Infallibility

Originally posted by googlefudge
Can you clarify exactly what is meant here by libertarian free will?

I am currently assuming it means the ability to have any physically possible
thought, and undertake any physically possible action "at will".


If that is what libertarian free will is... Then I can't see how it's possibly compatible
with a god (or anything else) KNOWING with ...[text shortened]... , can't be
changed and thus can't be free.

*By the supposedly free agent.[/b]

Generally, libertarianism is a subset of incompatibilism. Incompatibilists hold that one's acting freely is not compatible with determinism, that we could act freely only if determinism is false. Then, historically, such incompatibilists were subdivided into at least two groups: hard determinists (those who claim determinism is true and thus -- coupled with their incompatibilism -- it follows we are not free) and libertarians (those incompatibilists who think we are free and that determinism is false). But, generically, I think 'libertarianism' as I use it here and as it is often used would be proxy for a set of commonly held incompatibilist conceptions. It is often tied up with the "ability to do otherwise" as a necessary condition for freedom. So, under this, S is free with respect to doing A only if upon doing A it is the case that S could do or have done otherwise (or some such). Of course, there are different ways to flesh out the concept of "doing otherwise" as a necessary requirement for freedom; but in rough terms what this means is that for S to act freely at time t, it is necessary that there exist at t a plurality of relevant futures open to S. Another way of saying this is that there need to be other possible worlds identical to this world (or sufficiently nearby) up to time t in which S does/did do otherwise at t than in the actual world.

These sites may have additional relevant background info (or other similar sites on the same website):

http://plato.stanford.edu/entries/incompatibilism-theories/
http://plato.stanford.edu/entries/freewill/

Originally posted by googlefudge
I still maintain that for god to know with epistemic certainty the outcome of future
events then those events must be pre-determined and that there can be no libertarian
free will.

Because it seems to me that libertarian free will requires the ability to make a
non-absolutely predictable decision.


However I would like to hear from anyone who thinks this is wrong and why.

I still maintain that for god to know with epistemic certainty the outcome of future
events then those events must be pre-determined and that there can be no libertarian
free will.

I do not agree and have never seen a good argument that actually shows this. When I have some more time, I will revisit this thread and try to explain why I disagree.

Originally posted by LemonJello

I still maintain that for god to know with epistemic certainty the outcome of future
events then those events must be pre-determined and that there can be no libertarian
free will.

I do not agree and have never seen a good argument that actually shows this. When I have some more time, I will revisit this thread and try to explain why I disagree.

I look forward to that.

Originally posted by googlefudge
I look forward to that.

The statement you made was, again, this:

I still maintain that for god to know with epistemic certainty the outcome of future
events then those events must be pre-determined and that there can be no libertarian
free will.

The type of future events at direct issue here would be events such as S's doing A at some time t indexed to the future. And the claim you are making is that there is no coherent account under which God can know such events with epistemic certainty while S is still free with respect to A-ing at t (under some libertarian construal of freedom). By "coherent" here, let's agree that we would be talking about internal logical consistency and that what we need would just be an account that is internally consistent; the evidential plausibility of such an account would be a separate issue. Also, by "knowing with epistemic certainty" I take it that it would be necessary and sufficient to satisfy this that God knows such events in a way that is free from even the possibility of epistemic error (simply knowing them is not good enough; what we would need is an account that also guarantees that such knowledge is immune to the possibility of epistemic error; otherwise, there would be some epistemic possibility for God that He is mistaken -- even when in fact He is not mistaken -- and that is not good enough for epistemic certainty). Furthermore, I take it that by "libertarian construal" we are generically referring to some incompatibilist conception that requires that live alternative possibilities be open to S with respect to A-ing at t. So with this understanding, the following would be an account that I think satisfies all this and is internally consistent. Here would be the salient premises of the account (and, again, plausibiity of the premises is a separate issue; what we need to investigate here is just whether or not these premises entail a contradiction):

(1) God exists.
(2) God is a perfect knower, subject to the infallibility condition that Soothfast and I were discussing:
(2a) Necessarily, P if and only if God knows P.
(2b) Necessarily, not-P if and only if it is not the case that God believes P.
(3) S will do A at time t.
(4) Libertarianism is true (e.g., when S does A at t, S will be free with respect to A and there exist alternative possibilities wherein S refrains from doing A at t).

From these, it follows that S is free in a libertarian sense with respect to S's doing A at t and that there are alternative possible worlds wherein S refrains from doing A at t; and yet it is also true that God presently knows with epistemic certainty what S will do. The epistemic certainty here is guaranteed through the infallibility condition on God's knowledge. Even though there are other possible worlds wherein S refrains from doing A at t, there is simply no possible world wherein God is mistaken about what S ends up doing with respect to A at t. That's guaranteed by the modal term 'necessarily' in the infallibility condition. From the infallibility condition, the conditional "If P then God knows P" is true in any and every possible world. So, in a given possible world if it is true that S will do A, then it also follows immediately that God knows this in that world. If in another possible world it is true that S refrains from A-ing at t, then it is also true that God knows that in that world. So there is simply no possible world in which God is mistaken about what S will end up doing with respect to A at t. So, there is the epistemic certainty. And yet it is also the case that S is free in a libertarian sense.

I don't see any obvious contradictions entailed by this account. However, presumably what you will want to scrutinize here is whether (3) and (4) are consistent with each other. Implicit in (3) is the assumption that a future proposition like 'S will do A at time t' has a determinate truth value. Is this consistent with there still being live alternative possibilities open to S with respect to A at t? That would be the key question, I think. I would not be surprised if the answer turns out to be no, since I think libertarian construals of freedom are totally bonkers. But I'm not convinced. I would be interested to hear your thoughts here.

Also, I think one interesting comment would be this: the tension as I see it is simply between (3) & (4), neither of which have anything per se to do with God or God's knowledge. So the question is not about whether epistemically certain or divine knowledge about our future actions is consistent with libertarian freedom; the question is more basic than that. If a future event (say event E at time T) must be predetermined in order for a future-indexed proposition like "E will transpire at T" to have a determinate truth value; then the truth of libertarianism would seem to preclude any knowledge whatsoever (with or without epistemic certainty) regarding propositions about our future actions; the talk about 'epistemic certainty' would be a distraction on the periphery here.

Originally posted by LemonJello
The statement you made was, again, this:

I still maintain that for god to know with epistemic certainty the outcome of future
events then those events must be pre-determined and that there can be no libertarian
free will.

The type of future events at direct issue here would be events such as S's doing A at some time t indexed to t ...[text shortened]... ture actions; the talk about 'epistemic certainty' would be a distraction on the periphery here.

Thankyou for this response.

I need to think about it some before I respond properly.

I agree with all of your definitions at the beginning...
That's precisely what I was thinking of. So we're on a good footing.

I agree that 3 and 4 in conjunction are problematic.
I suspect my central argument was (and might be) that 3 requires a fixed and
unalterable future. Which is incompatible with 4.

But I need to think about that.


I am not sure that libertarianism would preclude all knowledge...
Given that the idea is so whacked out anyway, it's possible that you could argue that
only sentient beings have free will and that only their actions (and those consequences)
would be free and that the rest of the 'dead' universe, where not effected by sentient
beings, is utterly predictable?



I do have a question about 2a and 2b. Reveal Hidden Content

Are "Necessarily P" and "Necessarily ~P" exhaustive as well as exclusive for all P?

In other words can their also exist P and ~P that are not "Necessarily" determined?

Are there P that god neither knows (and thus Necessarily P) or disbelieves (and thus Necessarily ~P)


I ask, because it looks to me, that if P is "S will do A at time t." and 2a and 2b are exhaustive and exclusive for all P.
Then 2a and 2b taken together would also be in conflict with 4 in exactly the same way 3 is.

Because in that case it seems that 3 follows from 2a and 2b.



I'm working tomorrow, and probably wont get to this over the weekend (Christmas shopping must get done)
But I am off on Monday, and will probably get back to this then.

Originally posted by googlefudge
Thankyou for this response.

I need to think about it some before I respond properly.

I agree with all of your definitions at the beginning...
That's precisely what I was thinking of. So we're on a good footing.

I agree that 3 and 4 in conjunction are problematic.
I suspect my central argument was (and might be) that 3 requires a fixed and
u ...[text shortened]... istmas shopping must get done)
But I am off on Monday, and will probably get back to this then.

I think the main question we need to address here is the following. Consider some proposition about the future, such as "Event E will transpire at time T". For such a proposition to be now known, the proposition needs to be true and hence have a determinate truth value. The question is, in order for such a proposition to have a determinate truth value, does it need to be the case that E transpiring at T is causally determined? If the answer is yes, then (3) and (4) are contradictory. If not, then (3) and (4) should be compatible.

I ask, because it looks to me, that if P is "S will do A at time t." and 2a and 2b are exhaustive and exclusive for all P.
Then 2a and 2b taken together would also be in conflict with 4 in exactly the same way 3 is.

Because in that case it seems that 3 follows from 2a and 2b.

I am not sure I understand fully your questions about (2a) and (2b). But for sure (3) does not follow from (2a) and (2b). What follows from (2a) and (2b) is the necessary truth of some conditionals, such as "If P, then God knows P". This conditional, for example, then combined with a further contingent supposition that P would together entail God knows P. But the atomic propositions P do not themselves follow from (2a) and (2b). That's why (3) is a further supposition. Not sure if that helps....