evil God anymore. On that argument, both moral and immoral God are advocating that we do what it is in our own best interest, which is, therefore, a good God. Their disagreement would be over whether being moral is good or whether being moral is bad for us. And most of us, by default, are going to side with being moral being good.

Thus, Law’s argument doesn’t work. Evil or immoral Gods tend to either not care about being moral or end up being good Gods of some sort anyway, and always end up being less plausible than good Gods. This does not mean that we have good Gods and don’t have evil Gods, it just means that this little quirk on the Problem of Evil doesn’t work to establish the point Law is trying to make here.

Bear and Bloom’s Experiment on Free Will … and Why it Fails

May 20, 2016

So, from Jerry Coyne at “Why Evolution is True” comes another experiment that he, and others, seem to think supports the idea of hard determinism, the idea that we don’t really have any kind of free will at all, even the sort of free will that people like Dan Dennett think is worth wanting. Coyne describes the experiment thusly:

The first thing the authors did was expose the subjects (who had been trained) to five randomly-placed circles on a computer screen, asking them to choose one circle quickly. Then, after intervals of time ranging from 50 to 1000 milliseconds (0.05 to 1 second), the computer randomly turned one of the circles red.

The subjects were then asked if their chosen circle turned red. They had three choices: “yes”, “no” and “I didn’t have time to choose before the circle turned red”, all indicated by pressing one of three keys on a keyboard.

Without any “postdictive bias” of the kind I described above, one would expect “yes” to be answered about 20% of the time when subjects reported that they did make a choice, because the circle that turned red was one of five chosen randomly by the computer. Instead, regardless of the interval before the circle turned red, the probabilities that you said “yes, my chosen circle turned red” was always higher than 20%. That’s shown in the graph below, which plots “probability of a yes answer” against the interval after which the circle turned red.

What’s important about this plot is not only that the probability was higher than 20%, which means that people were saying that their “choice” turned red more often than they should, but that that probability was higher when the interval between the start of the experiment and the circle’s turning red was shorter. That is, people’s bias—that they had “chosen” the circle that later turned red—was higher when they had less time to “make” a choice:

Now, the experimenters thought of a potential problem with this — although you should be able to come up with some problems with it beyond that one — and tried to fix it:

The authors thought of one problem with the experiment above. If the subjects were confused about whether they had chosen the circle that turned red, they might simply randomly press the “yes” or “no” button. That would drive the “yes” answers, expected to be 20%, towards 50%, giving the higher-than-expected “yes” rate shown above.

To deal with this, they used an experiment in which they showed TWO randomly positioned, and colored, circles on a screen, with the two colors chosen from an array of six. The told the subjects to choose one color. They then added a third circle between the two that had a color randomly chosen from the two initially displayed. And, as in the five-circle experiment, the third circle appeared at intervals ranging between 0.05 and 1 second. This way a random punch of “yes” and “no”—”I chose the right color” or “I chose the wrong color”, respectively—a randomness due to confusion, would not bias the results. With only two circles, a random punch would just make the probabilities of “yes” and “no” closer to 50%, which is what they should be anyway.

And again, the same bias was shown: subjects generally reported that they chose the circle of the same color as the one that appeared later with a probability of higher than 50%: as high as 63% at short time intervals. And again, the shorter the time interval, the greater bias was seen in the self reports.

Supposedly, this shows something important:

What both of these experiments seem to show is that, as Bear wrote in the Scientific American piece, “Perhaps in the very moments that we experience a choice, our minds are rewriting history, fooling us into thinking that this choice—that was actually completed after its consequences were subconsciously perceived—was a choice that we had made all along.” The paper with Bloom cites earlier experiments that also support this result. We have to face the possibility, just as we now realize that choices can be made by the brain before we become conscious of them,” that choices may actually be carried out before we become conscious of having made them; and yet that we feel that the sequence was the opposite of what really happened.

So, this might show that we make a decision and then trick ourselves into thinking that the decision we made was the one that occurred, no matter what decision we actually made. Put this way … it seems almost nonsensical to have such a convoluted process to trick us into thinking that we made a specific decision based on our own experiences and consciousness only for it to pull the rug out from under us later and rewrite our memory to think that the result of our conscious experience of decision-making was the exact opposite.

Fortunately, there are two other really big issues with the experiment. The first is that the key is the conscious recognition of when a choice was made. It appears that in all cases the experiment relies on the participant being certain that they actually made a choice, and so they can report the cases where they made the decision after the circle had changed colour or appeared so that we can eliminate any bias in the decision-making from this new stimulus. But note that the time ranges are all incredibly small. The longest time interval was one second, so that gives lots of chances for the person to have simply not decided before the stimulus kicks in. Ideally, these would be eliminated, but making this choice isn’t going to be all that binary a process. We’re picking at random, which is mostly subconscious anyway, and so we’re just going to react with a gut “That one!”. When the stimulus and the choice reaction come close together, it’s quite possible that the people were on the cusp of deciding when the stimulus flashed, especially since we do seem to at times be able to react to a stimulus before we are consciously aware of this. Given all of this, the more likely scenario — which they haven’t eliminated — is that the people were generally still making a decision when the stimulus kicked in, and the stimulus impacted their subconscious decision-making processes. To really eliminate this, you’d have to make them hit a button to lock in a choice, and then have the new stimulus appear. But if they did this, it would almost certainly be the case that this effect wouldn’t appear. So at a minimum they need a better experiment, one that can let us ascertain that the choice was clearly and distinctly made before the stimulus appeared, and that the participants aren’t invalidly thinking that they made the choice clearly before the stimulus appeared.

The second big issue is the common one with all of these experiments: as noted above, they are testing decisions that are, in fact, mostly subconscious in the first place. We don’t reason out choosing something at random. But free will is all about the choices we make when we reason out decisions, not instinctive or gut reactions or random choices. So, for libertarians, we’re interested in being able to make choices for legitimate reasons, and these experiments test cases where we are told to choose something for no reason. There’s no reason to think that these experiments say anything interesting about those sorts of cases.

Ultimately, for hard determinists to make their case, they have to start getting into experiments that test the paradigmatic cases of what we think are free choices. These are much harder to test, but these controlled cases simply leave out everything that makes free choices free according to libertarians or even compatibilists, and so add little to the debate.