Why would omniscience and free will be mutually exclusive?

Discussion in 'General Philosophy' started by wynn, Jul 17, 2011.

  1. Big Chiller Registered Senior Member


    I would like to add something on this i don't think there is any "P must be true in all possible worlds" rather there is "P must be true regardless of all possible worlds." I.e. what must be true regardless of all possible worlds cannot be said to be a part of each possible world.
  2. Google AdSense Guest Advertisement

    to hide all adverts.
  3. Pierre-Normand Registered Member

    I can't understand what distinction you are trying to make here. To say that P is true in all 'possible worlds' just is another way to say that P is necessarily true. What do you mean by the phrase "regardless of all possible worlds."? Possible worlds aren't parallel universes that are postulated to exist. The expression just is a semantic device used to model statements of modal logic. A possible world just is a way the actual world could possibly have been.
    Last edited: Jul 30, 2011
  4. Google AdSense Guest Advertisement

    to hide all adverts.
  5. Yazata Valued Senior Member

    There was a whole thread about that a few weeks ago, that ended up with me getting flamed as I recall. But yeah, I made a similar (but not identical) point myself over there.

    'God' is just a word in the English language. It has a variety of meanings and uses, derived from the word's history. Those have been typically, but not entirely Christian. Sometimes they've been popular usages, at other times influenced by ancient scriptural Greek and Hebrew. And sometimes they have been philosophical, and sometimes a conscious mixture of philosophy and doctrinal Christian in the form of philosophical theology such as we saw in the medieval universities. Since the age of discovery and the realization that other cultures with very dissimilar religions exist out there, there's been a deist/perennialist tendency as well, that continues to imagine a single monotheistic Godhead but questions whether any of the world's historical religions understands this divine truth correctly and completely.

    All I've been interested in here in this thread is exploring the implications that a certain conjunction of philosophical/theological propositions about divine foreknowledge, infallibility, freedom, possibility and so on seem to have. It's an intellectual exercise, not a religious quest.

    Philosophical theologians, let alone more typical believers, needn't necessarily embrace all of those propositions. That's probably going to be an internal issue inside their own religious organization, namely how much lattitude individual thinkers have to deviate from and to improvise around what may or may not be considered traditional elements of doctrine.

    That's assuming that God exists. Nevertheless, we do have this very suggestive word 'God' and all of the various and sometimes inconsistent ideas, connotations and meanings that have stuck to it over the many centuries.

    The Platonist tendency in the Christian tradition might argue that it's precisely our reason that brings us closest to the divine. It's a faculty that we share, in however limited a fashion, with God himself and that allows us to perceive, however dimly, the existence of the eternal. Traditionally, man's reason has been thought of as what supposedly separates him from the rest of the earthly plane, the spark of divinity within him.

    I'm not convinced that there's anything there to find, apart from what you call the mere "claims of people". If by chance there is, then I don't have any faith that any of what you call the "theistic traditions" contains a special divine revelation that's going to lead me there.

    The religious traditions (theistic or not) all appear to be the work of human beings. That's all that I'm really concerned with. I'm not really seeking God, certainly not in the fideistic manner that you seem to favor. I'm not even entirely sure what the word 'God' means. (It seems to mean a variety of subtly different things, depending on the particular usage.) I just poke through the detritus and am sometimes surprised at what I find.
    Last edited: Jul 30, 2011
  6. Google AdSense Guest Advertisement

    to hide all adverts.
  7. Pierre-Normand Registered Member

    That's no difference at all. I just pointed out to you that this premise is the logical consequence of my premises (2) and (4). If it is the case that P (premise 2) and necessarily, whatever P is true, God knows that P (premise 4), then God knows that P.

    I could just as well have written down K(P) as my premise (4) and in that case the truth of P would have followed from my premise (1). The choice of which premise to use, K(P) or (P) makes no difference since in conjuctions with the other premises they are entailed by each other.

    Your present objection is in any case irrelevant and point missing. I challenged you to derive Nec(P) from the set premises #1 to #4. If you can't, then adding the premise that K(P) will not help you since it was already a consequence of the other premises. Further, you now claim that it never was your burden to demonstrate Nec(P). You now claim that the only sort of necessity you ever claimed to endorse was the necessary conditional expressed with the wide scope necessity operator: Nec(K(P) -> P). But this is just my premise (1)! So, adding the premise K(P) is idle. I obviously grant you the truth of my own premise (1)!

    This is confused. There is no switching or problem redefinition taking place. K(P), which is already entailed by my own premises, is the claim that God knows P in the actual world. Both P and K(P) are true in the actual world, we assume (or infer). Considering possible worlds in which K(P) is false does not contradict that assumption. The claim that Q could have been the case does not contradict the claim that P is actually the case.

    The point under dispute between those who claim that our four premises preclude or don't preclude free-will is the question whether or not it follows from those premises that P is the case in all possible worlds (in which God has infallible knowledge, etc). This is equivalent to the question whether or not some Q (inconsistent with P) could possibly have been the case instead. Our attention if of course restricted to those possible worlds in which God's knowledge is total and accurate. This is what is at issue. Do our premises entail or don't they entail that it is possible that I would not have performed some action A that I actually preformed (the proposition that I performed actions A is noted P). If you debar me from the outset from considering any possible world in which I did not perform A, then this amounts to assuming without argument that my doing A is necessary. Your task rather is to demonstrate from our four premises that P (and hence K(P)) is necessary. You are of course free to make use of the fifth premise K(P), since it follows from (2) and (4).

    It's just inconsequential to ask me to add a premise that was already there implicitly. It changed nothing to whatever you can deduce from the whole set of premises. You are free to add it in explicit form. I don't mind the redundancy if it pleases you.

    If we assume that God has infallible foreknowledge, and we suppose that I could have done otherwise than I actually did, then it follows straightforwardly that whatever I would have done instead, God would have known it. So, if I did P, to deny that there is a possible world in which I did Q instead and in which God knew I would do Q just is to deny that my doing Q would have been possible. Since this the the very point under contention, you can't just rule this out without argument.

    This claim has long since been agreed upon. It is my premise
    (1) Nec(K(P) -> P).

    Sure. But here I am assuming that for some action I preformed, God knew I would perform it. His belief about the future turned out to be a genuine case of foreknowledge, we assume. We just assume that his epistemic faculties are generally fallible but they didn't fail in this case. So my premise (2) Nec(P -> B(P)&K(P)) is now assumed not to hold. But in the case we are considering, premise (1) Nec(K(P) -> P) still holds. It is a premise concerning ordinary fallible knowledge. But this is just the necessary conditional you now claim it was your aim all along to defend and not anything stronger (such as Nec(P)). This is why I say that your new official position regarding the (wide scope) necessitation of our actions -- which God's infallible foreknowledge supposedly entails -- fails to make use of the infallibility condition. This is the trivial consequence of the fact that my premises (1) and (2) are independent premises. You don't need (2) to assert (1).

    Well, the trouble here is that the very possibility for anyone (any old mortal Joe) to remember some action you preformed provides premises that are formally the identical counterparts of the premises from which you infer the (wide scope) necessity or our action from God's having infallible foreknowledge of them. While foreknowledge entails justified true belief about the future, remembering, qua knowledge of the past, entails justified true belief about past events. (This isn't meant as an analysis of knowledge, as Gettier cases show J&T&B not to entail knowledge, but that's not an issue here.)

    The premise here simply is (1*) (1) Nec(R(P) -> P), (In each possible world where Joe remembers P, P is true)

    This premise just is the claim of (wide scope) necessary conditionality that you believe to be inconsistent with free-will.
  8. Big Chiller Registered Senior Member


    I agree that a possible world is another possibility for the world and also there are many different possible worlds ad infinitum. I just disagree that P should be based on any of these possible worlds if its true always. You don't have to take me seriously...
  9. wynn ˙ Valued Senior Member

    Which was a misunderstanding and which I tried to clear, but you were not to be reached.

    Not at all. It's my abstract understanding of the relevance of the concept of God.

    The way I see it, holding reason in such a high regard when it comes to God leads to effectual solipsism.

    A naive transcendentalism (as suggested here earlier by Hesperado, for example) is actually far more rational. Provided, of course, that one already naturally holds that position (I don't).

    Neither do I have much such faith for myself.

    On principle, I doubt that. But just for the sake of the argument, not because I was sure there is evidence that all religions are man-made.

    I resent fideism. I do think that someone who has not been born into a religion, doesn't have a special calling for one, nor has neurotic tendencies, cannot become religious other than in a fideistic leap of faith.
    I do wish that the members of esp. theistic traditions and especially those eager to proselytize would actively relate to this and consider it when they preach.
  10. Pierre-Normand Registered Member

    I still don't understand what it could possibly mean to say that P is based on a possible world. What is it to base a proposition on a way the world could have been -- apart from saying that the world could have been such that the proposition was true?
  11. Big Chiller Registered Senior Member

    All I'm saying is that for a propositon to be made in a possible world is that it's true or false in a possible world based on that world e.g. the speed of light is greater than the speed of light in this world in a possible world and in another different possible world the speed of light is equal to the speed of light in this world therefore any variation of the speed of light is not necessarily such hence it "depends" on the possible world. Now lets take a necessarily true concept e.g. Object A cannot fit in object B physically and be bigger than object B simultaneously. Such a concept is independent of any possible world.
  12. wynn ˙ Valued Senior Member

    But truisms don't prove anything.
  13. Pierre-Normand Registered Member

    OK, now I understand what you mean. But in that case we simply say of the propositions that it is true in all possible worlds. This is the case of all the truths of arithmetic or logic for instance. When we say Nec(P), we are just saying, then, that had the world been different in any respect it could possibly have been, then the proposition P would still be true in that case. The way you are putting it is equivalent.
  14. Pierre-Normand Registered Member

    What are you responding to? Some people believe of some propositions that, prima facie, they are truisms, while they can actually be demonstrated to be false. And some complex and prima facie contentious propositions can be shown to be logically equivalent to simple truisms. So, there may be use in demonstrating or refuting 'truisms'.
  15. wynn ˙ Valued Senior Member


    Take one like this:
    If John has three children, he has to buy three pairs of pants. Therefore, John now has to buy three pairs of pants.

    Many theistic attempts to persuade people to take up a particular religion have that structure.

    First a truism is stated, more or less clearly, sometimes even in the conditional form as in the example of John.
    But the conclusion (ie. instruction for action) does not follow.
    Yet they insist it does.

    Sure. But what gets accomplished with that?
  16. Pierre-Normand Registered Member

    You show an example of a fallacious inference that makes use of a truism as a premise. If you intend this to vindicate your claim that truisms don't prove anything, they you are of course indulging in fallacious thinking. If you only mean that truisms ought to be suitably interpreted -- taking care not to be misled through naively relying on their merely superficial grammatical form -- then your point is well taken. (Although I am still not sure what relevance this has to the issue Big Chiller was raising.)
  17. Big Chiller Registered Senior Member


    Well truisms are better than trivialism.
  18. Yazata Valued Senior Member

    Imagine the hypothetical situation that existed before P actually occurs (if it does) but after God has acquired his knowledge that P is surely going to happen. If God's foreknowlede of P was eternal, then that would presumably mean any time prior to the occurrance of P. In this situation, we already have K(P). Now our question is to determine whether something other than P can happen in the future without contradicting K(P).

    Perhaps you need to explain what relevance you think Nec(P) has to the omniscience and free-will question. I've already argued that challenging opponents to prove that P is true in all possible worlds, including those in which God doesn't exist or knows something other than P, is too strong and rather irrelevant. The question in this thread is given the premise that God knows P, can anything other than P occur without contradicting what God knows.

    So if God already knows P, how can Q occur instead of P without contradicting K(P)? Your (1) seems to say that it's impossible.

    Adding the premise K(P) isn't idle since your (1) is a hypothetical, 'It's necessary that(If K(P), then P)'. You've been exploiting that 'If' pretty aggressively, insisting that the possibility still exists that Q could happen instead of P, in which case K(Q) -- God would have always known that Q was going to happen. I'm saying that making that move contradicts the original hypothetical situation in which God already knows P (or at least thinks thathe does). If Q should occur instead, then God will have been mistaken, and that in turn would contradict the doctrines about his omniscience and infallibility.

    What I'm questioning is slipping into possible worlds in which God knows something other than P. There's no problem in imagining whoever's supposedly going to perform action P doing whatever he or she chooses. The question then becomes, what happens if the person chooses to do Q instead of P? God's foreknowledge would seem to be contradicted and his omniscience and/or infallibility fail.

    If you did Q instead of P, then God's K(P) fails. In other words, if K(P) is always true and if God is infallibly omniscient, then you can't do anything other than P without contradictions arising.

    Whether that implies some coercive force over what you can and can't choose to do is a theological question. The position that theologians take on that question will depend on how badly they want to protect the idea of God's perfection, I guess.
  19. wynn ˙ Valued Senior Member

    Truisms are trivial.

    Just like "Therefore, John now has to buy three pairs of pants" does not follow from "If John has three children, he has to buy three pairs of pants,"
    so "We should all accept Jesus as our Lord and Savior and join the Holy Roman Catholic Church" does not follow from "If the Holy Roman Catholic Church is the only one to hold the keys of heaven, then we should join it."

    Our reasoning is productive when it leads us to taking action that we assess as wholesome. Truisms don't do that, they just keep us stuck.
  20. Pierre-Normand Registered Member

    This had been granted already and in fact it is common ground on both sided of the debate. Remember that it holds true whether someone's knowledge is (1) infallible foreknowledge or (2) contemporary fallible knowledge or even (3) fallible memory (knowledge of the past). It is a premise about ordinary concept of knowledge that Nec(K(P) -> P). This proposition is logically equivalent to the truism Not(Pos(Not(P) & K(P)). It isn't possible for someone to know that P is the case if P is false. So, if we assume that Joe knows (at any time and from ordinary deduction/observation) that I did P, then, by the same reasoning you conducted above, I can't do anything but P. But you don't seem to hold that this hypothetical presents any incompatibility with free-will?

    First, the relevance of Nec(P) consists in its incompatibility with the principle of alternate possibility (PAP). Many libertarians about free-will hold that (1) I can't be held responsible for my action whenever there wasn't any possibility for me to do otherwise; and (2) that freedom of the will entails responsibility. Compatibilists usually agree with (2) but they will insist on replacing the notion of possibility in (1) with a notion of 'can do otherwise' that has conditional form and in which the antecedent refers to some orectic state of the agent (desire, intention, etc). This makes Nec(P) threatening to libertarian free-will but not necessarily to compatibilist free-will.

    Second, I've always excluded from consideration possible-worlds in which God doesn't exist or in which his knowledge is fallible. Whether or not one must also exclude from consideration possible worlds in which he knows something other to be the case (because something other indeed is the case in that world) is the issue under contention.

    Yes it's impossible. It's also impossible in the case where Joe knows P because he has seen me acting. But yet you don't seem to hold that the hypothetical that Joe may have known P constitutes a threat on my ability to have done otherwise.

    Forget about omniscience and infallibility already! If Q should occur instead, it would directly contradict Nec(K(P) -> P) however fallible and restricted God's epistemic faculties may have been. Your intended conclusion directly stems from the premise Nec(K(P) -> P) and K(P). Can't you see that those are the only premises that you really make use of in that paragraph above? Sure you mention omniscience (Nec(P -> K(P)) and infallibility (Nec(Believes(P) -> K(P)). But you make no use whatsoever of those two premises. Your last sentence is an afterthought. You've already derived a contradiction in the sentence before. Hence your proof didn't need the last two premises.

    Let me transpose you argument to another case. 'Suppose Evil Corporation' (EC) sells items to Chloe. Whenever Chloe buys something from EC, EC thereby sells it to her. We thus have the premise about selling/buying (1) Nec(Sells(P) -> Buys(P)); Whenever Chloe is engaged in a buying relationship with EC, EC is thereby engaged in a corresponding selling relationship with her. Does the hypothetical existence of EC threatens Chloe's freedom in choosing what stuff to buy from EC? Here is how one might reason:

    Adding the premise Sells(P) isn't idle since (1) is a hypothetical, 'It's necessary that(If Sells(P), then Buys(P))'. You may insist that the possibility still exists that Q could be bought by Chloe instead of P, in which case Sells(Q) -- In that case it would also necessarily have been the case that EC would sell Q to Chloe. I'm saying that making that move contradicts the original hypothetical situation in which EC was selling P, not Q, to Chloe. If Q should be sold instead, then this would contradict our premise that it is actually P, not Q, that was sold.

    Now, don't you think the above argument is a bit of a fallacy? Actually it's perfectly valid inasmuch as it only aims to conclude that whenever EV sells something to Chloe, Chloe buys it. It is only invalid if it aims to suggest anything stronger.

    But nowhere did I ever invalidly 'slip' into a possible world. Either there is such a possible world or there isn't. You need to provide good arguments. Your argument that there being such possible worlds would contradict the assumption of omniscience or infallibility has been addressed already. It doesn't. All the possible worlds I have considered are worlds in which our premises about omniscience and infallibility hold. Your argument that there being such possible worlds contradicts the assumption of K(P) also has been addressed. It doesn't. If Q and P are inconsistent, the claim K(P)&P does not thereby contradict Possibly(K(Q)&Q). What occurs in merely possible but not actual worlds need not be consistent with what occurs in the actual world, unless, that is, one wishes to endorse the philosophical doctrine of necessitarianism -- that everything actual is necessary -- in which case PAP is directly refuted, whatever God may know or ignore about our actions.

    Your insistence that assuming the actuality of P & K(P) already precludes consideration of the possibility of (K(Q)) (their being possible worlds in which Q & K(Q) is a covert and question begging claim of necessitarianism, it seems to me.
    Last edited: Aug 1, 2011
  21. arfa brane call me arf Valued Senior Member

    What the problem amounts to is:

    Humans can have foreknowledge, or a precondition exists, say, that means we know how to predict something. For instance, we have foreknowledge that a coin has two distinct sides, so we can predict that one or the other will be face up if we spin the coin in the air and it lands on a flat surface.

    We also make logical statements, predicates and propositions. We can form a logical argument about a being with complete knowledge or omniscience, but this argument cannot be applied to humans.

    So if omnisicence exists, does an omniscient "being" also predict events, and do they also "remember" events? Why do they do either (or both) of these things that humans do? Why is it "necessary" to apply this human capacity for prediction and memory to an omniscient being?
    Why are some of the posters in this thread doing this, as if it is necessary?

    Why would an omniscient being "need" to know about any more than an infinitesimally small interval of time, over the entire universe? Would this mean they know everything the universe is doing, and in what sense, seeing as how an infinitesimal interval of time is NOT what we think of as knowledge, which for us necessarily requires prediction and memory, both of which are "fixed" by much larger intervals of time than our predicated omniscient being seems to need?

    Does that mean all the discussion of necessary conditions, possibilities, and actualities is in fact, redundant to the initial question? I think so.
  22. Pierre-Normand Registered Member

    I am unsure to whom these comments are directed. I myself pictured infallible foreknowledge and omniscience on the model of ordinary human knowledge, through a negative idealization process, as it were, which involves the removal of any limitation on the scope of one's knowledge and any possibility of error. However God is supposed to manage such a feat is left indeterminate.

    It is not necessary to attribute these things to God or to assume God exists. Who are the posters doing this assumption of necessity? Most of us, I think, have only been concerned with the hypothetical consequences such divine powers would have on the possibility of free-will if a divine being existed and had them.

    Just as you can't know what a picture represents through looking at it just one pixel at a time and then forgetting what colour the pixels already seen were, a being that could only know about one time-like slice of the universe at once would miss on a lot of facts. How could such a being know that WWII occurred on Earth and how long it lasted?

    Sure, if you picture God himself as residing outside of time, as it were, then it may not make literal sense to say that he predicts or remembers. However, for present purposes, it changes nothing to the arguments if you replace all occurrences of "God foreknows at t1 that P will be true at future time t2" and "God remembers at t2 that Q was true at past time t1" with some sentence of the form "God eternally knows that P" where P expresses an explicitly time-indexed proposition rather than a time-variable state (i.e. some way for the world to be at a time).
  23. Big Chiller Registered Senior Member


    Nonsense, if anything truisms can only be by negating trivialism.

    These are not appropriate examples of truism as these are fallacious and are mere conjectures.

Share This Page