The possibility was always open. Universe is big. We are small. Grand surprises are always an open possibility. But this result does us the added benefit of shining a spotlight on one particular place where it might be worthwhile to look.
This is a slight shift in the discussion but you have previously brought up Conway’s Game of Life as having undecidable propositions. And this, of course, is true. It is Turing complete.
But any instantiation of the game is a deterministic system. It’s actually one of the best possible examples I could use, and I was actually debating whether to focus on the Mandelbrot set or Conway’s Game in my long post, and I landed on Mandelbrot because I figured most readers would have seen it before. They can visualize it. But I’m glad that you brought up Conway’s Game, since it’s frequently used to demonstrate how a “world” with some of the most astonishingly primitive “physics” possible can result in unintuitively complex phenomena. The complexity arises from simplicity. It’s worthwhile for anyone who hasn’t seen it to look at what it can do.
There might be universes out there whose core physics is nothing more than Conway’s Game. And those are deterministic universes, despite the fact that there are undecidable questions. If the people living there were to write the core physics of their world (assuming they had sufficient access to the micro-data to do so), they would be writing out simple deterministic rules which run their lives, despite the immense complexity that they see around them, and even despite the fact that Conway’s Game has undecidable questions.
Right. Right.
Okay, I’m going to try to simplify this part of the discussion and make only one request.
If I say X, and you read Y, no problem. I could have been unclear. I’ll just clarify that no, I didn’t mean Y. I meant X. But after I clarified that I actually meant X, and didn’t mean Y… if you once again attribute the position Y to me, I’m going to point out the deliberateness of your having done that even over my explicit denials. And then I’m gonna fuck off outta this place.
There are genuine seekers of truth out there, and they don’t behave like that. Their writing would be more worthy of my attention.
It does more: it also tells us that the possibility exists. It tells us that such physical systems aren’t inconsistent with our universe.
You say that (and what follows it) as if it contradicts my stance; but it doesn’t—rather, it emphasizes what I’ve been going on about: that there are undecidable propositions even within some of the simplest imaginable systems.
Perhaps it’s easier to explain my stance using what’s known as Frankfurt cases. Take, for instance, somebody planning to vote in an election. I, a supremely skilled neuroscientist of questionable morals, want to ensure victory for Party A. To this extent, I devise a procedure by which a monitoring device is introduced into a voter’s brain. When this monitoring device registers an intent to vote for Party B (or any other party besides A), it intercedes, instead stimulating the brain in such a way that a vote for A is cast.
Now, you, a voter having undergone the aforementioned procedure, go to vote. However, you fully intend to vote for Party A anyway; and this, you do. Consequently, the device implanted into your brain never comes active, and hence, you act as if it weren’t even present.
Most people’s intuition would hold that you acted just as freely as if the device weren’t present—after all, counterfactual possibilities don’t seem to exert the right kind of power to have any sway on that; the device never actually does anything, it merely might do something. Nevertheless, your actions are completely predetermined: you could not have done anything else but vote for Party A.
So if that’s right, then we must conclude that for the question of whether we’re free or not, whether we’re determined to do something is kind of a red herring.
However, as such, Frankfurt cases basically are circular: your action as a voter is free if and only if it would have been free without the device implanted into your brain. If that’s itself not a free action, then we’ve gained nothing; and to stipulate that it is is begging the question.
That’s where I believe undecidability can step up and play a role. It gives us options: both universes in which you vote for Party A and in which you vote for Party B are logically consistent; the complete data describing the universe before you actually do the voting does not logically fix that choice. From here on out, we can then continue as above: even though exactly what happens, happens (as it always does), this does not mean that it didn’t happen through your free volition.
Moreover, the only way to have what happens happen is via your choice (or something computationally equivalent to it); hence, that choice gives you responsibility for its outcome.
I’m not attributing a position to you; I’m claiming that you’ve been carrying out this discussion in a particular way, whether you meant to or not. And that way, I’m sorry to say, strikes me as rather condescending: your argument always appears couched in reminders of how our intuition misleads us, of our cognitive biases, and whatnot, in place of an actual engagement with my arguments. Furthermore, you paint yourself, simply by doing so, as the one person acutely aware of these biases.
Thing is, you’re not. These aren’t deep insights into the human psyche; they’re commonplace internet armchair psychology. Yeah, we’re flawed beings, and sometimes we believe things that aren’t true simply because we want to do so; but that doesn’t give you any license to write off opposing points of view as simply due to such biases, because in doing so, you’re exactly demonstrating these biases at work. (As you do, for instance, when despite my repeated insistence that I’m not adding ‘something extra’ to the ordinary physical world, you continue to argue as if that was my position, while simultaneously taking me to task for not taking on your corrections of my reading of your position; whereas I just go and explain myself again, to toot my own horn here for once.)
The simple thing, the only thing, we can do is to present our arguments as best we can, and to engage them on their own merits, not on their possible, alleged or assumed motivation. So quite simply, any admonishment regarding biases and so on has no place in discussion: it violates the principle of charity, and serves to discredit another’s arguments on reasons external to them—itself a fallacious tactic.
Nevertheless, I do admit to reacting overly strongly to this. It’s just that the sheer pompousness of griping about how ‘genuine seekers of truth’ are supposed to behave does rile me up a bit. I mean, come on. We’re random strangers bickering on the internet. Not some sort of noble warriors engaged in a grand metaphysical struggle against the forces of fallacy.
No. I know you know all this, and I know it doesn’t contradict your stance.
But I want the other people reading this thread, all two of them now, to know that the example you pulled is actually yet another example of a ridiculous simple deterministic system that can nevertheless exhibit extremely complex behavior.
You did not mention that in the way I preferred, and it is fully deserving of mention.
Your offered evidence of my inconsistency and bias: I “continue to argue” a point despite your “repeated insistence” that it’s wrong.
The problem is that this is a rank falsehood.
You’re complaining here about a single post of mine. I somehow “continue to argue” something that shows up in exactly one post of mine. One post. You can go back and look if you don’t believe me. I said it once to you. And what have I done since that post?
I have stopped attributing to you the position that you’re adding “something extra”. You said X, I read Y, you demurred and said you really meant X, and then – quite contrary to your own behavior in this thread – I stopped claiming that Y was your position (or implicit attitude or whatever you want to call it). This might shock you, but if you actually look at every post I’ve written since that one single “something more” post (and there have not been many for previously mentioned reasons), you’ll find that I am not in fact attributing that position to you any longer. I have not “continued to argue” this. I have stopped arguing this. Because you asked me to stop.
Go ahead and scroll up if you don’t believe me on that.
You paint yourself as someone with psychic powers, and when you get called on pulling something out of your ass and attributing it falsely to other people, you act like it’s an offense against the majesty of your pure truth-seeking self.
And yes. You are the one who claimed and extolled this particular virtue, not me.
I never claimed that “wanting to know is what drives me”. You did. My original sin in this thread, to which you have reacted so dishonestly, was to say I’m not quite ready to trust your self-evaluation of your pure truth-seeking motives, that I’m not going to take your epistemic virtue at your word but rather based on your behavior, and suddenly your monocle pops out of your eye socket and you start proclaiming to the world that I’m the self-serving, virtue-slinging, biased one. One small comment of doubt about your own motives results in an atomic blast of presumptuous comments back. It’s not just your presumptuous attitude toward other people’s motives that’s astounding. It’s also remarkable how much hypocrisy goes with it.
You have violently debauched the “principle of charity” you now claim to hold in literally every post written to me since post 140, most especially in the very same post that you posited this principle.
If you want to claim that I’m in the gutter? That’s fine. My behavior is not consistent with someone truly driven by a desire to know. I never claimed otherwise. (But you did.) Years ago I used to believe I was, until I actually met other people who were seriously, earnestly driven by the genuine desire to know. I realized I wasn’t really like them. Despite your virtuous paeans to the holy “principle of charity”, and the “drive to know”, you don’t act anything like the way they act.
Instead, you act exactly like me. Or rather, in exactly the manner that you so contemptuously describe me. Every single complaint you have thrown at my own posts describes your own posts as well. Extolling virtue? If I’ve done that, you’ve done that with your “drive to know” and the “principle of charity” to which you can’t hold for more than ten seconds. I don’t actually think your complaints are an accurate summary of my posts, but okay, you think I’m deluded about that. Maybe I am. But if so? You’re even more deluded about your own posting style.
Do you not realize that you called me “biased” in exactly the same post where you said that making admonishments of bias was a violation of the “principle of charity”? You did that. You, not me. I have never, not even a single time, called you biased. All I said was that I didn’t necessarily trust your self-evaluation because of general human biases, but that I could decide one way or another based on your actual behavior. So how do you think your behavior stands up now, when you state the importance of a lofty moral principle a few bare paragraphs after you’ve grossly violated it? I’ve never done that. Even if I’m deluded about my own posts, I haven’t sunk that low.
I’m not making any explicit paeans to high principles. You are. I have a hard enough time editing my posts to remain within forum guidelines to aim for anything loftier.
This is the one courtesy I asked from you in this thread. The only one.
That wasn’t an onerous request. By the “principle of charity” you’re now swinging around like a wiffle bat, you should have been fine with this.
But no. You accuse me of being biased using evidence that isn’t true, and then a few paragraphs down in the very same post you assert that admonitions of bias are a violation of the “principle of charity”. I marvel at the hypocrisy.
So yeah. The last clause of my request stands. This is my last post in this thread, and after one last comment I’m gonna fuck off outta this place.
Come on.
You’re obviously smart. You have interesting and original things to say. Me? I already told you: absolutely nothing I write in a thread like this is original. Even so, I think it’s a fair characterization of me that I am somewhat smarter than the average person. I’m not driven by a desire to know, not at core, but nevertheless, I feel that I’m kinda sorta the type of person you ought to be talking to with these ideas. The kind that might possibly, maybe, perchance, in a different context, could perhaps be receptive to reading about new ideas.
I’m not going to deal with this. Not right now. I made a simple request, and it was rejected, so I’m out today. But I’ll tell you what.
If there’s another “free will” thread in the future (and there will be), and you start participating in it, then if I participate, too… I promise I will scrupulously adhere to your “principle of charity” for the duration of that particular thread. Rather than editing my posts to barely pass muster with forum guidelines, I will scrub my words to an immaculate polish. In return, I ask that you also begin adhering to your own stated principles. If one or both of us is deluded, the time and space will give a breather and some necessary mental distance.
But that’s it for this thread for me. Maybe there will be another.
OK, I think this has gone far enough off topic. I’m not going to defend myself against your accusations in detail; anybody interested enough can just go back through this thread and see for themselves. I’m also not going to try and get in some cheap parting shots.
I am going to say one thing, though. Yes, I did in fact say that it’s wanting to know that drives me; furthermore, I stand by that. I did not, however, say that it’s the only thing that drives me, nor that I’m some sort of ideal rational actor pursuing this. I think the evidence here speaks for itself: I’ve thought about these issues at some length; I’ve revised my position on them (more than once, actually). Why do you think I did that, if not because the whole thing is simply of interest to me? It surely hasn’t helped me acquire fame or fortune.
Moreover, I write long posts with detailed arguments. All I’ve asked is to have these taken on at face value, if they’re taken on at all. Defending myself against accusations of motivated reasoning is part and parcel of that. Charging into a discussion with grandstanding remarks of how something is ‘easy to fool oneself into believing’ is its exact opposite.
I don’t understand why this debate has to get into grandiose territory of good versus evil or weird philosophical minutiae.
I had Cinnamon Life for breakfast this morning. I also have a box of frosted mini-wheats. I enjoy the taste of the two equally, and I assume their nutritional value to be equivalent (I haven’t checked the labels). I live alone. I’m not aware of any constraint placed on my decision between the two. By all appearances at least, I could have just as easily chosen the other way. But having already decided to have cereal for breakfast, I had to choose between the two, so I did.
It seems to me that either I exercised free will in choosing to eat Cinnamon Life, or my brain-computer has an awesome random number generator that I’m not conscious of. If it’s the latter, that would be neither free will nor determinism, in my understanding of the terms. But this “RNG hypothesis” seems unlikely to me, because it’s been shown that in many different contexts, humans don’t have much sense of randomness.
Many of the arguments against free will sound an awful lot like a hidden variable theory, which, in physics, is considered discredited.
I believe that the common version the “RNG hypothesis” is simply that the universe is nondeterministic, in the sense that teeny tiny subatomic reactions and such are presumed to be somewhat random and to impact physical matter slightly now and then - enough to impact the future universe state in a material way. And further, it presumes that this happens a lot, in people’s heads, enough that these reactions impact cognitive calculations on a regular basis. And beyond that, it assumes that some significant percentage of the time the human brain literally has no preference whatever between the options it perceives, such that it abstains from making a rational decision and defers to the ‘RNG’.
That last assumption there is the one I reject. I honestly think that as a reasonably functional calculation machine the brain can’t possibly be ceding control to randomity with any frequency - the randomity would be occurring randomly throughout the matter of the brain, and if the brain didn’t have some way of dealing with that and nulling it out it wouldn’t operate coherently at all. So yeah - I think that the human brain makes decisions on its own based on its own preferences, without deferring to randomity, and that that complies with every reasonable definition of free will.
@HMHW: Sigh. Now that I’m back I suppose I should try and resolve this discussion with you.
As best I can tell, your position comes down to two points:
You reject that determinism implies that anything in the universe that is ‘sufficiently complicated’ can be predicted by any method. Including simulation, apparently. While you speak of the “choice” being the thing that introduces this unpredictability, logically speaking choice and cognition is just another physical process, with nothing magical about it. Which means that speaking generally you must think that the universe’s future state cannot be predicted from the current state at all - given any process involving sufficient physical complexity.
You think I care the slightest whit about whether future actions can be predicted.
As it happens, I actually don’t give a flying function whether anything can be predicted; I care about whether the universe is deterministic. If the universe is deterministic, then it doesn’t matter how uninformed, obtuse, or lacking the ability to simulate the universe people are when they try to predict the outcome of choices - the outcome of the choice is predetermined nonetheless, because determinism. And that being the case, one must needs consider whether it’s possible to both make a choice and be enslaved to the physics in and around your mind. And it is! Yay!
I also happen to believe that the universe being deterministic axiomatically provides a sure (if theoretical) method of predicting future events: simulation. Because duh.
But nothing in my thoughts on free will and determinism is dependent on believing in simulation, because predictability isn’t the problem. Inevitability is. People have an intuitively difficult time reconciling the concepts of choice and inevitability. Which is why compatiblism requires a certain amount of explanation to most people.
I don’t care about predictability as such (and what you think or not doesn’t feature at all in my position). I care about whether prediction can be ‘short-cut’, or perhaps even better, I care about reducibility. If we can take the state of the universe today, or yesterday, or at the big bang, and from it, derive whether your box of cereals at some time tomorrow will be more empty than it is now, and if we can do it ‘on the fly’—in the same way we can predict the point where a thrown ball touches down without taking into account the way it took there—, then whether you’re having cereal for breakfast is reducible to the state of the universe at some point in the far past. Consequently, the actual choice you make, the deliberation, the—perhaps—computations you carry out, don’t really figure in bringing about that outcome: they’re extraneous, epiphenomenal. They’re not even a contributing reason to you eating cereal: everything is completely fixed without even thinking about the process of your choosing. There’s no question of that choice being in any way responsible for the outcome: it doesn’t even figure into it.
Furthermore, you’re still missing the point of undecidability. As above in my discussion regarding Frankfurt cases, the above still doesn’t seem to suffice, to me, in order to give rise to actual freedom worth being called that. While being constrained, even to one single outcome, doesn’t necessarily make you any less free, it doesn’t make you any more free, either; so at best, we’re question-begging: you’re free if and only if you’re free.
That’s where the undecidability comes in. It ensures that there are several logically possible courses the universe could take. It introduces options consistent with the data at some time in the past. One of these options is then realized via a perfectly ordinary physical, and yes, deterministic process; but that process has to be taken into account; hence, it is responsible for the outcome. Without it, there is no way to decide which option occurs; the question does not have an answer.
In the Frankfurt example above, without the logical possibility of voting for either A or B, the choice is already fixed before the neurosurgeon’s procedure is implemented. So while introducing this constraint to completely determine how the patient votes does not lessen their freedom, it also does not add to it. But if there is the possibility of alternatives, then this means that even being constrained to only realize one of them does not diminish the freedom thus introduced.
I’m a new user and I haven’t read through the whole debate up to this point, but I thought I’d link in some relevant material to the subject.
Sam Harris argues that humans do not have free will. Whichever ‘side’ you are debating for, Harris brings up some fair points and I do recommend watching the talk.
From what I have read on thread, a lot of posts seem to be concerned with whether the universe is deterministic or not. Why does that matter? Whether the universe is 100% deterministic or there is some degree of randomness or uncertainty, I don’t see where a free will variable comes into play? Perhaps I am not using the correct definition of free will as everyone else.
This is exactly like saying that because you can use your equation to predict the thrown ball, gravity doesn’t really figure in bringing about that outcome: it’s extraneous, epiphenomenal. Despite the fact that gravity is written straight into that equation you used as explicitly being an integral and utterly necessary component in reaching your prediction about the outcome.
To be as gentle as possible, this is wrong. Flatly, transparently, obviously, wrong. I can’t even entertain the idea that this is a correct way to view the situation, it’s that wrong.
And because this is so wrong, and because it is so incredibly obvious that the process of choice, like the effect of gravity, could not possibly be excluded from ether the process of predicting or the process of actually deciding, then it is clear that choices occur and we don’t need to screw around with this undecidability nonsense which can’t possibly introduce anything useful because determinism is determinism and under determinism, only one logically possible course the universe could take exists. Because every instant of time along that logical course is also predeterminined, because determinism dictates every future time, not just the one after the decision is made. If there are two roads to Rome, and your position when halfway to Rome could be more than one place, then you’re not talking about determinism.
Suffice to say, if freedom needs to “rise” from something, then it needs to rise from the cognitive processes of the brain, and the fact that uninterfered-with cognitive processes making choices totally meets with the definition of free will.
Oh, and this thing with the Frankfurt box? In a deterministic world, the frankfurt box was always going to have been there at the time of the decision being made. So when you say talk about adding it or not failing to have an impact on the amount of free will because determinism already robbed them of it, you’re completely failing to understand what determinism and free will are.
The frankfurt box robs people of free will not because it prevents the universe from ceasing to be deterministic for the duration of the decision making process, but simply because it violates the definition - it messes with cognition to impose outside interference with the decision-making process. Simple as that. (Whether it counts as interference if it only interferes if you disagree with it is another discussion, but not one that will help your argument any.)
I don’t have a spare hour at the moment; can you summarize it?
From my understanding of things, the discussion on free will originally was a question about whether people made choices for themselves, or whether God has already made their decisions for them due to his ability to predict their actions. The idea being that if God says there’s going to be a war, then all the humans can’t all sit down and decide to watch Netflix instead. The business in some translations of Exodus where God is “hardening the Pharoah’s heart”, repeatedly forcing him to continue playing bad guy for Moses, comes to mind.
With that sort of situation, where god is actively messing with people’s hearts/minds, it’s pretty clear that there’s an abrogation of free will by any definition; God is changing his mind for him. But on another note there’s the idea that God knows the future ‘just because’ - he doesn’t have to manipulate you to predict what you’ll do; he just knows. This notion leads quite rapidly to the idea that the universe must be locked into one specific path of events forward into the future, forever - if it wasn’t, God wouldn’t be able to predict things reliably.
This view of God’s foresight was disturbing, because it seemed to rob people of the ability to choose - if Adam was fated to fall, then he had no choice in the matter; he was doomed to fall. This line of thinking troubled people, because lots of interesting nilistic things can be deduced if you decide that people are unable to make choices.
Exit god, and enter physics. Under physics the goalposts change - rather than God’s mind versus people’s minds it’s physical domino-falling causation versus sheer randomity. The arguments don’t translate over too well, and so people get confused. Many people think that determinism caused by physics is as much an oppressive, choice-stealing force as God’s domineering will, so they try to come up with some way to sidestep or get around physics to attain something they call “libertarian free will” which is rather poorly defined, to say the least.
People who say, “Wait a tick, physics and the physical world includes us, so if we’re chugging along in the physical world reacting to things and making choices, that gives us the same free will we had before people introduced God’s interference!” are called compatiblists. This approach seems to be vaguely looked down on, for some reason.
I’m honestly not sure if you’re being serious here, but, in case you are: no. It’s the exact opposite: I’ve said that you can leave out the process of choice because doing so does not affect the outcome, i.e. the ball landing; leaving out gravity does affect the outcome, in that the ball simply wouldn’t land. So gravity is instrumental in bringing about the outcome; but, for instance, the detailed sequence of states the ball traverses—the equivalent of the computations you go through in making a choice—does not play a role: by simply solving the equation, you can compute where the ball lands, without any necessity of talking about intermediate states.
Furthermore, I’ve been at pains to always include the laws of physics—which include gravity—within the data needed for prediction.
This is right. Which is why you can’t leave it out. But sans computational irreducibility, you can leave out all intermediate states and simply use the boundary conditions plus the evolution law to arrive at a given outcome; hence, that data is sufficient as an explanation for the outcome. So, in such a case, when asked what you will eat for breakfast, I can just point to this data, as the answer will readily follow from that; but if there is computational irreducibility, the only way to realize this outcome is to have you essentially make that choice, and consequently, every explanation of the outcome is going to include you making that choice.
Again, this bears little to no relation to what I’m actually saying. I’m not saying that determinism robbed somebody of free will; I’m saying that in the scenario with the manipulative, but inactive (this is crucial) box, the decision is only free if it would have been free without that box—i.e. while the box does nothing to diminish free will, it also does nothing to suddenly add it.
It’s possible to construct a generic algorithm that allows you to predict the outcome of decisions in advance without in any way referencing or accounting for the mechanical cognitive practice of decision making itself.
Since there is such a predictive algorithm, this allows you to state that the standard analysis of free will under determinism is not allowed to so much as reference the mechanical cognitive practice of decision making.
Because standard determinism is now disallowed from recognizing the mechanical cognitive practice of decision making, this allows you to propose an allegedly alternate method of analyzing free will with the special and unique property of accounting for the mechanical cognitive practice of decision making - which, again, only your model is allowed to do, despite taking place in a totally deterministic universe.
Having carefully carved out a special unique status for your model, you then dress it up with references to undecideability and Godel and the halting problem and so forth in order to draw further distinctions between it and the normal analyses of compatiblism, which recognize that the inevitability of determinism makes such considerations moot even if they were correctly understood and applicable.
The thing is, though, your assertion that it’s possible to create a model for predicting future choice that doesn’t include the mechanics of choice within it (as the thrown ball equations include gravity) is complete fantasy. You certainly can’t present such a model yourself. (If you could, you wouldn’t be screwing around here - you’d be using it to predict the stock market!) And I am personally absolutely certain that no such model could possibly be created, because that’s just stupid - all predictive models take into account all the factors that influence the results in any significant way, and you’d have to be batshit insane to believe that the mechanics of the mind have no influence on people’s decisions. (Well, that or a particularly uninformed theist who doesn’t believe that brain matter does anything at all, but that isn’t the argument you’re making. I think.)
Excluding the delusion that under determinism choice doesn’t really change your model, though, because determinism on its own renders all your blather about undecideability and Godel and the halting problem and so forth moot - at a mechanical level the entire choice proceeds in an unvarying manner, regardless of decideability, because determinism by definition doesn’t allow for potential variations in the cognition process. But the absurdity of pretending that choice under determinism can’t depend on the process of choosing is still worth noting because your entire argument relies on obfuscation, distraction, and confusion to serve as smoke and mirrors concealing the fact that your model takes place in a deterministic universe and thus none of this undecideability garbage you like so much can possibly have any impact on the outcome at all.
Your argument utterly relies on the claim that determinism robs people of free will, because otherwise there would be no way for your alternate view of the process to give free will back to people.
But that aside, the Frankfurt box is an interesting scenario that, with its apparent paradox (how can a (conditionally) inactive box effect free will?) compels us to reexamine our idea of what free will is. Or it would, if I weren’t a compatiblist, and more importantly, one that owns a dictionary. As a compatiblist, I recognize that mechanical brain activity is the font from which free will springs (to be a bit poetic about it). And I recognize that strapping a device to the brain that observes and conditionally modifies brain function is an external modification of the decision-making process itself. That means the “will” isn’t quite as “free” anymore - period. So to me it’s not a contradiction, and nothing to be concerned about.
It’s not, though: I can do so, for the ball thrown in a gravitational field. In principle, all the universe could be like that. It’s not, but recognizing that isn’t trivial, and hinges on the fact that certain complex systems can only be predicted by explicit simulation.
Well then, glad you settled that once and for all.
Which is my point: there are systems such that the models we use to predict them can take ‘shortcuts’, i.e. disregard certain aspects—say, the set of states the ball traversed between the highest point of its trajectory and its impact. Thus, those states don’t play a role for where the impact is: the ‘decision’ made has no impact on the outcome. It takes some insight into computation and complexity to see that not every system is like that. (Well, it does if one isn’t you, and therefore not satisfied with just calling things ‘stupid’.)
I’m trying to be as clear as possible. I’m sorry if these things still seem difficult to you, but I will continue to patiently try and explain them.
Again, my model has three simple ingredients. First, undecidability means that the set of data before a given choice does not fix its outcome—this outcome isn’t logically implied by that data; you can’t sit down and derive the outcome. Hence, futures in which it goes one way or another are logically consistent.
Second, computational irreducibility. A system is computationally irreducible if there exists no model capable of significantly short-cutting its evolution; in particular, this means that every simulation of such a system will be computationally equivalent to just setting up the original system and letting it run its course. This is different from, say, a ball thrown in a gravitational well, where just the initial data (and, of course, the laws of gravity) suffices to immediately compute the point of impact, without having to go across all the steps in between.
Now, combine undecidability and irreducibility to get the necessity that, in order to realize a given outcome, the system has to traverse all the steps to get there; and only after including all those steps does a given outcome become manifest. Thus, only the initial data (and evolution laws) together with that process make it possible to logically derive the outcome of a given choice; only this set of data makes one alternative real, and the other impossible.
This gives us systems that are free in a sense in which the thrown ball isn’t. Perhaps the first thing for you to try and understand is how such a system differs from that ball. Of course, we then need to introduce will to get free will—but that merely means that the free system has to be of a certain kind, namely, an active agent, with goals, aims, and the like.
Sure, the whole thing is still deterministic; I’ve never claimed it isn’t. But that’s where Frankfurt’s thought experiment comes in: mere determinism does not suffice to rob an agent of its freedom. If there is freedom in the first place—which comes from undecidability—then that freedom may be preserved even if one outcome is fixed. And then Bob’s your uncle: we have all the ingredients for a meaningful free will.
Again, that’s getting it the wrong way round: Frankfurt’s point was that even though the agent is completely fixed to take only a certain choice, that doesn’t mean they’re not free, precisely because the box remains completely inert and hence, does not coerce them.
I think I just got hit by a moving goalpost. Let’s try to be clear here:
You can disregard the mechanical effect of the ball hitting air molecules because the effect on the outcome is insignificant. Unless it’s not - if the ball is tossed in a high wind, such as in a hurricane, the simple models fail to predict the outcome.
So, when you talk about the tossed-ball models, you are saying that you believe air molecules to be insignificant to the result - but you consider gravity to be significant to the result. All factors that are significant to the result are accounted for in the model; factors that you consider insignificant to the result are not.
By claiming that you can leave the mechanical process of cognition out of your predictive model for choice, you are saying that you believe that those cognitive processes have an insignificant effect on the result. If they did not have an insignificant effect on the result, then like the hurricane winds your omission of them would be a critical failure in the construction of your model.
So suffice to say that if you claim that there exists a predictive model that excludes the mechanical process of cognition as a factor, then you sure as hell can’t then turn around and say that waiting for the mechanical process of cognition to complete is a necessary part of your model.
Which can’t be true in a deterministic universe.
There are differences between a deterministic universe and a nondeterministic one. These differences can be expressed in terms of logical statements. Here’s one of them:
Given a specific prior universal state, the universal state at every future time is fixed to a specific universal state.
So, logically, given a complete set of universal data, the outcome is fixed - not by your models or ability to predict them, but by the fact the universe is deterministic and thus logically only allows one outcome: the one that will inevitably happen. The assertion that there are multiple possible futures is logically incompatible with determinism.
Which, if we accept your definition of undecidability, logically proves that there is nothing undecidable in any deterministic universe.
This has never bothered me, because I have always proposed “just simulate it” as an always-available model to use when trying to predict the future in a deterministic universe. Or put another way, the reason that nothing is undecidable in a deterministic universe is because you are always provided at least one way to determine it - by logical necessity based on what is logically implied by determinism itself, simulation always works.
(Obviously for it to be actually useful as a predictive method in the real world, you’d need your complete-universe-simulation to run faster than real time, allowing you to ‘see the future’ in it - but that’s an implementation problem, not a philosophical problem.)
You don’t have undecidability, so I believe this disappears in a puff of logic. Determinism has no commentary on whether you’re forced to traverse all the steps to get to the outcome, or whether the mechanics of the brain allow simplifying assumptions to be made to disregard certain things. But it does say with absolute logical certainty that the traversing of those steps has absolutely no impact whatsoever on whether one alternative is real and the other impossible. There was already only one possible alternative, because the definition of determinism logically eliminates the possibility of multiple logically possible outcomes.
I am super okay with there having to be an active agent with goals, aims, and the like for there to be free will - it’s either explicitly required by the definition or very, very strongly implied. But that has nothing to do with the “sense” in which you’re claiming that the systems are different. The agent and the ball are most certainly not different due to the agents forcing in a kind of undecideability that is logically impossible under determinism; they are different because the ball doesn’t meet the expectations we have of the kind of agents that are compatible with free will.
Of course determinism doesn’t rob agents of free will; free will is defined in a way that is compatible with determinism. However, free will is not defined in a way that is compatible with a Frankfurt box. When you start strapping boxes to people that reduce their ability to make decisions based on their preferences, you are reducing their free will, which states that they are able to make decisions based on their preferences.
Hilariously enough, when I went to look up Harry Frankfurt’s cases, it turns out he’s flat-out arguing against determinism being a problem for free will - which is to say, he’s arguing against you. At which point I have to say, now that I’ve read the original argument, which is about moral responsibility, not free will directly - I don’t like his analogy much when you try to take it out of its original context, as you have. It works fine for saying that people are responsible for their own actions if they are made entirely on their own choices, uninfluenced by the unknown threat of external interference. It does not work fine for saying that just because he managed to make that one decision without running afoul of the limitations on his will, that he has unfettered free will. You don’t have to smack into a wall for it to be there.
Not what I’m talking about. I should probably be more explicit. Take a system S, which across some time interval traverses a sequence of states. For simplicity, assume discrete time-steps t[sub]1[/sub], t[sub]2[/sub], …, t[sub]n[/sub], and consequently, discrete states S[sub]1[/sub], S[sub]2[/sub], …, S[sub]n[/sub]. These states could be virtually anything: from, say, the precise position of a ball thrown in a gravitational field, to the steps of a computation realizing a decision mechanism.
Then, my claim is the following: for simple systems, all you need is S[sub]1[/sub] (or, in fact, any S[sub]i[/sub]) and the laws of physics in order to immediately derive every other S[sub]i[/sub], including S[sub]n[/sub]. That is, in order to know where the ball lands, you only need to know how it’s thrown, plus the laws of physics (being real careful here, but just for future reference, this is always implicit, even if I neglect to mention it).
Consequently, S[sub]2[/sub] to S[sub]n-1[/sub] don’t play a role in bringing about S[sub]n[/sub]: S[sub]n[/sub] is already fully determined via S[sub]1[/sub]. You can short-cut the system; there’s an equation such that if you enter the initial conditions, you can immediately derive the state of the system at any point in time.
This is no longer so for complex systems, which are computationally irreducible. Here, you need to, essentially, move through all the states S[sub]1[/sub], S[sub]2[/sub], …, S[sub]n-1[/sub]. Each of these states, and hence, the whole process, is a necessary ingredient for bringing about S[sub]n[/sub]. Whenever you want to see what S[sub]n[/sub] is like, you’ll have to traverse the whole sequence: hence, irreducibility.
Now, is the difference clear here?
I know that this is a really hard intuition to shake, but I would think that my explicitly showing you a model in which there is undecidability in a deterministic universe would have some effect on you; but it’s as if I’m just talking to the wind.
But it’s in the end not anymore difficult than the Gödelian phenomena: you have a set of truths (the state of the universe at a given time/the axioms of a formal system), a set of derivation rules (the laws of physics/the rules of inference) and statements whose truth is not derivable from the above (the existence of a spectral gap/the Gödel sentence). Nevertheless, their truth values are completely fixed: the Gödel sentence is true if the axioms are consistent; and they either are, or fail to be. There either is a spectral gap, or there isn’t; a Turing machine either halts, or fails to.
Do you agree that both the truth and the falsity of the Gödel sentence are consistent with the axioms of the formal system within which it is formulated?
But in a world in which the outcome does not depend on these brain mechanics, in the same sense as the ball’s landing point does not depend on its prior states, how could one hold that those brain mechanics are responsible for said outcome? (Again, this isn’t the world we live in; it’s a hypothetical. It’s about the way the world could have been, but isn’t, luckily; it shows that there are deterministic worlds in which no free will exists.)
I need not know about a sack of rice falling in China to determine where the ball lands; hence, that sack of rice’s fall is irrelevant to this outcome. Likewise, I need not know about the ball’s intermediate states; hence, they’re just as irrelevant.
Again, this has nothing to do with the point I’m making. I’m using the ball as an example of a system that isn’t free; the combination of measurement apparatus and gapped-or-not quantum system for one that is; and finally, an agent that is free in this sense, but also has aims, goals, and intentions as an example of a system with free will.
So, you are arguing that counterfactuals have a bearing on whether we have free will. That is, because the box could have intervened, if the choice had been the wrong one, even in the case where the right choice is made, there is no free will; even though this case is indistinguishable (as regards causal factors) from the one where the decision is
made in the absence of a box. Consequently, causal factors, to you, are not sufficient to determine whether an agent has free will: we also have to look at what could have happened, but didn’t; we have to look at what does not actually influence a decision at all. But then, we can never really claim whether anything has free will: something unknown could always have done we don’t know what.
Yes. I said as much in the first post bringing them up. I then proceeded to note a common objection against them, that while free will is not lessened in Frankfurt’s examples, they also can’t generate it where there previously was none, ultimately rendering them circular. Then, I pointed out how my model circumvents this criticism.
But you do in order for it to be relevant to your movement. Frankfurt’s examples basically are like closing off areas of a labyrinth you wouldn’t have gone down anyways: however free your movements were beforehand, they’re still just as free, as nothing is hindering them.
In a universe that relies on causality, like ours, then your ‘consequently’ is clearly false; the laws of physics operate on the state of the universe incrementally to create each successive universal state from the prior one. In this situation all the substates clearly play a role in bringing about the later substates. The existence of predictive models that don’t explicitly account for intervening states doesn’t change this fact; it merely indicates that the laws of physics operate in a consistent way across time.
This need not be the case…
If you are unable to analyze state of the universe and the laws of physics to predict outcomes, that means one of two things:
Your understanding of the laws of physics is incomplete.
You are not in a deterministic universe.
Because under determinism, by definition, there is one “law of physics” that MUST be true:
At time t, for all times t, the state of the universe is U[sub]t[/sub].
I’ll call this “law of physics” “DET(t)”. And by applying DET you can straightforwardly find the result of any system with the universe, complex or otherwise, without working through the intermediate steps.
Of course, this is not the sort of law that you could derive the complete details about by examining prior states of S - the thing is basically just one big scary lookup table whose later states don’t necessarily rely on prior states at all. Nonetheless it is a big scary lookup table that MUST exist for every deterministic universe, by definition, and if you knew what it was, no problem in any deterministic universe can meet your definition of “complex”. Heck, even if you don’t know it, even then no problem in any deterministic universe can meet your definition of “complex”, because ignorance is no protection from the (natural) law - DET exists, whether you know it or not.
It’s as if you’re deliberately ignoring the definition of the word “deterministic”. Because if the word means ANYTHING, then it means that undecidability cannot exist - DET doesn’t allow it.
But I’ll play ball. Please explain what YOU think determinism is.
I don’t agree that you understand ANYTHING about Gödel, Turing machines, or any other named argument that you’ve brought up. And frankly I’m tired of discussing them with you.
Fortunately, I don’t have to! The existence of DET proves beyond doubt that none of this misrepresented obfuscating garbage can possibly legitimately apply here.
And here’s how it does!
Axiom 1: Anything you like that attempts to insert or imply undecideability.
Axiom 2: The universe is deterministic.
Axiom 3: If the universe is deterministic, then a DET exists.
Axiom 4: If a DET exists then undecideabilty doesn’t exist.
Lemma 1 (from A2 and A3): A DET exists.
Lemma 2 (from A4 and L1): Undecideabilty doesn’t exist.
Conclusion (from A1 and L2): Logical contradiction!
As you know, when a logical contradiction is encountered in a proof, then either the logic is bad or one of the Axioms are bad. I maintain Axioms 3 and 4 are rock solid, as they derive from definitions, and Axiom 2 is taken as given in our discussion. Which means the only weak link, the axiom that has been proven false, is Axiom 1.
No matter how you try to support it.
Unless you’d like to dispute one of the other axioms?
(Note: My apologies to anyone else wandering through this thread - I had been hoping to avoid bringing in formal logic, it being scary. But my patience has worn thin. I’m hoping it doesn’t wear thinner, because if I’m pushed to stating things in formal symbolic logic, then this thread will become freaking incomprehensible.)
In a hypothetical universe where the outcome doesn’t depend on brain mechanics, that means that causality doesn’t exist. And of course if causality doesn’t exist then nobody has free will, because free will is when choices are caused by preferences.
(If the universe is deterministic it still has a DET, but the varying time/states in DET are arbitrary and not caused by each other. For a convenient example of this, consider the average fiction book. Each page (well, each sentence or even noun/verb clause, really) represents a different moment in the book’s “time” - but the contents of the pages are not dependent on the contents of the prior pages in any causal way. And thus, unsurprisingly, we outside the books don’t consider the characters within to have free will.)
Yep! But this is a distinct and interesting discussion that has no bearing on the rote mechanical disproving of your undecideability model. The way I see it, we should resolve that boring stuff first before moving onto a discussion of whether slapping people with a bunch of hypnotic compulsions that turn them into slaves unable to disobey your will counts as an abrogation of free will.
So, just to get clear about it. You’re actually claiming that this research is wrong based on your a priori argumentation?
This ‘law’ would be true even for an indeterministic universe; it says nothing but that the universe has a definite state at each point in time.
This is false. Take a very simple universe consisting of a point particle on a line, between the points zero and one, with the full evolution simply being that it moves from one end to the other. There then can’t exist any such lookup table. This is a simple diagonalization argument.
Each line in the lookup table corresponds to a position of the particle on the line, i.e. a real number. Now, let the table be infinitely long: it’s easy to see that still, not all possible positions of the ball can be contained in that table. For we can construct a valid position of the ball in the following way: take the first digit of the first entry in the line, the second digit of the second, the third digit of the third, and so on; then, add 1 to each digit, with the convention that adding one to nine yields zero. We then have a possible particle position different from every position on the list—it differs from the first line in the first digit, from the second in the second, from the third in the third, and so on.
Consequently, there is a particle position not contained within DET(t). Now, if you add it to DET(t), then we can just run the same argument again: the lookup table can never be complete. So, determinism doesn’t imply the existence of DET(t).
Thankfully, my understanding of those topics doesn’t really depend on the agreement of somebody who had to have it explained to them that Gödel’s theorem has nothing to do with solving equations.
So that’s false, as shown above. As of course it must be: we already have an explicit example of your contention, that there is no undecidability in a deterministic universe, being false.
Which is something you seem to reject in the case of Frankfurt’s examples.
Actually, I’m arguing that if that is actually an example of undecidability in our universe it proves that our universe isn’t deterministic. Proofs by contradiction can disprove any of the axioms, after all.
Which is a place I’m willing to go, by the way, but it changes everything. The entire discussion changes if we’re discussing a nondeterminisitic universe, and I get to bring a whole big shiny new argument.
Okay, it’s a shiny old argument: that any deviation from determinism must by process of elimination be complete randomness, and therefore can’t improve free will even the tiniest bit.
Okay, not so shiny after all, but pretty solid. More so than specious analogies.
Don’t be silly; don’t you understand what nondeterminism is? It’s a model under which only the past is fixed, but the future is not. Which is to say it enforces the awareness of a ‘current time’ t[sub]now[/sub], such that for times t where t > t[sub]now[/sub] it’s impossible to assert the truth/existence of any specific U[sub]t[/sub]. (For t > t[sub]now[/sub] the timeline exists in a determined state, and thus a DET(t) that applies only to the range [0…t[sub]now[/sub]] exists.)
Which is to say, it’s very literally and specifically what you assert when you claim that the universe at the time of S[sub]0[/sub] includes undecideability about the state of time S[sub]t[/sub]. It’s the whole core of your position.
Well, until you turn around and declare that your model operates in a deterministic universe, but that’s just because you don’t know what the hell you’re talking about and are hellbent on not just fighting hypotheticals, but definitions.
Determinism MUST imply a DET(t), by definition. And you can fight definitions as much as you like, and will continue to be flat wrong as long as you like.
However, there is a possibility that you just proved that there must be a ‘Planck time’ - a level of granularity/discreteness to time such as to allow the information describing a universe to be finite. And I am perfectly willing to accept the existence of Planck time! So yay!
And you’ve also proven that no universe can be of infinite duration - all universes must logically have a discrete start time and, if the universe is deterministic, must also have a discrete end time! (Nondeterministic universes can ‘grow’ - at any given metatime they have a finite amount of time, but as metatime passes the amount of determinied time increases.) I’m totally cool with this, too! So double yay!
You think you understand these things well. I think you have gone to some length to demonstrate that you don’t, to all informed observers in this thread. And I think we can agree to disagree about that, because fortunately for me I don’t have to even address these tortured analogies, since cold hard definitions back me up. The analogies are specious in any case.
You not only haven’t disproven DET (which is impossible to do), you also haven’t proven the universe we’re in is deterministic. So you’re zero for zero on these assertions of falsehood.
Also, after having bothered to read that short article past its title, I’m not sure that it demonstrates an example of undecideability in the sense of “something that introduces potential variation in possible outcomes” - which is the kind of undecideability you’re relying on. It may simply be stating it’s uncalculable. Which is to say, the actual state of things may be fixed and unvarying, and thus compatible with a deterministic universe; we might just not have any way to run the calculation about it ourselves in finite time. The thing to note is that if this is the case, then it totally doesn’t support your model either - there would be no ‘logical possibilities’ created; only human ignorance about the state of things. Which isn’t exactly hard to find in any case.