Well you said “all [metaphysical models that I listed] make exactly the same predictions”. Prediction means something in science, it’s not enough to talk about “regularities”. Like I say, it needs to be explicitly a consequence of the model and it needs to be falsifiable.
Start by not trying to put words into someone’s mouth, especially after they have just told you that they do not hold that opinion.
This is an important concession though, because I am saying that the very concept of free will is nonsense, and here you’re alluding to models that just have free will as a black box. Not only can we not break the box down any further, but also we can’t even discuss how it relates to things like memories (if you disagree, then please let’s talk about that. I would very much like to talk about the decision-making process).
That’s convenient, eh? We invent a concept, provide no evidence for it, then say even a model including it cannot actually make any concrete claims about anything.
Yes, I know, this is exactly the point.
If it is merely a matter of interpretation, then on what basis would we claim that a causal model with local hidden variables has been ruled out? Why only non-local models?
It’s because causal models do make predictions here, and they have been falsified for the case of local hidden variables via bell’s inequality. It’s conceivable that non-local hidden variables could be ruled out someday too.
You’re using a completely nonsensical notion of free will again. And it is messing you up.
Free will can be easily analyzed, because (in common parlance) it’s not stupid fantasy gibberish nonsense. Free will is what is happening when an agent makes choices itself. It’s as simple as that. The term is in common use and people using the term don’t get promptly sucked down a puzzling rabbit hole.
Fun project: Figure out what the term “agent” means. It’s a little tricky, because people are reluctant to call fleas and Furbys ‘agents’ because it makes humans seem less special.
Causation, sensibly enough, is when later events are the result of prior state. Causation is reliant on physics, because for A to cause B (or to cause anything) things have to be interacting, and interacting requires physics.
Fun project: Explain why physics happens. Win awards.
Randomness, again, is not puzzling. It is when events happen which are not fully determined by causation, and more specifically, it is the degree to which the events are not fully determined by causation. It’s worth noting that randomness is not guaranteed to propagate - a process that includes randomity could have outcomes that are completely determined by non-random parts of the process. This is why computers are able to work - the electricity that runs through them includes perturbations, but the hardware corrects for it to keep things running as expected.
Fun project: Prove whether or not random elements are corrected for in human cognition, or whether randomness influences the outcome. Show your work.
It’s worth noting that between causation and randomness, there is no third category, pretty much by definition. Attempting to wedge ‘free will’ in between them doesn’t work, not even if you lubricate it with an air of mystery or whatever. In any case free will is not a ‘basic, primitive, given, a priori, what have you’ - it’s a category of process. (Or a term that can be used to describe certain specific types of processes; whichever.) It’s not a thing, and you aren’t going to find free will particles sitting next to the higgs boson or whatever. At least, not when you’re talking about “free will” as the term is used in the english language. Which honestly doesn’t seem to have much relation to whatever you’re talking about at all.
Good post @begbert2 , and I want to be clear that I was not disagreeing with you when I said this:
I’m of course alluding to the “spooky” free will here, as found in philosophical discussion, that is somehow neither determined from past states, nor randomness (nor merely a mix of these two things).
The kind of free will that supposedly solves the problem of evil, and many other problems in theology, just as long as you don’t ask how.
I agree with you that free will in the simpler sense of making choices of course exists. Offer me coffee or tea and I get to think about it, and pick the one I want. And no-one could absolutely predict what I would choose without simulating my brain and my thoughts exactly.
Oh, I dunno if they have to go to all that much work to predict you. Depending on what they know about your preferences (and whether you have a preference), they might be able to predict you with absolute certainty merely by ruling out the possibility that outside interference is controlling you.
That’s one of the things that has always confused me about free will discussions - there’s this large contingent that seems hell-bent on asserting that they cannot possibly be predicted about anything. I, on the other hand, take a certain amount of pride in the fact that I’m neither an insane person nor a personalityless lump. I’m a rational person with preferences, sometimes clear preferences. So if you were to offer me the choice between eating a tasty-looking chocolate chip cookie and a ghost pepper, pretty much all you have to do is check to see if somebody’s got a gun to my head before making your prediction. And you can be 100% certain you’re right.
I mean, I get that back when free will discussions were first building up steam the big question was whether your will was free or whether you were a puppet of god or the fates. In that context, being unable to deviate from the plan meant being literally controlled by external forces, so of course free will included the ability to do ‘anything’ (aka ‘something other than God’s will’). But just because that was the crux of debate when gods and fates were in the mix doesn’t mean that deviating from the plan is still the be-all and end-all when you’re making your own plans. Seriously, why is that distinction so hard for people to understand?
I think you know what I meant.
Yes of course I can predict with almost certainty that a random human will choose a backrub over having their ribs ripped out (with the “almost” being as close to certain as anything ever is).
What I am getting at, is that people who cling to the idea of “spooky free will” often take issue with their decisions being determined, because they often seem to think that that makes their thoughts irrelevant. That my choice of coffee was preordained, and I was kidding myself that I freely chose it.
But if your decisions are “on rails”, as it were, then your thoughts, conscious and subconscious, are part of the rails, not the car. In fact, they are the most important part of the rails, most crucial in determining where the car goes. It’s just that, things like “pain is bad” are part of the initial set up.
I was fairly sure, but this is Great Debates. We not only have licence to nitpick, but it’s practically a calling!
But yeah, I’m pretty sure we’re basically in complete agreement - the whole problem with the free will discussion are the people who for some reason think that determination = outside control. When in actual fact the mere existence of personality and preference literally means that you no longer are completely random and unpredictable. Thoughts and preferences are why we act like we do, and if a person is arguing that they contradict free will then they are arguing that things with free will aren’t allowed to have wills at all.
What makes the prediction is the assumption that there are stable regularities. As the models I listed all lead to these stable regularities, they all make the same predictions as one with causation does. But that just shows that in neither case, it’s actually the metaphysical superstructure that makes these predictions, but simply the presence of stable regularities.
To claim that a certain prediction confirms some assumption, it must be the case that the assumption is relevant to arrive at the prediction. So, for an example, it used to be claimed that the Casimir effect confirms the reality of vacuum fluctuations—since if you assume there are vacuum fluctuations, you can derive the effect that conducting plates placed close to each other in vacuum will experience an attractive force. However, lots of people are, for (good) theoretical reasons, suspicious of the notion of vacuum fluctuations. But that, on its own, is of course not enough to attack the idea: if the experiment doesn’t have any other explanation, then it simply shows that there are vacuum fluctuations, theoretical scruples be damned.
However, it has since been claimed that the Casimir effect can be explained without relying on vacuum fluctuations. Consequently, if that explanation is right, the Casimir effect doesn’t provide evidence for vacuum fluctuations after all, as it can exist without them.
It’s the same with your claims regarding causality. The predictions it makes, by virtue of leading to stable regularities, can be explained without reference to it; hence, the fact that we do see stable regularities in the world doesn’t constitute evidence for causality, as they are explicable without it.
In other words, you’re not going to tell me what your actual argument is, you’re not going to tell me what’s wrong with my best-effort reconstruction, but you are going to complain that I didn’t somehow intuit what it actually is you’re saying. I gotta say, that makes honest debate kinda hard.
Again, that’s the same boat we’re in regarding causality; it’s just that you, for reasons you’re not telling me, claim that we have some further reason to believe in causality. If you were to hold that it’s necessary for making predictions, I could understand your stance; but since you explicitly disavow this notion, it seems you believe in causation just cuz.
What Bell’s theorem shows is that there can be no joint probability distribution for a certain set of experimental outcomes. This can only be the case if 1. there are no predefined outcomes, or 2. the outcomes influence one another. Hence, we can exclude the conjunction of both.
What we can’t exclude is plain, old determinism. Indeed, it’s easy to see that we never can: for each way the world might be, we could easily simulate it on a computer; but if we can simulate it on a computer, we can simulate it deterministically; but if we can simulate it deterministically, there exists a deterministic system giving rise to exactly the same predictions—IOW, causality isn’t falsifiable.
You cling to unanalyzed metaphysical baggage, and it’s dragging you down.
You say that this is sensible, but how does any event actually make another event happen? This is a question we usually don’t ask, because we implicitly accept causation as primitive: A causes B just cuz. But there is no account whatever about how A causes B; it’s a wholly mysterious notion.
And yet, if there is more than one possible option, how is it that only one gets realized? What determines which one? And again, the only answer is—that’s just how it is. There is no account of how one thing happens, as opposed to the other. It’s a wholly mysterious notion. We’ve just grown used to accepting it.
That wasn’t always the case, by the way. Leibniz would balk at the idea of randomness. To him, with the ‘principle of sufficient reason’, anything whatsoever can only happen if it has sufficient reason to happen. I mean, that’s just sensible: how could anything happen for no reason?
These days, we’ve accepted that things, probably, do happen without reason. And that’s fine: it’s a metaphysical stance which one might well accept. But it’s a black box: there’s no account of how something happens without reason, anymore than there’s an account of how something causes something else to happen.
See, that’s where you’re wrong. I have given above explicit examples that are neither causal nor random, such as the block universe. Now, I get what you (probably) want to argue (although I seem to have little luck lately trying to find a sensible restatement of other people’s arguments): everything either happens for a reason, or it don’t. If it happens for a reason, it’s not ‘free’ in the libertarian sense, since it’s happened for just that reason—there was no choice. It it’s random, it’s not any more free, since it just happened, without any ‘will’ behind it. So, either no free, or no will, makes no free will.
Except of course as such, that argument doesn’t work. Everything does happen for a reason, but that’s not to say that reason couldn’t be an agent’s will. Ah, but then, of course, why is the will as it is? Is it thus for a reason? Then, that reason is why the agent did as they did, in fact, do. Is it determined randomly? Then, whatever decision was made, could’ve been just as well made randomly. (This is the ‘regress’ argument mentioned above.)
But of course, that leaves out one possibility: the reason the will is as it is could be the will itself. Sure, this is circular, but many causal models include the same circularity—a universe that caused itself, for example. Then, there’s nothing beyond the will that determines it—it’s free, in that sense. And, there’s no randomness—the agent acts as they do because their will is as it is.
No, we don’t know how that could work—just as we don’t know how things could ‘just happen’, or how A could cause B. The same black box will always crop up in our descriptions. You can either own up to that, or pretend that causality and randomness are somehow unproblematic, while you take those hewing towards free will to task for not providing a model of how it could work.
So really, the only reply that needs to be made is, in the end, the following: can you explain causation? Can you explain randomness? Because if you can’t (and you can’t), you can’t use the same inability with respect to free will as a point against it.
There’s nothing metaphysical about my position at all. Literally - free will, as I understand/define it, is based entirely in completely materialistic processes.
Did you even read my post? Physics makes events happen. Physics is the mechanism by which causation works. There is nothing mysterious about the fact that physics causes causation.
Now, you could say that physics are mysterious, in the sense that we don’t know why physics happen the way they do. Why is gravity? “Fucking magnets, how do they work?” Now that is a legitimate metaphysical question. However the uncertainty about why physics happens does not translate into uncertainty about what it’s doing (which of course has been thoroughly studied), and thus none of that uncertainty translates into uncertainty about how causality works.
This (unlike free will or causality) is another metaphysical question, and there are several proposed models - most visibly the many worlds theory floats that it’s not the case that “only one gets realized” at all. Alternatively one could posit that randomity “just happens”, and the fact that randomity happens is just a property of the universe, the same way physics appears to be. Or one could posit that there’s actually no such thing as randomity at all, anywhere, and anything we believe is random is actually deterministic but based on factors we are unaware of or do not understand.
The thing is, though, that (just like physics), the question of why it’s happening doesn’t really impact what it does. And what it does, explicitly, is exclude the middle that that libertarian free will bullshit proposes to hide in. If something isn’t caused, it’s random. Randomity is defined as when something happens without cause, whenever it happens without cause. There is no crack in the middle to slip bullshit into.
What you’ve actually given are examples where causality is routed through a ‘black box’ by one method or another. This doesn’t mean that causality isn’t happening; it just means that we don’t understand what causality is happening.
To pick an arbitrary example - consider the “reality is a role playing game with a DM” model, which I think describes about half the models you presented, including the block universe. In this model the DM declares “rocks are falling”, and then declares “everybody dies”. To the characters dying, it appears that the rocks falling cause their deaths - but they’re mistaken - both outcomes are caused by the DM’s autocratic declarations. (Which are in turn driven by other things - but let’s not look back into free will just yet.) So while the squashed adventurers are wrong about what the causes were, they’re not wrong about causation existing - they’re just making a correlation/causation error based on limited information.
Oh, and:
Nothing is “free” in the libertarian sense, because libertarian free will is nonsensical bullshit that sprung from people conflating god with determinism.
It may be worth clarifying that the causation we’re talking about here is, explicitly, metaphysical causation, in that it involves the actions of things not within what we consider the physical universe. This doesn’t mean that it’s mysterious or ineffable though - it’s just outside our view. Black boxes aren’t magical, their workings are just hidden.
Also worth noting is that free will doesn’t require metaphysical causation - and at a conceptual level the process doesn’t really care whether the causation it’s running on is in the physical universe or not. If you are a physical meat machine, or a simulation, or a digital avatar being controlled by somebody plugged into the Matrix, the same cognitive processes are still happening either way.
I miss the point of your criteria. The brain is closer to a copy machine than a numerical computer. The addressing structure is associative. So, the n+1 state of the brain is determined by the n state. It’s the way we memorize poetry:
Input state n= “Twas brillig”
Output state n+1 " and the tlithy toves"
So, the brain is capable of handling deterministic operations. Why would that be exclusive? Why would there be ‘free will’ implications?
If you read “Twas brillig”, can you choose to think of something else or are you mechanically pre-determined to think “and the tlithy toves”?
If you are mechanically pre-determined to think “and the tlithy toves”, and a brain state can be associated to reading “Twas brillig”, and a brain state can be associated with thinking “and the tlithy toves”, then there should (or might) be a deterministic model that relates the two brain states.
If you can choose to think something else in response to reading “Twas brillig”, then no such deterministic model should exist, and any resulting brain state model should appear to be stochastic because your free will is interceding so as to appear to add randomness to the model.
P.s. - Choosing something from Alice in Wonderland/Through the Looking Glass is a good choice I think.
This seems rather oversimplified - doubtlessly on purpose for obvious reasons. However I think that the simplification my be introducing a problem.
I would reasonably expect that a whole host of brain states can be associated with thinking “Twas brillig” - the brain has a lot of real estate and whole bunches of those cells will be concerned with things other than poetry. Similarly there are going to be a bunch of different brain states for finishing the line. And it’s entirely possible that while some (most?) of the “Twas brillig” states will lead to a “and the slithy toves” state, the fact that some of them might instead lead to “-Squirrel!”, “-Oh my god hit the brakes hit the brakes”, or “zzzzz” isn’t an implication that determinism isn’t a thing or that free will is a magical intercession fairy.
Definitely simplified for the purposes of answering Crane’s post. And certainly, your point is 100% valid. That changing circumstances can change what we might think about. So my OP is (very) loosely based on my research, which is dealing more with people in fMRI machines. In other words, are not likely to be suddenly beset by ravenous squirrels or other distractions. The more generalized form is if I get some sequence of brain states from the fMRI machine, and I can deterministically say your next brain states will be X, and it is always right, then can you still be said to have free will?
There are a lot of real world complexities that would make doing this pretty impossible. Even when in an fMRI machine, we know that the brain is thinking about a lot of things in parallel, and in some sense, no two states are ever going to be sufficiently alike to be able to actually model things in such a perfect way (I cannot even begin to imagine how much data might be needed to that… 10s of millions of brain scans? Although given the temporal resolution of fMRI, I’m not even sure any number would be enough). It is a hypothetical question of if such a model does exist would it be compelling one way or the other with respect to free will.
And then, the other side, is a very accurate stochastic model.
I’ll confess that I keep wanting to pick at that “free will” thing - until you clearly define it, you can’t possibly answer your questions about it, and it could even be argued that your questions aren’t even meaningful speech if even you don’t know what you mean by your words.
Add that to the fact that I’ve never heard a definition of free will that isn’t either fully compatible with determinism or composed entirely of desperate handwaving and sticking one’s fingers in one’s ears, I find myself looking at your questions and wondering how you could be asking them - of course there are no free will implications, because free will has no problems with being predictable. I mean, yes, lots of people don’t accept that, but I feel like the discussion should be had about that, because until free will is properly sorted out your questions can’t be asked, and after it’s properly sorted out they’re already answered.
All that said - I’m still confused how come the stochastic models are more accurate. Are they just giving vaguer answers? If they say “there’s a 40% chance they’ll choose coffee and a 60% chance they’ll choose” then I suppose they’re ‘less wrong’ when they’re wrong than something that takes a specific position.
No, I don’t think you’re completely wrong at all. This is a very nebulous topic because we don’t really have a good sense for what free will might be, and certainly not as to whether we have this free will thing or not. I’ve really enjoyed reading the conversation, because yes, to me, free will has always been synonymous with non-determinism. In my mind, I cannot be free to make a choice if something (or someone) else can predict the choices perfectly based on a priori information. Some degree of prediction is sensible. If someone knows I like Chinese food (and I do), and I say “I’m going to order out tonight”, then taking a stab at “I bet you’ll order Chinese food” make sense. But reading this thread has given me some pause with respect to that.
I’m trying to think of it like this. Suppose the brain is a machine with levers (A,B,C,… and so on). And I as the free will agent can pick any lever any like at any given time. If something can predict that I will definitely pick the lever every time, given some set of a priori information, then am I still making a free choice. It feels like that shouldn’t be. That sometimes the predictor should be wrong. But I can see the other side to that. That I might be free to pick any lever I like; however, the choice is predictable.
It is interesting, and I’ve really enjoyed the thread, even if I’ve not participated in it as much as I originally intended.
Doesn’t it strike you as odd that you’re talking about yourself as a “free will agent” that is separate from your own brain?
That’s the situation with most libertarian free will positions; they place themselves literally in opposition with their own physical presence. And it’s not all spiritualists, either. They seek some source of volition that is supposed to be making decisions completely independent from their own knowledge, opinions, thoughts and preferences - because such things are deterministic inputs. When of course all that’s left when you remove all determining factors is complete randomity - which hardly seems like meaningful volition to me.
You might, but it would be difficult. The two states were placed in your brain by your volition. Possibly by intentional repetition. And, once there and exercised repeatedly they become deterministic. However, with effort, you could change it and you would have a new deterministic pair.
See, this is the sort of thing I’m talking about. In the first sentence, you decry all metaphysics, only to then state your commitment to metaphysical materialism in the next one. No wonder you get yourself into such a muddle!
Anyway, I’m actually glad to see the discussion has gotten back on track about @BeepKillBeep’s research, and I think I’ve made my point about as well as I can, so I’ll bow out for the time being. Anybody that calls for an explanation of free will, but is unable to give one for causation and randomness (or whatever other principles they hew to), is just the pot calling the kettle black, even though my efforts at holding up a mirror have largely been met with blank stares.
And no, it doesn’t help to replace one black box with another, such as physics—physics doesn’t explain how causation happens as long as we don’t know how physical laws make stuff do things (not what the precise physical laws are), and we don’t have any knowledge about that, either (in Schmidhuber’s model, we actually get an explanation of why there are physical laws, but that model depends on events being drawn from a probability distribution, and doesn’t give an account of that, either).
So unless anybody can give an account of causality, or of randomness, or of physics, that doesn’t itself rest on another black box, I don’t think there’s any sensible way to continue. Ultimately, it all comes down to this:
You can’t both hold this and demand others give an account of free will. It could just be a property of the universe.
I mean, what would it take for you to consider that (spooky) free will is a nonsense concept? You can’t define it, you can’t give an account of it, you can’t say how it interrelates to known phenomena like memories. Let’s at least consider the possibility that this kind of free will is like the concept of vitalism; something which was plausible enough prior to various scientific discoveries, but ultimately, not only was there no evidence for it, there was no need for it.
Also, there’s a big difference between
The concession that we cannot reduce phenomena forever, and that perhaps a primitive input like the apparent randomness of some quantum phenomena cannot be further broken down
and
Handwaving that we cannot describe at all a phenomenon that should be responsible for complex effects and should be at least associated with neurology which is amenable to study.
The brain is not a closed system, so assuming determinism, you cannot make predictions based on internal state alone. (There is a small amount of time you can theoretically predict, no more than two seconds I would guess, probably less for most actions)
Under no theory can a perfectly detailed model of my brain now predict whether I will be murdered before sunrise. Nor can it predict what words I will be reading five minutes from now.
