Free Will and Theism

It couldn’t get off them.

It seemed to me what you were suggesting, but OK, I misunderstood. Here’s how I understand the usual definition. Back to the trilemma:

  1. Determinism is true.
  2. If determinism is true, free will does not exist.
  3. Free will exists.

If one denies #1 and accepts the other two, one believes in libertarian free will. That’s all there is to it. It has nothing to do with what an imaginary machine like the Futureviewscope or an imaginary being with the powers of omniscience will or will not be able to detect. I imagine libertarians would have varying views on the question. TBH, I have no idea how to try resolve such an issue, and suspect we have already wasted enought time trying to. :wink:

Yes. It seems to be just a vague concept made up by some apologists to try rescue their arguments. My completely biased opinion.

Yes. But this argument is over whether future events can be considered “facts” if they are the result of libertarian free will. And there I go, wasting more time. Probably time to let the matter drop.


Déjà-vu !

That was, from memory, quite a sufficient handful to keep track of.

You are assuming we were adding the results. We only had to count the number of sixes. Less work, but a more random result.

You also have to account for the Percentile Purists, who insist that the only way to roll d100 is an actual hundred-sided dice, not 2d10. :wink:

I could make a joke about the thread being already meticulously clean and refurbished, but will let this latest typo pass. :slight_smile:

1 Like

Actually it’s the other way around, since they know that the d100s are slightly biased towards extreme results.

Counting the sixes gives you a Binomial, and for large samples we can approximate that with the Normal. Its hard to escape the LLN! :slight_smile:

I’ve never seen a d100 I would trust to roll fairly. Does that make me a Discrete Uniform-itarian???

1 Like

The possibility of an omniscient being or a Futureviewerscope is not necessary to characterize various positions pertaining to free will, but I do not agree that there is nothing these have to do with one another.

If it is the case that there can exist a being (or a machine) that knows (or can reveal) the future infallibly, then it is not the case that the future is uncertain. If the machine or being cannot err, then it is the case that things can only go the way it knows they will. Therefore, no being in such a world can meaningfully direct the future in a way other than the one that is known to come.

There may be subjective feelings of freedom, and without access to that being/machine there may not be a way to tell the difference, but none of that would satisfy a libertarian’s definition of free will. Many libertarian conceptions of free will require agents to specifically be able to choose which way things go. It must therefore be allowed, if libertarian free will of such kind is to exist, that there be the possiblity for things to go another way, were one to reset them back to the moment a choice occurred. This is explicitly excluded, however, if there only is exactly one way things can go. As already explained, this in turn is a prerequisite to there existing a being that can infallibly know said single way things can go.

The libertarian freedom-to-choose-otherwise kind of free will, in my opinion, is incompatible with the idea that there is exactly one option at all times and one was never free to pick differently.

In terms of your “trilemma”, allowing for the future to be infallibly knowable amounts to a rejection of #3 (for libertarian meanings of “free will”), without definitive position on the other two.

Our disagreement, it seems to me, is over whether it is metaphysically possible to know the future if determinism is not true. In my opinion it is possible. But I have no idea how to determine which of us is correct. I suspect philosophers will disagree on this question, as they do on all other questions.

No, I don’t think that’s the disagreement at all, nor can I identify what of anything either of us said made it seem that way to you.

I am making a conditional statement: If it is the case that future events can be infallibly known, then it is the case that events can only play out in one particular way. That’s it. Whether this is metaphysically possible, or under which conditions, was not in dispute.

1 Like

It’s undecidable today. My guess it will remain undecidable for the human intellect.

How true, and how well put. I shall smile all day after reading that. Deterministically, of course.


OK. But, in your opinion, would that also mean that libertarian free will does not exist?

Libertarian free will is (often, anyway, and at most with subtle variations or elaborations) defined specifically as an ability to choose between multiple ways things could go. In a world where there are not multiple ways things could go, there would by definition not be libertarian free will.

So yes, if it is the case that future events can be infallibly known, that would mean that therefore things could only go one way, which in turn would mean that libertarian free will so defined would not exist.

It would be more accurate to say that is a common definition of free will in general. Libertarianism includes the additional stipulation that determinism is not true.

Which is the position that I had earlier interpreted as: If it is (metaphysically) possible to know the future, then determinism is true. You disagreed with that description of your position, but it still seems reasonably apt.

Recall that this discussion began with @gbrooks9’s assertion that “The mere possibility of a deity, or a magician, or a telepath being able to have known YESTERDAY what someone will do tomorrow… does entail the LACK of freewill.” And that is what I am disagreeing with. As far as I can see, a being with the ability to “see” what has happened in the past should be no less able to “see” what will have happened in the future. The idea of an omniscient being knowing with certainty that you had scrambled eggs yesterday (without relying on any present-moment evidence or testimony) seems no less far-fetched than the same being knowing that you will have had waffles tomorrow. And in neither case is it entailed that free will not exist or that determinism be true.

1 Like

All the work in that definition is being done by “ability to choose”. What constitutes an ability to choose? In libertarian free will, it boils down to something that isn’t determined and presumably, isn’t random either. So we’re back to “What would that third thing be?”


Only insofar as that libertarian free will is a common meaning for which the term “free will” is used. There is no general definition of free will, however, and there are common understandings of free will that do not necessarily require an ability to choose between multiple courses of action.

There exist conceptions where an almost-determinism is true, but volition is explicitly excluded from being (completely) causally determined, though can itself form a first(-ish) link in a causal chain. But yes, I suppose pure determinism as defined earlier in this discussion would have to be incompatible with libertarian free will.

At any rate, the name of the thesis that free will and determinism cannot coexist is “incompatibilism”.

The thesis that the world in question operates by rules that allow predicting said future from the world’s state at some point in time is determinism. The thesis that things can only go one way is fatalism. They are compatible, to be sure, but not identical, nor do they imply one another. I disagreed with the “… then determinism is true” part, because the possibility of the future being known is not determinism, nor does it imply determinism.

Correct. What is entailed in case a being should know the future without the possibility for error is that there is not multiple alternatives for what things can happen.

This does not mean that determinism is true, because the possibility to know the future is not the same as the possibility to predict it from a knowledge of things at another time. Also, free will in general is too ill-defined to be conclusively dismissed in such a hypothetical, so the possibility that the future be infallibly known does not entail a non-existence of free will.

What it does entail, though, is that there cannot be an ability to choose between multiple possible courses of action, because an ability to choose between multiple possible coursees of action requires that there be multiple possible courses of action. If things can only go one way, there are not. Hence, in a world where the future can be infallibly known libertarian free will (by that definition) cannot exist.

1 Like

@Faizal_Ali ,

The only way you can get away with “libertarian freewill is still possible” is if the prediction or forecast is not always correct. But, if the forecast is 100% correct every time - - how is the actor able to choose anything other than what was forecasted?

Then there is this awkwardky worded qualifier you like to add:
“… where there are NOT multiple ways things could go, there would, by definition, NOT be libertarian free will.” Agreed… as long as you also agree that even if there ARE multiple ways things could go, if the forecast is always correct then there will always be suppressed freewill.

No problem if the prediction is unknown to the subject. The future choice then exerts backward causation on the predictor, who sees the eventual result of the choice. If the subject had chosen differently, the prediction would have been different too.

Just to cause trouble (:wink: ), I think we can define free will as the ability to be random.

There are some Information Inequality theorems, one of which states that you cannot increase the amount of shared information between two random variables by using any deterministic function (You can decrease the shared info in this way - that’s the inequality) . The only way to Increase the shared Information is to add random information from an independent fandom variable.

This is mostly compatible with George’s definition, in the sense that if you have all the same Information as a future state, then that state is perfectly predictable by deterministic processes.

This may create a problem for the origin of free will. Does a person make a “random” choice independent of previous information, or could random error in some otherwise deterministic process give the same appearance as free will?

ETA: Using (or choosing) the wrong deterministic function “loses” some information about the future state, which would also give the appearance of free will/random information added.

Those two (now three) things might be indistinguishable - I’ll have to think on it. :person_shrugging:


Clearly you mean the right thing, but the expression is a little imprecise here. It could just so happen that the oracle is correct every time, but without being infallible. It is conceivable that it could err, but simply never does, for what ever reason. In that case, it is still conceivable that things can go more than one way, but the oracle in question just happens to correctly identify the way they go ahead of time. The availability of choice between multiple things can only exist if there is the possibility of error, but actual error is not necessary.

For that matter, error and incorrect prediction is also not the same. The oracle could lie, say, even if she has perfectly infallible access to the future.

This is assuming that (a) there is causation in general, and that (b) the rendering of the prediction is in one way or another caused by the facts so predicted. What role the subject’s knowledge of the prediction plays, I must admit, I cannot see.

Even then it is not clear that libertarian free will is recovered. For one could suppose a meta-prediction that predicts both the prediction and the action it predicts. Of course one could then say that that meta-prediction could be retroactively caused by the subject’s choice as well. In this way we commit ourselves to the subject’s choice causing as much of the future and the past as can be caused. And if causation is transitive, then that’s quite a lot indeed.

There may not be anything wrong with this logically, but if that’s a bullet we are willing to bite - as if retro-causation alone wasn’t tough enough to break our teeth as it is - I’m not sure.

1 Like