Rutten’s argument for the existence of God
October 23, 2011 — 11:12

Author: Alexander Pruss  Category: Existence of God  Tags: , , , ,   Comments: 106

Emanuel Rutten sent me the following interesting argument which I am posting with his permission. Please make sure to be clear that if you cite this post, everything except the title, the preceding sentence and this sentence, is taken verbatim from Rutten. He has some other interesting arguments on his blog, some of which alas are in Dutch.
Take the following metaphysical principle, connecting possible worlds, knowledge and truth: ‘If it is impossible to know that p, then p is necessarily false’. This principle seems cogent. For, if a given proposition p could be true, then, plausibly, there is some possible world in which some subject knows that p is true. In other words, if in *all* possible worlds *all* subjects do not know that some proposition is true, then, plausibly, that is because that very proposition cannot in fact be true.
Well, on a Cartesian view of knowledge, that is, to know p is to be certain that p is true, the above principle has an interesting consequence. For, take for p the proposition ‘God does not exist’. It seems reasonable to hold that it is impossible to know that God does not exist. For, whatever the arguments against God, there will always be some (perhaps an extremely remote) possibility that God does exist after all, so that we can never truly say, on the Cartesian view, that we know that God does not exist. But then it follows that it is necessarily false that God does not exist. Hence, it is necessarily true that God exists. The principle thus entails theism. Is this new argument for theism convincing?


The atheist might object that it is also impossible to know that God exists. And thus, by similar reasoning, it would follow as well that it is necessarily true that God does not exist. However, I would argue that there is a possible world in which some subject can truly say that he or she knows that God exists. Take a possible word in which God exists and in which there is an afterlife, such that all who enter the afterlife in that world will encounter the divine. In that case, those subjects who enter the afterlife will in fact know that God exists. So, it is not impossible to know that God exists. Note that a similar move to reject the argument for theism is not open to the atheist. For, if God does not exist, then, plausibly, there is no afterlife. And besides, even if there would be an afterlife, then entering it would not bring a subject in the epistemic condition of knowing that God does not exist.
Now, the atheist might want to offer three further objections to the proposed argument, which shall I present and respond to in what follows. The second objection (#2) was suggested to me by Alexander Pruss.
#2
The principle on which the argument for theism is based can be formulated as: ‘If p is possibly true, then p is knowable’. This principle entails that every truth is knowable. But from that, as Fitch has shown in his 1963 paper ‘A Logical Analysis of Some Value Concepts’, it can be logically deduced that every truth is in fact known! An atheist might reason that this is a very problematic, if not absurd, consequence. Thus, as the atheist would have it, the proposed argument is not convincing and should be rejected.
I would respond as follows. Now, it is indeed the case, following Fitch, that the principle entails that every truth is in fact known (call this consequence T). But why hold that T is false? After all, for all we know, there might be an omniscient being in the actual world knowing all truths. (I recently read an excellent refutation of Dennis Whitcomb’s argument that omniscience is impossible.) So, even though T does seem problematic for atheism, it does not follow that T is false. It would be begging the question for the atheist to deny T solely because T does not fit nicely with atheism and favors theism (since the theist can hold that God knows all truths). Indeed, the fact that my principle entails T is not sufficient to reject it. For, it would be unreasonable for the atheist to initially accept the principle as plausible (which I would contend it certainly is), but then, when it becomes clear (after a non-trivial complex deduction) that it has a consequence unpleasant for atheism, to reject the principle.
#3
The atheist might refute my response to the first objection. After all, someone could, even encountering God in the afterlife, believe that he or she is dreaming, or hallucinating, or being deceived. Therefore, on the Cartesian view of knowledge, it is impossible to know that God exists after all. But then, by parallel reasoning, it also follows that, necessarily, God does not exist. And thus the new argument fails. My response would be that even if someone could always think that he or she is dreaming, hallucinating or being deceived, it still does not follow that it is impossible to know that God exists. For, take a possible world in which God exists. In this possible world there is a subject that knows that God exists, namely God. Indeed, in that world God knows that God exists. So, it is not impossible to know that God exists.
#4
Another objection would be to argue that there might be some true mathematical Gödel sentence G that cannot be proven by any proper mathematical system. Hence, G is unknowable. But then not all truths are knowable, and therefore my principle (which entails that all truths are knowable) fails. My response would be that G is in fact knowable. For, there is a possible world in which G is known. Take again a possible world in which God exists. In that world God can be taken to know (at least) all mathematical truths by direct immediate intuition, and therefore God knows G as well.
Emanuel Rutten
VU University
Amsterdam
The Netherlands

Comments:
  • Now me again. So theism lets you connect truth and being known, possibility and knowability, necessity and necessarily being known. That’s a neat move which seems to be some evidence for theism: the more orderly a scenario is, the more likely that it is actual. But I think what makes a piece of order be even better evidence is when the piece of order is explanatory. But here I am not sure if there is something explanatory going on–though I am also not sure that there isn’t.
    Maybe one can get something explanatory out of this by positing the following theory of truth: p is true iff p is known. Or maybe: p is true iff p is known by God. If theism is correct, both of these theories of truth are simple and extensionally correct. (I am not endorsing this theory of truth.)

    October 23, 2011 — 11:25
  • Dianelos Georgoudis

    It seems to me that there are propositions that make sense only in one possible world. If there is no God in such a world then the truth value of such a proposition may not be known by anybody in that world, but still that proposition may be true in that world.
    Here is an example of such a proposition. Consider a possible world not dissimilar to ours, but in which no God exists, and where quantum phenomena are random. Suppose in that world somebody constructs a machine that uses a random generator based on quantum phenomena in order to print on a piece of paper a one digit number and then incinerates that piece of paper to ash before anybody has a look at it. Consider now the proposition p=”the number printed on that paper was 5”. Given the random source of the printing event it can only obtain in that one possible world, therefore p refers to an event that can only obtain in that one possible world, therefore p only makes sense in that one possible world – and in that world it is impossible to know whether p is true. But it may well be.

    October 23, 2011 — 12:42
  • P1: It is impossible for a human being to know that God exists, as this would undermine faith understood as belief in the absence of evidence.
    P2: Rutten’s argument purports to supply human beings with the knowledge that God exists.
    C: Rutten’s argument is flawed in some way.
    I would suggests that an agnostic who finds the above argument plausible has no reason to accept Rutten’s premise that all true p statements are knowable in some possible world, since on the agnostic’s view there is some possible world in which God does not exist, and so some possible worlds contain unknown p statements. Thus the first premise is only plausible after it has been tested against our prior belief or disbelief in God.
    In general, I’d be interested to know how you respond to the limits of reason arguments like I’ve offered above.

    October 23, 2011 — 12:57
  • Joshua:
    1. In Christian tradition, faith is not primarily concerned with the existence of God. Faith is faith in Christ. Scripture says that even demons believe that God exists–but it doesn’t help them.
    2. The notion of faith as belief in the absence of evidence is not typical (in fact, I don’t know of any case of it) among the Church Fathers or the great medieval theologians.
    3. And, besides, Emanuel’s argument only requires the logical possibility of someone knowing that God exists–it does not require that we be the ones who could know it.
    Dianelos:
    I don’t see why you think that the event only occurs in one possible world. It seems plausible that there is another possible world where the piece of paper is not annihilated.

    October 23, 2011 — 13:31
  • 1. Hebrews 11? Lots of non-Christ-related faith there, unless you think Abraham’s faith was faith in Christ. Yet Abrahamic faith is what the author of the epistle recommends.
    2. It’s strange to move from the modal argument to a historical argument or an argument from popularity. Many contemporary lay theists hold this view. Keirkegaard describes faith in this way. Kant’s theism looks a lot like this. As for the claim that “No GREAT medieval theologian” held this view, I worry that I will point at Augustine or Abelard, really anyone other than Aquinas, and you will say he is not truly great.
    3. If Rutten’s argument has true premises and a valid inference, then under epistemic closure we have knowledge that God exists.
    In general, we should only accept Rutten’s premise if we believe that there is a possible world in which a self-knowing God exists.

    October 23, 2011 — 14:15
  • Dianelos Georgoudis

    Alex,
    The described event only occurs in one possible world because it depends on a random source, and it is impossible to copy a random source. By the very nature of what “random source” means we know that that there are no two identical random sources, either within one world, or in two possible worlds. (Of course it may happen that the two random sources produce the same data.)
    Here’s a case in point: Consider two worlds, the world X I described above, and the world X* which is identical to X up to the printing event. Suppose further that in X* the piece of paper is not annihilated and reads 3. Given the random source in both worlds, the 3 produced in X* says nothing about which number was produced in X. Thus p on X* is false and knowably so, but p on X may be true albeit unknowably so. After all the piece of paper in X may have had a 5 printed on it before being annihilated.

    October 23, 2011 — 15:58
  • Mike Almeida

    Take the following metaphysical principle, connecting possible worlds, knowledge and truth: ‘If it is impossible to know that p, then p is necessarily false’
    The proposition expressed by ‘this proposition is unknown by anyone’ is not necessarily false, but it’s impossible for anyone to know it. It’s not necessarily false whether or not there are omniscient beings. So, I think the principle is false.

    October 23, 2011 — 16:09
  • Clayton Littlejohn

    “Take the following metaphysical principle, connecting possible worlds, knowledge and truth: ‘If it is impossible to know that p, then p is necessarily false’. This principle seems cogent. For, if a given proposition p could be true, then, plausibly, there is some possible world in which some subject knows that p is true. In other words, if in *all* possible worlds *all* subjects do not know that some proposition is true, then, plausibly, that is because that very proposition cannot in fact be true.”
    My initial reaction was quite different from yours. It seemed that propositions expressed by sentences of the form, ‘p, but nobody knows that p’ can be true, but there’s no world in which anybody would know that the propositions expressed are true.
    True, you can say that the example doesn’t work because you think there’s an omniscient being in every world, but doesn’t this get things backwards? If the principle is going to do any work in an argument for theism, shouldn’t it rest on principles that seem plausible even if we’re agnostic as to whether there are any omniscient beings?

    October 23, 2011 — 16:09
  • The result that the knowability principle (for any proposition p, if p, then possibly it is known that p) implies a quasi-omniscience principle (all truths are in fact known) [1] was credited to an anonymous referee by Fitch (1963). Thanks to the detective work of Joe Salerno (2009), we know that referee was Alonzo Church. I like to call the proof that yields the relevant result, the “Church-Fitch” proof, giving proper credit to Church as well. We now have access to Church’s original remarks on the issue in Church (2009).
    I should add that discussions of Church (2009) and (Fitch 1963) ordinarily treat the knowledge operator K as short for the locution ‘it is known by some finite being at some time that…’ (see for example Williamson (2000: 273)). This keeps us from worrying about gods.
    The above remarks aside, I’d like to make a comment about the argument above. There is independent motivation for typing the knowledge operator (see Paseau (2008)). Without getting into the formal details, typing the knowledge operator blocks the derivation of the quasi-omniscience principle from the knowability principle even given Rutten’s understanding of that operator (though see Hart (2009) who quantifies into the type-index to secure the result). So the Atheist can side step the entire issue by proferring motivation for typing.
    Christopher G. Weaver
    Rutgers University (New Brunswick)
    Rowan University
    ————-
    [1] The principle is qualified as “quasi”-omniscient because if you treat the knowledge operator as equivalent to the clause “it is known by some being at some time that…”, there is no guarantee that the propositions known are all known at the same temporal index. Every proposition may be known at different indices.
    ——————————-
    Church, A. (2009) ‘Referee Reports on Fitch’s ‘A Definition of Value’’ in J. Salerno (ed.) (2009): 13-20.
    Fitch, F.B. (1963) ‘A Logical Analysis of Some Value Concepts’, Journal of Symbolic Logic 28: 135-42.
    Hart, W.D. (2009) ‘Invincible Ignorance,’ in J. Salerno (ed.) (2009): 320-323.
    Paseau, A. (2008) ‘Fitch’s Argument and Typing Knowledge,’ in Notre Dame Journal of Formal Logic 49:2: 153-176.
    Salerno, J. (ed.), (2009) New Essays on the Knowability Paradox New York: Oxford University Press.
    Williamson, T.(2000) Knowledge and its Limits. New York, NY: Oxford University Press.

    October 23, 2011 — 17:01
  • Dianelos Georgoudis

    Joshua,
    Hebrews 11:1 defines faith as “the evidence of things not seen”, which sounds very different from your “belief at the absence of evidence”. Actually, the ancient Greek word “elegchos” has a stronger sense than the modern concept of “evidence”, and is closer to “examination”, or “proof”, or “conclusive evidence”.
    In any case, I don’t think that any of the Fathers or any of the Scholastics would agree with your suggestion that “it is impossible for a human being to know that God exists”.

    October 23, 2011 — 17:12
  • Dianelos, my citation of Hebrews 11 was intended as a counter to Alex’s claim that “Faith is faith in Christ.” That’s why I cited to the whole chapter and not just the first verse. That first verse is frequently cited as evidence of this kind of faith, but you’re right that ἐλέγχω has the legalistic sense of proving guilt or rebuking through refutation. But I think we can say that Abraham’s faith that Isaac will be spared is not faith in Christ, nor is it very much like a prosecutor’s closing arguments, and that is one of the major examples in Hebrews 11.

    October 23, 2011 — 20:27
  • But I think we can say that Abraham’s faith that Isaac will be spared is not faith in Christ
    Actually, as far as I can tell (from my admittedly non-comprehensive reading), this seems to be precisely how the Church Fathers and medieval scholastics read it; and, for that matter, the reading makes some sense of the chapter when read in light of 11:39-12:2. I suppose to some extent it depends on (1) how seriously one’s exegesis takes typology as a more-than-merely-literary phenomenon; and (2) how seriously one takes the doctrines of the Trinity and the Incarnation, since genuine faith in God would then be the same thing, even if not as precise, as faith in Christ. Regardless, anyone reading the chapter with broadly Middle Platonist or Neoplatonist eyes would at least not be as quick to dismiss the interpretation as you are; and, of course, much of Church history consists of exegetes reading it with broadly Middle Platonist or Neoplatonist eyes. (The book of Hebrews itself, for that matter, has features suggestive of Middle Platonist influence.)
    This is all side issue, of course. I’m a little puzzled, though, at your number (2) in the 2:15pm comment above, and feel I must be missing something. The issue had been, I take it, that one could reject P1 simply by denying the definition of faith on which it explicitly depends; and that accounts of faith that would require such a denial have been common and are very respectable. I don’t quite see what your counter-response does to address this. Even if the opposite is as common as you say, it would still leave the response intact: subcontrariety is not contradiction, so it would still be, apparently, common and respectable to reject (P1) without too much worry. (Like Alex I’m very skeptical of the notion that it is at all so common; in every specific case I’ve come across, even in fideist cases, the person in question turned out to mean by ‘evidence’ ‘evidence of such-and-such kind’, as specified by context. But even if we assume this is wrong, I don’t see your response would have any force against someone saying, “Look, there’s plenty of evidence that one can, in an entirely non-ad-hoc and non-arbitrary way, simply deny P1” or even “One can simply claim that P1 is not at all obvious and in fact easily controvertible”.) So what am I missing?

    October 23, 2011 — 21:17
  • Joshua:
    I, too, think that Abraham’s faith was implicitly in Christ, but I don’t want to insist on that in this thread.
    I know that they were just examples, but Augustine gave arguments for the existence of God while Abelard “insist[ed] that not merely the existence of God, but also his nature as a trinity of power, wisdom and love, is and has been knowable to all through reason and through the testimony of the created universe” (John Marenbon, The Philosophy of Peter Abelard).
    Paul in Romans 1, probably basing himself in part on Wisdom 13, insists that God is knowable through the things he has created. The Catholic Church in turn formally taught in the First Vatican Council that the existence of God can be known by reason.
    Granted, some contemporary Protestants think that the existence of God cannot be known but must be taken on faith. That was not the position of Luther and Calvin. Luther thought that there was a “legal knowledge” of God that philosophers had, and Calvin thought we all had a sense of divinity.
    Besides, it is simply a conceit of modern secular society that faith is belief in the absence of evidence or knowledge. That is not how Scripture sees faith. The New Testament constantly talks of us knowing God, the grace of Jesus Christ, and all sorts of things that are the content of faith. In Scripture, faith and knowledge are quite compatible. What may not be compatible is faith and vision, or direct apprehension of the truth. In fact, I think the way to right distinction in a Christian context is between knowing naturally and knowing by faith: but both are species of knowledge.
    Aristotle in the Rhetoric defines “pistis” (“faith”) as a persuasion by means of the character of the speaker. In the New Testament, “faith” has two aspects: there is the aspect of entrusting oneself to Christ and the aspect of believing. The belief aspect fits very well with what Aristotle says: what we believe by faith is that which we believe on the basis of the perfect character of God.
    Belief on the basis of another’s character can certainly be knowledge. A friend tells me something. She’s got the sort of character that I can’t imagine her saying it unless she knew it. I believe her. That’s “faith”, but it’s also a species of knowledge.
    If one takes faith to be unevidenced belief that isn’t knowledge, then the Christian view of faith as a virtue becomes a straw man.
    p.s. Closure is more complicated than you make it sound. First of all, for your point to go through, the premises need to be known and not merely true. Second, there can be dwindling probabilities. P may have sufficient probability for knowledge, Q may have sufficient probability for knowledge, but their conjunction need not. And even an argument that doesn’t give knowledge can be useful by raising probabilities.

    October 24, 2011 — 9:09
  • Aaron

    At first blush, I see no reason to accept the metaphysical principle Rutten employs. That aside, in the third paragraph Rutten writes:
    ‘It seems reasonable to hold that it is impossible to know that God does not exist. For, whatever the arguments against God, there will always be some (perhaps an extremely remote) possibility that God does exist after all, so that we can never truly say, on the Cartesian view, that we know that God does not exist.’
    This does not seem correct at all. Now, I take it that the operative notion of possibility used here in the quoted above is epistemic possibility. Hence, I see no difficulty with conceiving of a possible world in which someone constructs a proof which shows that, for any conception of an omni-type deity, G, G contains a contradiction- surely this is epistemically possible. Thus, it is conceivable that there is a possible world in which ‘God does not exist’ can be known.
    Of course, the opposite is true. It is epistemically possible that, in every possible world in which there is a proof that purports to show that G entails a contradiction, it is not the case that G entails a contradiction, but rather that G is true. From these epistemically possible states of affairs, however, we do not want to infer that ‘God exists’ or ‘God does not exist’ is necessarily true. To do so would be to assume what is at issue, namely, whether ‘God exists’ is necessarily true or false. Rutten’s argument does not progress the matter at all.

    October 24, 2011 — 9:28
  • Tim Pawl

    Two points, one of which already raised by Mike A. and Clayton L.:
    1. Rutten says, “Take the following metaphysical principle, connecting possible worlds, knowledge and truth: ‘If it is impossible to know that p, then p is necessarily false’.”
    Mike and Clayton mention propositions of the form “P, but nobody knows P” as counterexamples to this principle. Here’s another example of the same sort. Most non-theists will think that it is possible that there are no knowers. Rutten finds his principle cogent; the non-theist will no doubt find the possibility of no knowers cogent as well. Name one of the worlds in which there are no knowers W1. “That W1 is actual” is a possible truth. But it is not knowable–for the only world in which that proposition is true is a world without things to know it.
    2. Consider the third paragraph of this post. There Rutten argues that, “It seems reasonable to hold that it is impossible to know that God does not exist. For, whatever the arguments against God, there will always be some (perhaps an extremely remote) possibility that God does exist after all, so that we can never truly say, on the Cartesian view, that we know that God does not exist. But then it follows that it is necessarily false that God does not exist. Hence, it is necessarily true that God exists. The principle thus entails theism.” Consider a parody of this argumentation, drawn from van Inwagen’s Mr. Know-No. Call a man who knows that God does not exist Mr. Know-No. Can we know that Mr. Know-No could not exist? Well (now starts the parody), whatever the arguments against the possibility of Mr. Know-No’s existence, there will always be some (perhaps an extremely remote) possibility that Mr. Know-No could exist, after all, so that we can never truly say, on the Cartesian view, that we know that Mr. Know-No could not exist. But then it follows that it is necessarily false that Mr. Know-No could not exist. Hence, it is necessary true that Mr. Know-No could exist. So Mr. Know-No’s existence is possible. So he exists in some world. So, in that world, God does not exist (since knowledge entails truth, and Mr. Know-No knows God doesn’t exist). Since God is supposed to be a necessary being, and he doesn’t exist in some world, he doesn’t exist in any world. So God is impossible.

    October 24, 2011 — 10:15
  • Marc

    I don’t know whether the following objection has any merit, but I thought I’d suggest it nonetheless.
    Let T represent the aforementioned metaphysical principle on which Rutten’s argument is based: “If it is impossible to know that p, then p is necessarily false.” It seems to me that if we accept the Cartesian view of knowledge, one of the consequences is that we’re prevented from saying we know that T is true. This, by itself, might not necessarily be problematic for the argument. However, another one of the consequences is that we appear to undermine any good reasons we might have for affirming that T is even plausibly true. Consider the assortment of intractable epistemic difficulties we immediately encounter if we hold that certainty is a necessary condition for knowledge. For example, several of our (epistemically important) properly basic beliefs don’t constitute pieces of knowledge. So, if we’re forced to abandon even these beliefs, then it seems that our epistemic position is such that we can’t rationally construct or accept an argument for theism, such as the one in question.

    October 24, 2011 — 13:08
  • Brandon: One can certainly deny that definition of faith! After all, that’s how we proceed with Rutten, isn’t it? The reason his argument fails is that he has the wrong account of truth. If my argument fails because it has the wrong account of faith, then it fails! My point was simply that we ought not to deny it simply because it wasn’t held by an ad hoc assortment of theologians.
    Alexander: Thanks! That’s a great set of responses, though I disagree that Augustine’s arguments are intended to be dispositive for atheists (“It is certainly perverse and preposterous to desire to see the truth so that you may purify your soul, which should rather be purified that you may see.”) I see this as a kind of practical syllogism that seems to motivate arguments in the philosophy of religion, even though no one explicitly enunciates it. I disagree that it somehow denigrates or “straw mans” faith, and personally have a great deal of respect for the philosophically-rigorous fideists I know.
    I’m sorry to bring this up in a thread on Rutten’s argument: my only defense is that his account of knowability reminded me of a lingering “unknowability” question.

    October 24, 2011 — 15:50
  • Yeah, I think the Cartesianism is a major weakness. You could try to boost the principle: “p is possible iff p can be known with certainty”. But now the principle is much less plausible.
    Here’s a somewhat different worry. If I were an atheist, I would worry whether it’s even possible for an epistemic agent to have certain knowledge of empirical matters.
    A way to get out of these problems would be to use an ordinary notion of knowledge instead of the Cartesian one, but argue that no one could know that there is no God, and then aim the argument not at the atheist (who may well think she knows that there is no God), but at the agnostic, who might be thought to be receptive. But that would be hard. The agnostic may well insist that if there is no necessarily existing God, there will be worlds which are so utterly horrible that it’s easy for agents in them to know that there is no God. (Possible response to agnostic: in worlds that are so horrible, the agents shouldn’t trust their moral or metaphysical intuitions, and hence can’t get knowledge from the argument from evil?)

    October 25, 2011 — 8:49
  • Thank you all for your interesting critical comments to my argument. Highly appreciated! Due to a course that I’m following this week I have unfortunately not found time yet to respond. Upcoming weekend I will have sufficient time to respond to the various proposed objections.

    October 25, 2011 — 10:53
  • CliveStaples

    Could it be possible to construct a kind of Godel sentence p? Say, p := “Nobody (or, no accurate knower) knows p to be true.” Might such a sentence be true, but necessarily unknowable?

    October 26, 2011 — 2:20
  • Today I found time to respond. The first objection I would like to respond to is Marc’s objection that my principle cannot be known to be true if we accept the cartesian view of knowledge. I believe this objection is not convincing. For, I’m definitely not claiming to know that the principle is true. I’m just claiming that the principle (rendered a bit differently, as I will explain in what follows) is sufficiently plausible or reasonable to accept, so that we are justified to use it as one of the premises of a reasonable (thus not necessarily conclusive) argument for God’s existence.
    Let me continue with Aaron’s objection that the concept of God is possibly self-contradictory (and thus, I would add, contradictory in all possible worlds if we, plausibly, take it that the fundamental laws of logic are true of all possible worlds). I believe that this objection is not convincing. Consider for example the following conception of God: ‘An immaterial uncaused person that is the direct or indirect originating cause of all other concrete particulars’. This definition, I would argue, is logically consistent. And there are many similar examples. So, in order for the objection to have force one would at least have to suggest some sketch of a proof that the concept of God is logically contradictory.
    Clayton has it that my argument get things backwards because the principle I adopt seems only plausible if we already accept that there are omniscient beings. His objection amounts to the worry that my argument begs the question since I appeal to God’s existence to show that God exists. But, it seems to me that I’m not appealing to God’s existence. What I did to refute the third objection of the atheist was only to appeal to the mere logical possibility that God exists, and subsequently to derive from this possibility that it is not logically impossible to know that God exists (since, on my appeal, there is a possible world within which God knows that God exists). Now, my appeal to the mere logical possibility of there being a God is surely not the same as assuming that God actually exists. Moreover, my argument does not rely on the claim that God is by definition a necessary being, so my appeal to the mere logical possibility of God’s existence does not reduce my argument to the ontological argument.
    The principle on which my argument is based is that every proposition that is possibly true is also possibly known. A number of counter examples to this principle have been proposed. I do not believe that the counter example of Dianelos (“randomly printing a number on a piece of paper and then annihilating that a piece of paper”) is convincing since I agree with Alexander that, plausibly, there is another possible world where the piece of paper is not annihilated. Indeed, if we understand possible worlds as epistemic alternatives, so that there is a possible world in which Obama (taken ‘de re’) would not have won the elections, then there is also a possible world in which *this very piece of paper* (taken ‘de re’) would not have been destroyed.
    The counter examples proposed by Mike (consider “this proposition is unknown by anyone”) and Tim (“Let W1 be a world in which there are no knowers and consider “That W1 is actual”) are surely interesting. One could for example also suggest the following counter example: “There are no known propositions”. Yet, these counter examples seem a bit a-typical or artificial. They appear to be ‘loophole’ cases. Therefore I believe there is an adequate way to avoid them by opting for a slightly different, more realistic, rendering of my argument. For that I need three definitions.
    A state of affairs exists within one or more possible worlds. First, let a ‘c-state’ be a state of affairs of one or more concrete particulars, having each zero or more properties, and standing to each other in zero or more relationships. Second, let a ‘c-proposition’ be a proposition that either affirms or denies there being some c-state, such as for example “The car of Peter is blue”, “Eva is a friend of James”, “There are no horses”, “Jim does not know anything”, “Linda knows that snow is white”, “God exists” or “God does not exist”.
    Third, let a K-world be a possible world in which at least something is known. Now, only subjects (i.e., agents or persons) can know things, thus a K-world is a world that contains one or more subjects. And, vice versa, a world that contains at least one subject is also a K-world. For, every subject knows at least that he or she exists. Beside, according to the negative introspection axiom of S5 epistemic modal logic, if a subject S does not know p, then S knows that S does not know p. So, indeed, every subject knows at least something.
    Given these definitions, my principle can be alternatively rendered in the following way: ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’. I would say that this principle is quite similar to the original one, except that it is more modest. For, it only applies to c-propositions, and its antecedent now also requires that there must be a K-world in which p is true. Moreover, it seems more realistic than the original rendering. For the proposition p in the antecedent is quite basic in the sense that it is about concrete particulars. Besides, p is already required to be true in at least one world that contains one or more subjects that know things. So, in a sense, p is “closer to” the possibility of there being a subject that knows p.
    Well, the aforementioned three counterexamples do not apply to my alternative rendering of the principle. For, Mike’s proposition is not a c-proposition, and the other two, “That W1 is actual” and “There are no known propositions”, cannot be true in a K-world.
    Now, let p be the c-proposition that God does not exist. As argued before, and taken my response to Aaron into account, there is no world in which p is known. Hence it follows that there is no K-world in which p is true. Therefore God exists in all K-worlds, including ours.
    Note that the proposed alternative rendering of the principle comes with a price. It now no longer follows that God is a necessary being. After all, for all we know there might be one or more non-K-worlds, and in those worlds God (being a subject) does not exist. But still, it follows that God exists in all K-worlds, and thus also in our world.
    And, moreover, we could opt for an even weaker version of the principle. Let us consider the world that we actually inhabit. Now, we might only hold as a principle that at least everything that is true of our world is possibly knowable. That is, for every truth of our world there is some possible world within which there is a subject that knows that truth. But then, by parallel reasoning, it follows that God exists in our world, either necessarily or as a brute fact.

    October 26, 2011 — 11:14
  • A doubt that I have is:
    Is there only one God in all worlds combined, or is there a God in each of the worlds. (Multiple Gods in a world does not change the argument.)
    If there is only one God in all worlds combined, there are no issues. That God might be in our world, or might not be in our world. Status Quo.
    But if there is a God in each of the world’s there is a problem. Take any two Gods. Take the only property God has being omniscient. God A knows that God B exists. God A knows that God B knows that God A exists …ad infinitum. So there is mathematically no end to what God A must know to be omniscient. Hence two omnicient beings are not in being. Hence at max only one God can exist. Back to single God paragraph above.
    This problem does not arise with a single God. That one God knows how much any individual knows.

    October 26, 2011 — 15:01
  • Clayton Littlejohn

    “Given these definitions, my principle can be alternatively rendered in the following way: ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’. I would say that this principle is quite similar to the original one, except that it is more modest. For, it only applies to c-propositions, and its antecedent now also requires that there must be a K-world in which p is true.”
    Hi Emanuel (if I may),
    Just a quick follow up. I don’t see why the proposition expressed by “Nobody knows p” couldn’t be a c-proposition if the proposition expressed by “p” is a c-proposition. If both propositions are c-propositions, why wouldn’t the proposition expressed by “p, but nobody knows p” be a c-proposition? If the proposition expressed is a c-proposition, don’t we have a counterexample to your modified and more modest principle? It’s contingent whether the principle is true, but there’s no world in which it’s known.
    If the proposition associated with Moorean absurdities are contingent and the proposition expressed by my Moorean absurdity is unknowable (either because it’s false or because if true, not known), I don’t see how the more modest principle deals with the objection.

    October 27, 2011 — 7:05
  • Joshua Rasmussen

    Rutten,
    Clever argument. I wonder, though, if a parity argument could be given. Consider that the claim that it is logically possible that God does not exist is surely about as modest as the claim that it is logically possible that God exists (a claim you appeal to in response to objection 3). Yet, if your reasoning is correct, then these two claims are mutually inconsistent (even without assuming that God is by definition a necessary being).* Thus, perhaps we should be agnostic about the logical possibility of God prior to further data.
    *Suppose it is logically possible that God does not exist. Then if what can’t be known is impossible, then it is in fact possible to know that God does not exist–even if we doesn’t know how one might know that. But then from the Fitch principle that whatever can be known is known, it follows that God does not exist.

    October 27, 2011 — 15:05
  • Hi Clayton, your proposition “p, but nobody knows p” would be a c-proposition only if it either affirms or denies there being some c-state. But what would be the c-state affirmed or denied by “p, but nobody knows p”? It seems to me that there is no c-state X such that “p, but nobody knows p” either affirms X or denies X. Note that X would have to be one or more concrete particulars having certain properties and standing in certain relationships. Perhaps you would respond that “p, but nobody knows p” both affirms the c-state expressed by “p” and denies the c-state expressed by “There is somebody that knows p”. Yet, this is not the same as either affirming or denying some given c-state.

    October 28, 2011 — 3:51
  • Clayton Littlejohn

    Hi Emanuel,
    One quick follow up. I thought that for some c-proposition, p, there’s another c-proposition: that somebody knows p. The negation of that second c-proposition is nobody knows p. (Given your definition).
    Initially I thought that you thought that the propositions expressed by “Nobody knows p” and “Somebody knows p” aren’t c-propositions, but I see that the proposition that nothing is a horse is a c-proposition as is the proposition that something is a horse (given your example and your negation rule). So, I think you have to grant that the conjuncts p and nobody knows p are c-propositions.
    It now seems that your examples of c-propositions are misleading or you block the counterexample by denying the conjunction rule:
    CR: If p and q are c-propositions, the conjunction of p and q are c-propositions.
    If you accept CR, I think I’m sticking by your description of c-propositions and can say that the proposition expressed by “p, but nobody knows p” is a c-proposition.
    It seems plausible that this is a world in which there’s some p that nobody knows. So, the proposition expressed by “p, but nobody knows p” is true in the actual world. The actual world is a K-world. There’s no world in which anybody knows that the proposition expressed by “p, but nobody knows p” is true. So, counterexample? (If not, where does everything go wrong? The conjunction rule?)
    You wrote: “your proposition “p, but nobody knows p” would be a c-proposition only if it either affirms or denies there being some c-state. But what would be the c-state affirmed or denied by “p, but nobody knows p”?”
    I think my answer is just this. It affirms two c-states ((i)p and (ii) that nobody knows that p). If it affirms two c-states, it affirms at least one c-state.

    October 28, 2011 — 9:15
  • Hi Clayton, I grant that, for many p, the conjuncts “p” and “nobody knows p” are c-propositions. For, indeed, “p” may *affirm* some c-state, and “nobody knows p” *denies* the c-state expressed by “there is somebody that knows p”.
    (Note that “nobody knows p” does not affirm some c-state, since c-states are defined “positively”, that is to say, a c-state is one or more concrete particulars having zero or more properties and standing in zero or more mutual relationships.)
    However, I do not accept CR. Why? Well, every c-proposition either affirms or denies some single c-state. But the propostion “p, and nobody knows p” does not affirm or deny some single c-state. (After all, its first conjunct affirms some c-state, while its second conjunct denies some other c-state, and this does not amount to affirming or denying some single c-state.) Hence, CR fails, and therefore your counter example does as well.

    October 28, 2011 — 10:27
  • Hi Joshua, let me start by saying that my argument does not appeal to the claim that it is logically possible that God does not exist. All I need for my argument is the claim that it is logically possible that God exists.
    However, let us grant that, precisely because I claim that it is logically possible that God exists, I should also affirm that it is logically possible that God does not exist. So, here we go, I affirm that there is a logically possible world in which God does not exist. What does follow from this?
    Now, as I mentioned above in my first response, the weakest alternative rendering of my principle has it that for every true c-proposition p of *our world* (taken ‘de re’) there is a logically possible world within which p is known. The second premise of my argument has it that it is impossible to know that God does not exist. From these two premises it follows *only* that God exists in our world. But then, indeed, there might be many other logically possible worlds in which God does not exist. Therefore, the claim that it is logically possible that God does not exist does not conflict with the premises or conclusion of the weakest alternative rendering of my argument.

    October 28, 2011 — 13:08
  • Clayton Littlejohn

    Are these c-propositions?
    1. God loves my kindness.
    2. God understands my socialism.
    How about:
    3. God hates my atheism.

    October 28, 2011 — 14:08
  • Dianelos Georgoudis

    Emanuel,
    Given that the piece of paper was printed by a random source there is not another possible world in which this very piece of paper (taken “de re”) exists, simply because there is not another possible world with a random source that *will* produce identical effects.
    For each possible world where p turns out to be true, there will be 9 possible worlds where p turns out to false. So whatever another person in these other worlds may observe will not warrant knowledge about the truth value of the c-proposition p in the world in which it is defined.

    October 30, 2011 — 16:25
  • John Alexander

    Emanuel: What about a possible world that contains beings that know many things, but it is not possible that they know everything that is knowable? It seems to be possible that God could create a world with such beings and then make it such that He cannot know what occurs on it. Maybe He eliminates the memory of creating this world and makes it such that it cannot be discovered even by Him. In such a world there would be propositions that it is impossible for the beings in that world to ever know, but they are not false because they are knowable in the same way there are propositions in this world that are unknown at the present but may become known. If this is possible then the argument for theism fails because it is possible to have propositions that it is impossible to know, but are not false.

    October 31, 2011 — 1:15
  • Dianelos: If there is a possible world in which *this* very man, Nixon, has not lost the elections (an example of Kripke), then there is also a possible world in which *this* very piece of paper, randomly produced, is not destroyed. Your response does not indicate why this would be problematic.
    John: You write that “[i]n such a world there would be propositions that it is impossible for the beings in that world to ever know”. Well, perhaps, but in any case we can posit some other possible world in which these propositions are in fact known. Hence, there is no problem for my argument here.

    October 31, 2011 — 14:50
  • John Alexander

    Emanuel: I take it that propositions are fact specific therefore world specific (even though different PW could contain the similar facts). Therefore, we can posit a world that has specific propositions that apply to that world only. It would seem that PW is not relevant to the issue of whether the metaphysical thesis is true or false. That being the case there can be propositions that it is impossible to know, but are not false.

    November 1, 2011 — 8:35
  • John: Propositions do not have to be world specific. A proposition, for example “Mark has a blue car”, might be true in many possible worlds, and false in many others.

    November 1, 2011 — 14:47
  • Emanuel:
    Suppose “God” is short for a definite description, like “the omniscient, omnipotent and all-good creator of all contingent beings [OOAGC].” Then “God exists” isn’t a c-proposition, since it doesn’t affirm the existence of a particular being, but should be analyzed as: Ex(OOAGC(x)).
    Suppose “God” is a proper name. Then the atheist may not grant that “God exists” expresses a proposition. After all, it’s plausible that “Thor exists” doesn’t express a proposition, there not being any way to anchor “Thor” to a particular individual.

    November 1, 2011 — 16:33
  • Dianelos Georgoudis

    Emanuel,
    Please consider the following sets of propositions.
    1. There are possible worlds in which X is false.
    2. Therefore, it is possible that X is false.
    3. But in fact (i.e. in the actual world) X is true.
    4. Therefore, X is true but not necessarily true.
    If we use X=”Nikon won the 1968 election” we get:
    A1. There are possible worlds in which Nixon did not win the 1968 election.
    A2. Therefore, it is possible that Nixon would not win the 1968 election.
    A3. But in fact (i.e. in the actual world) Nixon won the 1968 election.
    A4. Therefore, the proposition “Nixon won the 1968 election” is true but not necessarily true.
    Let p be the proposition I defined above. If we use X=”nobody can know that p” or “it is impossible to know that p” we get:
    B1. There are possible worlds in which somebody can know that p.
    B2. Therefore, it is possible that somebody could know that p.
    B3. But in fact (i.e. in the actual world) nobody can know that p.
    B4. Therefore, the proposition “nobody can know that p” is true but not necessarily true.
    or in other words: The proposition “it is impossible to know that p” is true but not necessarily true.
    Now the epistemic principle you suggested is: “If it is impossible to know that p, then p is necessarily false”. It seems then we have here a counterexample to your principle, because it may well be the case that p is true, even though it is in fact impossible to know that p. Thus, perhaps your principle should read: “If it is necessarily impossible to know that p, then p is false”. A more concise way to put it may be this: “If it is necessarily impossible to know the truth value of p, then p has no truth value (i.e. is meaningless)”.

    November 2, 2011 — 6:11
  • John Alexander

    Emanuel: That is what I said. Given that propositions relate to facts – the way things are – it is possible that there is a fact that no one can possible know to be false(there is a diamond buried under 1000 feet of granite that no one will ever find), but it is not necessary that the corresponding proposition is false. Also, it seem plausible to maintain that there is a PW where its members know that God does not exist. The fact that we may not know this does not warrant the conclusion that no one does not know it in all PWs. If God is actually Descartes’ Evil Demon then the ED knows that God, as defined by the theist, does not exist.

    November 2, 2011 — 8:38
  • Tim Pawl

    Emanuel,
    Two thoughts:
    1. You write: “Beside, according to the negative introspection axiom of S5 epistemic modal logic, if a subject S does not know p, then S knows that S does not know p.”
    That conditional in the above quotation is false in cases where S believes he knows p and is justified in his belief that p. Consider: I believe that my wife is at home now. And I have justification to believe it. She told me she would be there, and she is a very reliable source on such matters. But, it turns out, unbeknownst to me, that she is not at home. She’s at the hospital instead, since she needed to run there for some emergency. So the antecedent of the conditional is true: I don’t know that she is at home (since it is false that she is at home). It follows by the above conditional that I know that I do not know that she is at the store. But I know no such thing! I don’t even believe such a thing.
    2. What do you make of the Mr. Know-no objection I gave earlier?

    November 2, 2011 — 10:11
  • Tim Pawl

    Oops, I changed examples in writing my above comment but forgot to change all the instances where the old example occurred. the phrase “the store” in the third-to-last sentence in point 1 should be “home” instead. The first point should end like this:
    So the antecedent of the conditional is true: I don’t know that she is at home (since it is false that she is at home). It follows by the above conditional that I know that I do not know that she is at *HOME*. But I know no such thing! I don’t even believe such a thing.

    November 2, 2011 — 10:39
  • John Alexander

    Emanuel: you wrote, “For, if God does not exist, then, plausibly, there is no afterlife.” Why is it plausible to think that if there is an afterlife that that there is a God? It seems to me that the question of God’s existence and the existence of an afterlife are independent questions such that one can assert that there is an afterlife, but also assert there is no God. Or, maybe more resonably, one can assert that he or she believes in an afterlife while not believing that there is a God, or at least a God as defined by the theist. It seems plausible that even if there is a God that created the world, including an afterlife, that the nature of this being cannot be infered from what it creates, only that it has the knowledge and power to create what it did. If we simply mean that a belief that is plausible is one that is consistent with other beliefs held to be true, then theism meets this criterion, but so do many non-theisitc interpretaitons, i.e., Descartes’ Evil Demon. This does not seem to get us anywhere.
    Also, can you define how you are using ‘truth’ and ‘knowledge?’

    November 2, 2011 — 12:45
  • Dianelos: When I write that it is impossible to know that p, then, as should be clear from the first paragraph of my original presentation of the argument, I do in fact mean that there is no possible world in which p is known, that is to say, I do in fact mean that it is necessarily true (true in all possible worlds) that p is unknown.
    John: Your write: “Also, it seem plausible to maintain that there is a [possible world] where its members know that God does not exist”. Now, I do not think that the example of such a world you provide is convincing. For, how could some Evil Demon deceiving human beings know that God does not exist? Also, there seem to be many possible worlds in which it is in fact known that the diamond buried under 1000 feet of granite exists. Take for example a possible world in which some human being (or perhaps some extraterrestrial intelligent being) observes the diamond by using advanced technical equipment. Now, you might respond that the fact that no one will ever find the diamond is part of the state of affairs in question. However, in that case your state of affairs is incoherent, since, as I pointed out, there will be in fact many possible worlds in which a diamond buried at 1000 feet will be discovered.

    November 2, 2011 — 12:57
  • Alexander: Let T be some type of concrete particular, such as ‘conscious being’ or ‘material thing’. Now, I take it that the fact that there is a T is a proper example of a concrete state of affairs, that is to say, a c-state (although I should reformulate my definition of c-state above to make this clear). But then the fact that there is an omniscient, omnipotent and all-good creator of all contingent beings, is also a c-state, so that Ex(OOAGC(x)), affirming this c-state, is a c-proposition.

    November 2, 2011 — 16:14
  • Tim: I agree that the conditional in the quotation is false for the case you specify. Thanks for pointing this out! Yet, it does not impact my argument, since, in the relevant context I would still argue that every subject knows at least that he or she exists. Regarding Mr Know-no, I would say that we can specify the reason for the impossibility of such an agent. For, the basic concept of God is not contradictory, so that Mr Know-no cannot know that God doesn’t exist by logical proof. Neither is it possible to exclude God’s existence by testimony, observation or intuition. Thus, there cannot be a Mr Know-no. Now, you might respond that in that case it’s also impossible to know that, say, superman exists, so that it would follow that superman does in fact exist! To that I would reply that it is not impossible to know that superman does not exist. Take for example a possible world in which God exists but in which God is alone because He decided not to create anything. But then that world’s God will in fact know that superman does not exist.

    November 2, 2011 — 17:07
  • Oops, the fourth-to-last sentence of my previous post should have been: “Now, you might respond that in that case it’s also impossible to know that, say, superman does not exist, so that it would follow that superman does in fact exist!

    November 3, 2011 — 1:30
  • Emanuel:
    “Let T be some type of concrete particular, such as ‘conscious being’ or ‘material thing'”
    Here’s a type of concrete particular: a photon such that nobody knows any proposition that entails that there are photons. The atheist will say that possibly there is an individual of this type.

    November 3, 2011 — 8:41
  • Tim Pawl

    Emanuel,
    Thanks for the response. I agree that the counterexample to the conditional I give in my first point does not impact your overall argument.
    Concerning Mr. Know-no, your original argumentation, the argumentation that I parodied, had it that no matter what the arguments against God’s existence are, there might still be an extremely remote possibility that he exist. I thought, likewise, the same would be true for Mr. Know-no. No matter what arguments we have against his existence, we might still get it wrong. You say, for instance, that Mr. Know-no cannot exist b/c it is impossible for one to learn that God does not exist by intuition or testimony or observation. But isn’t there an extremely remote chance that one could? And if not, why is there a remote chance that someone’s arguments against God’s existence could fail, but no remote chance that someone’s arguments against Mr. Know-no could fail? Furthermore, what would these arguments against Mr. Know-no’s existence be that have premises with no chance, no matter how remote, of being false?
    Thanks,
    Tim

    November 3, 2011 — 12:54
  • Tim: Let W be a possible world for which it is true that God exists. Now, surely, in W, for God the belief that God exists can be taken to be self-evident or incorrigible, and thus an instance of knowledge under the Cartesian view of knowledge. Yet, there is no agent for which the belief that God does not exist is self-evident or incorrigible. Therefore, there is no possible world in which it is known that God does not exist. But then a Mr. Know-no is not possible. Now, to refute this response you would have to at least indicate how the belief that God does not exist could be self-evident or incorrigible for some agent in some possible world.

    November 3, 2011 — 14:07
  • Alexander: A photon such that nobody knows any proposition that entails that there are photons does not seem to be a type of concrete particular. For, the fact that nobody knows any proposition that entails that there are photons is not a property of the photon itself. Now, surely, there is a possible world in which a photon exists and in which nobody knows any proposition that entails that there are photons. But then, there are also other possible worlds in which *these* photons are in fact known, so that there is no problem for my argument.

    November 3, 2011 — 15:04
  • Clayton Littlejohn

    Hi Emmanuel,
    Three quick points about the argument.
    (1) I think that “God understands my atheism” expresses a c proposition if “God understands my socialism” does. Since “God understands my socialism” seems to express a c-proposition, I don’t think rejecting CR will save your principle.
    (2) Even if you restrict the principle and rule out my latest case as well, I wanted to address your question about whether the argument is convincing. If this wasn’t a philosophy of religion blog and somebody asked me what I thought about this principle:
    (*) If it is impossible to know that p, then p is necessarily false
    I would say that I didn’t think it was all that plausible. Why not? Because if the propositions concern claims about, say, the number of moments in the universe, whether my phone number occurs in the decimal expansion of pi an even number of times, what the minimal supervenience base is for my obligation to feed my pet dog, and things like that, I wouldn’t think that any possible mind could sort these things out. That’s because I don’t think it’s metaphysically possible for there to be an omniscient being. Notice that your response to these concerns is to say that there’s a possible world in which God could know the relevant propositions. I don’t know many atheists who would grant you two assumptions you need to block the (putative) counterexamples and complete your proof:
    (A1) There’s a possible world in which there’s an omniscient being;
    (A2) God is actual if possible.
    Without these assumptions, it’s hard to see how your principle could be both counterexample-free and dialectically effective in constructing a proof of God’s actual existence. With these assumptions, why don’t you just argue that since there’s a possible world in which God exists, God actually exists?
    (3) You work with a Cartesian conception of knowledge and I’m curious to know why this is. On the Cartesian conception, mere hyperbolic doubt destroys knowledge. On the ordinary conception, it doesn’t. Let’s introduce two expressions, “c-knows” and “o-knows” to pick out these two knowledge relations. You’ve argued that the atheist can’t c-know that God doesn’t exist. Let’s grant that. That’s consistent with the further claim that the atheist o-knows that God doesn’t exist (say, on the basis of the argument from evil).
    Let’s now distinguish two readings of your principle:
    (o*) If a given proposition p could be true, there is some possible world in which some subject o-knows that p is true.
    (c*) If a given proposition p could be true, there is some possible world in which some subject c-knows that p is true.
    I still think there are counterexamples to (o*) of the sort sketched above, but let’s assume you’ve handled them somehow. The atheist should think that she knows that God doesn’t exist, say, on the basis of the argument from evil. Given (A2), the atheist should conclude that there’s no possible world in which God exists. So, the atheist should conclude that the proposition that God doesn’t exist gives us a counterexample to (c*). Indeed, she o-knows that it’s a counterexample (even if she doesn’t c-know it).
    If you can show that this counterexample fails, you have to show that the atheist cannot o-know that God doesn’t exist or because you think you can show that o-knowledge just is c-knowledge. You might be able to do this, but I don’t think you can just assume at the outset that the atheist won’t take herself to o-know that God doesn’t exist.

    November 3, 2011 — 19:15
  • Clayton:
    1. Accepting “God understands John’s atheism” as c-proposition does not result in a problem for my argument. For, there is a possible world in which God exists and knows that He understands John’s atheism. Besides, God might understand why John is an atheist, even if John doesn’t know that God does not exist.
    2. I do not appeal to (A2) for my argument. For, as I said earlier, ‘existing necessarily’ is not part of the definition of God that I employ for the argument. Further, (A1) seems to be a sufficiently reasonable assumption, not more unreasonable than the assumption that it is possible that God exists. (As I mentioned in my original presentation of the argument, I recently read an excellent refutation of Dennis Whitcomb’s argument that omniscience is impossible.)
    3. The Cartesian conception of knowledge I use can be rendered as being an internalistic account of epistemic foundationalism for which the set of proper basic beliefs is restricted to, say, propositions known by logical proof or direct self-evident incorrigible intuition. Hence, on the Cartesian conception, mere hyperbolic doubt does not destroy all instances of knowledge. For, on the Cartesian view, even under hyperbolic doubt, there are proper instances of knowledge, such as “I exist”, “I’m having at this very moment the experience of seeing red” or “1+1=2”.
    4. It might be justified to say that in some possible worlds an atheist o-knows that God doesn’t exist. But there is no possible world in which somebody c-knows that God doesn’t exist, and that’s all I need for my argument.
    In general I would like to point out that there is a substantial, and epistemically relevant, difference between S asserting that S exists (e.g. God asserting that God exists), and S asserting that some other subject external to S, say U, does not exist (e.g. John asserting that God does not exist). For, in the former case S has direct access to its own mental states, whereas, in the second case, S does not have direct access to U, or to U’s mental states. Thus, while the former case, S asserting that S exists, is properly said to be maximally epistemic ideal, the second case is not maximally epistemically ideal. So, on the Cartesian view, we are justified to hold that the former case is an example of a proper basic belief, while the second case isn’t.

    November 4, 2011 — 3:12
  • Dianelos Georgoudis

    Emanuel,
    So I take your argument is about necessarily unknowable propositions. Now it seems to me that necessarily unknowable propositions are strange beasts indeed, which I think are intrinsically meaningless (have no truth value). If I am right there is a version of your argument that would go like this:
    1. If p is necessarily unknowable then it is meaningless.
    2. If p is meaningless and not-p is meaningful then not-p is true.
    3. “God does not exist” is a necessarily unknowable proposition.
    4. Therefore, “God does not exist” is meaningless.
    5. “God does exist” is a meaningful proposition.
    6. Therefore, “God does exist” is true.
    (2) sounds kind of strange, but I am inclined to think is true. If a proposition is meaningful but its negation is meaningless I’d suggest it must be the case that it is true.
    In any case there is one more issue with the possible worlds analogue. Since possible worlds are by definition different than the actual world, it is not necessarily the case that a proposition p that refers to the actual world can be transferred to a different possible world and keep its meaning. Therefore we should limit the possible worlds analogue to representations of counterfactual states of the actual world. Which implies that, if God exists (i.e. if the actual metaphysically ultimate is the perfect person) then God exists in all possible worlds. (And, similarly, if naturalism is true then it is true in all metaphysically possible worlds.) But then, if God exists then under the common understanding of omniscience there is *no* meaningful proposition which is necessarily unknowable. Conversely then the premise that some necessarily unknowable proposition does exist entails that God does not exist. Here is the respective argument:
    1. If God exists then God exists in all metaphysically possible worlds.
    2. If God exists then God can know the truth value of all meaningful propositions.
    3. Therefore, if God exists there are no necessarily unknowable meaningful propositions.
    4. There are necessarily unknowable meaningful propositions.
    5. Therefore, God does not exist.
    I think the only premise a theist can object to is (4). Thus it would seem the theist is committed to the view not that “God does not exist” is false, but that “God does not exist” is meaningless. Which comports well with the tendency of naturalism to slip into nihilism.

    November 4, 2011 — 5:27
  • Clayton:
    1. Accepting “God understands John’s atheism” as c-proposition does not result in a problem for my argument. For, there is a possible world in which God exists and knows that He understands John’s atheism. Besides, God might understand why John is an atheist, even if John doesn’t know that God does not exist.
    2. I do not appeal to (A2) for my argument. For, as I said earlier, ‘existing necessarily’ is not part of the definition of God that I employ for the argument. Further, (A1) seems to be a sufficiently reasonable assumption, not more unreasonable than the assumption that it is possible that God exists. (As I mentioned in my original presentation of the argument, I recently read an excellent refutation of Dennis Whitcomb’s argument that omniscience is impossible.)
    3. The Cartesian conception of knowledge I use can be rendered as being an internalistic account of epistemic foundationalism for which the set of proper basic beliefs is restricted to, say, propositions known by logical proof or direct self-evident incorrigible intuition. Hence, on the Cartesian conception, mere hyperbolic doubt does not destroy all instances of knowledge. For, on the Cartesian view, even under hyperbolic doubt, there are proper instances of knowledge, such as “I exist”, “I’m having at this very moment the experience of seeing red” or “1+1=2”.
    4. It might be justified to say that in some possible worlds an atheist o-knows that God doesn’t exist. But there is no possible world in which somebody c-knows that God doesn’t exist, and that’s all I need for my argument.
    In general I would like to point out that there is a substantial, and epistemically relevant, difference between S asserting that S exists (e.g. God asserting that God exists), and S asserting that some other subject external to S, say U, does not exist (e.g. John asserting that God does not exist). For, in the former case S has direct access to its own mental states, whereas, in the second case, S does not have direct access to U, or to U’s mental states. Thus, while the former case, S asserting that S exists, is properly said to be maximally epistemic ideal, the second case is not maximally epistemically ideal. So, on the Cartesian view, we are justified to hold that the former case is an example of a proper basic belief, while the second case isn’t.

    November 4, 2011 — 5:58
  • Clayton Littlejohn

    Hi Emmanuel,
    Thanks for your response. Just a few points in response to your response.
    (i) My example wasn’t “God understands J’s atheism” but “God understands my atheism”. Somebody could say “God understands my socialism” and express a proposition they know to be true. Nobody can say “God understands my atheism” and express a proposition they know to be true.
    (ii) I don’t see why A1 is a reasonable assumption. Rather than get into a quibble about the semantics of “reasonable assumption”, I guess I’d want to know what’s wrong with this. I think this is a reasonable assumption:
    (RA) There’s no possible world in which a being knows whether the number of times 867-5309 occurs in the decimal expansion of pi is even or odd.
    Given (RA) your principle implies that the number is neither even nor odd. If forced to choose, I’d reject your principle and retain my (RA). Can you help me see why my assumption (RA) isn’t reasonable? I think it’s reasonable to doubt the possibility of a mind that can know what is contained in some infinite collections. (Without saying that there’s some possible world in which God knows the answer.)
    (iii) I don’t want to get bogged down in a debate over the Cartesian conception, but when I said that hyperbolic doubt destroys knowledge, I meant doubts that applied to particular propositions. Never mind all that. You say:
    “It might be justified to say that in some possible worlds an atheist o-knows that God doesn’t exist. But there is no possible world in which somebody c-knows that God doesn’t exist, and that’s all I need for my argument.”
    I don’t think you want to concede this. If there’s some possible world in which an atheist o-knows there’s no God, there’s a possible world in which there’s no God. If there’s a possible world in which there’s no God, the proposition _God doesn’t exist_ isn’t necessarily false. If that proposition is not necessarily false, we can c-know God doesn’t exist or your principle fails. I think you’ve made a mistake about what your argument needs. Your argument needs a principle that admits of no counterexamples and will be persuasive only if you can give the atheist some justification to believe your principle admits of no counterexamples. You’ve just told me that I’m justified in believing your principle admits of counterexamples. So, as I said, you’ll need an additional argument to show that the atheist cannot o-know that God doesn’t exist.

    November 4, 2011 — 10:41
  • Tim Pawl

    Dear Emanuel,
    Thanks for the response.  What if we beef up Mr. Know-no to Mr. Eknow-no, a guy who essentially knows, in the cartesian sense, that God does not exist, and essentially knows himself to know this?  Then we can run an argument that starts with a world, W1, in which it is true that Mr. Eknow-no exists.  Surely, in W1, for Mr. Eknow-no, the belief that God does not exist is self-evident and incorrigible (it’s written right there into the essence of Mr. Eknow-no, just like the omni-properties are written right into the essence of God).  But then there is a possible world in which someone knows that God does not exist, and so God is not necessary.
    You write: “Now, to refute this response you would have to at least indicate how the belief that God does not exist could be self-evident or incorrigible for some agent in some possible world.”  Three independent responses of descending strength. First: his belief in God’s non-existence is essentially incorrigible to Mr. Eknow-no, just like God’s belief in my existence is essentially incorrigible to God. Second: just as the theist doesn’t need to explain how God could come to know every truth in order to put forward the possibility that there be an omniscient thing, so likewise one needn’t have an explanation of how Mr. Eknow-no comes to know God doesn’t exist to put forward that possibility. Third: even if I cannot indicate how Mr. Eknow-no comes to know God doesn’t exist, that doesn’t mean that he doesn’t or couldn’t come to know that God doesn’t exist. For, even if I can’t explain how it could happen, there could still be a remote possibility that it does happen. You say (re: arguments against God’s existence) that a remote possibility is all we need to rebut arguments that something doesn’t exist. I don’t see why we don’t have the remote possibility of Mr. Eknow-no’s existence, even if I cannot explain how he comes to know what he knows.

    November 4, 2011 — 10:47
  • Clayton:
    I would say that no rational agent S can believe that God understands S’s atheism. For, believing this would entail that God exists, which contradicts S’s atheism. Perhaps I should restrict the scope of my metaphysical principle to propositions P for which there is a possible world in which some rational agent could believe P.
    Moreover, I thought that for you o-knowledge is more or less the same as justified belief. Therefore I agreed that there might be atheists who o-know that God doesn’t exist. But, if o-knowledge entails truth, then, surely, if my argument is sound, there is nobody who o-knows that God doesn’t exist. For, if o-knowledge entails truth, then a falsehood cannot be o-known.
    Further, I would indeed say that there is a possible world in which God exists, and in which God knows all mathematical truths by, say, direct intuition.

    November 4, 2011 — 11:58
  • I think that the principle “If it is impossible to know that p, then p is necessarily false” is not true. Consider the proposition p:
    (p)I do not exist
    The truth statement behind that proposition is possibly true, in fact it was actually at some time and will be true again at some later time. But, that proposition in its indexical form cannot be known, since no one can know that they do not exist. To put it in more standard terminology, the content is possibly true but the linguistic meaning or character renders it impossible to know. So I think that the metaphysical principle on which the argument is based is false.

    November 4, 2011 — 20:05
  • Atheological: No rational agent S can believe “S does not exist”. For, believing this would entail that S exists, which contradicts S’s belief that S doesn’t exist. As I proposed to Clayton, the scope of my principle should be restricted to propositions P for which there is a possible world in which some rational agent could believe P. But then your counter example fails.

    November 5, 2011 — 7:11
  • Clayton Littlejohn

    Hi Emmanuel,
    (i) On restricting your principle…
    You can restrict the principle however you like, but at some point I think it’s fair to ask why we should think the restricted principle would be true when the unrestricted principle isn’t. Why should a principle that purports to uncover what’s metaphysically possible or impossible depend upon what’s rationally believable?
    (ii) On o-knowledge…
    o-knowledge is just knowledge according to ordinary standards (i.e., standards that don’t require anything like grounds that rule out any possibility of error). It’s a factive notion. I thought the whole point of this exercise was to offer an argument that an atheist might find convincing. Since most atheists think they o-know that there’s no God, most atheists would think that they o-know that the proposition that God doesn’t exist is a counterexample to your principle. That’s why I don’t think the principle you offer will be dialectically effective on its own. You might be right that atheists don’t o-know that there’s no God, but if you’re trying to offer an argument that convinces atheists don’t you have to appeal to considerations they might find remotely plausible?
    For example, you think that God might know whether 867-5309 occurs an even number of times in the decimal expansion of pi. You might be right. I seriously doubt that you’re right, but I think I have this much right–you can’t expect atheists to give you these points for free. The onus isn’t on the atheist to demonstrate the untenability of your positions. I thought the onus was on you to motivate your views. As I said, I can’t think of any reason to think that it’s metaphysically possible for a being to know that 867-5309 occurs x times in the decimal expansion of pi apart from thinking that, say, God is metaphysically possible. As an atheist, I think God is metaphysically impossible and so think that I have another counterexample to your principle. If your goal is to offer a valid argument with premises that only theists might take seriously, you might have done that. Indeed, you might have offered a sound argument consisting of premises that the theist accepts. But, you asked if it would convince an atheist and I think you’re expecting the atheist to take quite a lot on faith.

    November 5, 2011 — 18:10
  • Clayton: Regarding the topic of c/o-knowledge I refer to my previous response(s). Further, I take it that, reasonably, the prior plausibility of the premises of my argument is higher than the prior plausibility of the proposition that God necessarily exists, which, I would say, makes the argument relevant for the debate. Now, I believe this specific thread of the discussion has reached its ‘natural’ end.

    November 6, 2011 — 5:07
  • The proposition (p) I do not exist is knowable in some possible world. What I believe you’re confusing is the proposition (p) with the statement (p). Certainly no rational agent can believe the statement (p), but I wasn’t aware that you’ve so restricted your principle.
    In any event, there are some other issues with your argument. For one, even accepting the Cartesian view of knowledge, it doesn’t seem to be impossible to know that god exists. Consider that the ideal reasoner exists in some possible world and has an argument to the effect that god does not exist, and this argument is not an evidential argument but is rather deductive, for instance a logical argument from evil or an impossibility argument. In this case the ideal reasoner would know, on the Cartesian view, that god does not exist. This seems that it could be true in some possible world, but then possibly the proposition god does not exist is known.
    Also consider your rebuttal of the objection that it is equally impossible to know that god does exist. Your refutation is that in some possible world god exists and people can encounter him in that possible world, and in that possible world god knows that he exists. Well this presupposes that god does exist in some possible world. But the usual concept of god is of a necessary being, either existing in no possible world or in every possible world. And so your refutation begs the question by already assuming that god does exist, since if he exists in some possible world, he exists in all possible worlds. What the atheist maintains is god, so conceived of as a necessary being, exists in no possible world.

    November 6, 2011 — 12:06
  • Atheological: Your two objections have already been proposed by others. See for example my responses to Aaron and Clayton.

    November 6, 2011 — 15:34
  • Aaron

    Emanuel,
    Re: “‘An immaterial uncaused person that is the direct or indirect originating cause of all other concrete particulars’. This definition, I would argue, is logically consistent. And there are many similar examples. So, in order for the objection to have force one would at least have to suggest some sketch of a proof that the concept of God is logically contradictory.”
    Epistemically, it seems entirely possible that at some future time (or at some time in the past) an argument can be (or has been) given which shows that, say, it is necessarily the case that there are no necessarily existing entities, or that, e.g., ‘an immaterial person’ is incoherent (as I would argue) or inconsistent.
    If you admit this epistemic possibility, then, in light of my initial comment to you, your argument does not appear to work.

    November 8, 2011 — 23:05
  • Aaron: On the Cartesian view of knowledge a conclusive argument for the claim that God does not exist would have to be based upon self-evident incorrigible basic beliefs. But then, since, as I would argue, the definition of God that I provided is not logically inconsistent, there cannot be an a priori proof of the non-existence of God.

    November 10, 2011 — 1:09
  • eliram

    This whole business about existing in possible worlds just doesn’t wash. Consider the statement “In some possible world W, God exists”. If W does not actually exist (as opposed to merely being possible), then the statement is meaningless, for it is meaningless to exist in a world that does not exist.

    November 10, 2011 — 14:33
  • eliram: Your comment shows a lack of understanding of the concept of possible worlds semantics. See for an introduction for example http://bit.ly/ohcwhR

    November 11, 2011 — 3:42
  • Aaron

    Emanuel,
    Yes, right, but is it not *epistemically* possible that, at some future time, you will read / hear / think of and accept an argument which has as its conclusion a proof that your favored definition of god is inconsistent (or incoherent) or entails inconsistencies? I do not see how you can deny this possibility. E.g., Though I am not a dialetheist, many of Priest’s arguments are compelling, and, for all I am aware, perhaps some of his future arguments will persuade me to revise my stance on the principle of non contradiction. I cannot imagine why we cannot do the same for god (this is not to say ‘god’ is on as sure grounds as the principle of non contradiction), no matter the flavor.

    November 11, 2011 — 9:29
  • Aaron: For the argument I define God as ‘personal first cause’. Now, both ‘person’ and ‘first cause’ can be taken to be logically coherent concepts. Moreover, they, as I would argue, do not have any mutually conflicting attributes. Hence they can be taken to be logically independent from each other, so that the combined concept of ‘personal first cause’ does not result in a logical contradiction. Thus, unless someone provides us with a convincing reason for believing that the combined concept is nevertheless inconsistent, we are sufficiently justified to hold that there is no proof of the claim that ‘God’ is a logically inconsistent concept, and this justification is all I need for the argument.

    November 11, 2011 — 16:12
  • Dianelos Georgoudis

    Clayton,
    I’d like to comment on a couple of points you raise:
    “if you’re trying to offer an argument that convinces atheists [snip]”
    I think that the goal of arguments is to justify one’s beliefs, and not to convince others. What’s more I find that the whole project of convincing atheists is a fool’s errand, for a sufficiently strong version of naturalism is unfalsifiable. Consider a naturalistic worldview based on two premises which are already amply supported by the physical sciences, namely 1) that the universe is physically closed in the sense that no supernatural causes or effects are ever observed, and 2) that there is a perfect correlation between brain states and mental states. Such a worldview necessarily fits with the whole of our experience of life including any philosophising we may entertain, and thus trounces it. For example suppose a new anti-naturalistic argument X is such that many people find it convincing. The naturalist will point out that in a naturalistic world people would feel that X is convincing, and that therefor this fact represents an undercutting defeater for X. Ultimately naturalism is like a black hole of reason: if you fall in it then reason can’t pull you out of it.
    “Why should a principle that purports to uncover what’s metaphysically possible or impossible depend upon what’s rationally believable?”
    When all is said and done an argument ultimately hangs not on the premises and epistemic principles on which it is built (for they are proper elements of the argument) but on one’s personal – or perhaps I should say experiential – commitments. So, for example, if one commits oneself to the view that the world is good one will be pushed away from naturalism and will be pulled towards theism. If one commits oneself to the view that there is personal freedom, responsibility, creativity, or the mere possibility of the new, then, again, ore will be pushed away from naturalism and will be pulled towards theism.
    Now in the context of the question above, if one commits oneself to the view that the world is intelligible, or in other words that the world is ultimately of a nature which is amenable to rational deliberation, then, again, one will be pushed away from naturalism and will be pulled towards theism too. So, I’d like to suggest, the answer to that question is this: As long as we use reason to justify our metaphysical beliefs we commit ourselves to the view that the world is intelligible, and thus should only consider what’s rationally believable. The naturalist who claims that rationality is not dependable when one considers metaphysics (and perhaps even explains why this is so) moves herself into a self-referentially incoherent position.

    November 13, 2011 — 2:32
  • Clayton Littlejohn

    Dianelos,
    “I think that the goal of arguments is to justify one’s beliefs, and not to convince others.”
    Even if you’re right about the goal of an argument (which I doubt, if only because of the definite article) remember that Emmanuel asked, “Is this new argument for theism convincing?” If you think trying to convince the atheist is a fool’s errand, shouldn’t you take that point up with the person who expressed an interest in convincing an atheist?
    “The naturalist who claims that rationality is not dependable when one considers metaphysics (and perhaps even explains why this is so) moves herself into a self-referentially incoherent position.”
    That’s a bold claim, but I don’t think you’ve offered any reason for it.
    One of the points I was trying to make in response to Emmanuel’s argument was just this. Consider two principles:
    P1: For any p, if p is epistemically impossible, p is metaphysically impossible.
    P2: For any p such that p is an X-proposition, if p is epistemically impossible, p is metaphysically impossible.
    (Assume that there’s some p that’s not an X-proposition and P2 will be weaker than P1.)
    Because there are counterexamples to P1, some of us are skeptical of the move from claims about what’s epistemically possible to claims about what’s metaphysically possible. Emmanuel tried to restrict the original principle to X-propositions where the X/non-X distinction is an epistemological one, not a metaphysical one. This doesn’t address the general worry that you can’t derive metaphysical claims from epistemic ones. It assumes that the past failure to do so was due to some sort of funny trickery. The only assurance that we’ve been given that this move works is that the counterexamples are impossible to create because God is a possible creature who blocks them. As I said, that’s not very assuring to an atheist.

    November 13, 2011 — 10:41
  • Not a philosopher

    Gentlemen, don’t forget that possible worlds are merely possible. So anything happening in a possible world is just possible. It is possible that god exists. It is also possible that he does not. We don’t know. It is possible that this world is the only world there is. It is plausible that this is the only world you will ever see. So stop arguing and make something of your life!

    November 13, 2011 — 15:12
  • Joshua Rasmussen

    Emanuel,
    I’m probably missing something, but it’s still not clear to me why the argument cannot be reversed if you grant the logical possibility that God does not exist (where the proposition that God does not exist is a c-prop, I take it). For suppose it is logically possible that God does not exist. Then if what can’t be known is impossible, then it is in fact possible to know that God does not exist–even if we doesn’t know how one might know that. But then from the Fitch principle that whatever can be known is known, it follows that God does not exist. What did I miss?

    November 13, 2011 — 16:42
  • Joshua: As I proposed earlier, the principle on which my argument is based can be alternatively rendered in the following way: ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’. But then, given that on the Cartesian view of knowledge it’s impossible to know the c-proposition ‘God does not exist’, it follows that God exists in all K-worlds, and thus also in our world (being a K-world). And this would be compatible with there being one or more non-K-worlds in which God doesn’t exist. By the way, I believe that Fitch showed that if all truths can be known, then all truths are known, which is different from the claim that whatever can be known is known. Or perhaps I’m missing something here?

    November 14, 2011 — 6:49
  • Peter van Velzen

    Premisse
    Whether A is true cannot be known
    (Whether not A is true could be known if not A were true)
    So it follows that A is not true
    So it followes that Not A is true
    LOL!
    As usual the real prove for the existence of god lies in the assumption thant God exists (for only then it could be known that he existed)

    November 14, 2011 — 7:01
  • Joshua Rasmussen

    Thanks, Emanuel. I think I can see better how parity might be broken.

    November 14, 2011 — 7:50
  • Clayton Littlejohn

    Hi Emmanuel,
    I had a question about omniscience and God. You wrote:
    “… take for p the proposition ‘God does not exist’. It seems reasonable to hold that it is impossible to know that God does not exist. For, whatever the arguments against God, there will always be some (perhaps an extremely remote) possibility that God does exist after all, so that we can never truly say, on the Cartesian view, that we know that God does not exist. But then it follows that it is necessarily false that God does not exist. Hence, it is necessarily true that God exists. The principle thus entails theism. Is this new argument for theism convincing?”
    Your argument for the claim that it’s impossible to know by Cartesian standards that God doesn’t exist seems to be this:
    (P1) To know that God doesn’t exist by Cartesian standards, you’d have to eliminate any epistemic possibility that God doesn’t exist.
    (P2) It is metaphysically impossible to eliminate the epistemic possibility that God doesn’t exist.
    (C) Thus, it is metaphysically impossible to know by Cartesian standards that God doesn’t exist.
    Once you have (C), you have God (given your principle).
    Why accept (P2)? I don’t know. Here’s one reason to doubt whether (P2) is true. Let ‘Schmod’ be the omniscient being that isn’t omnibenevolent. Here’s an argument for Schmod’s existence:
    (P1′) To know that Schmod doesn’t exist by Cartesian standards, you’d have to eliminate the epistemic possibility that Schmod doesn’t exist.
    (P2) It is metaphysically impossible to eliminate the epistemic possibility that Schmod doesn’t exist.
    (C’) Thus, it is metaphysically impossible to know by Cartesian standards that Schmod doesn’t exist.
    Once we have (C’), your principle shows that Schmod exists. Schmod isn’t God, but presumably if there’s no God, Schmod could satisfy the Cartesian requirements for knowing this. So, I don’t think you can just assert that there’s no possible being that could believe there’s no God and satisfy the Cartesian requirements for knowing that there’s no God. (We can define omniscience in such a way that for any p that’s true, the being knows with Cartesian certainty that it is and for any p that’s false the being knows with Cartesian certainty that it is. So, let’s assume Schmod is God’s epistemic equal.)
    What assurance can you give us that there’s not some possible knower distinct from God that could determine with Cartesian certainty whether or not there’s a God? If you can’t provide that assurance, what justification can you give for (P2)? (Or, is your argument for (C) an argument that doesn’t rest on (P2?)

    November 14, 2011 — 8:40
  • Peter: My argument doesn’t rely on the assumption that God exists. Of course not, for that would make the argument clearly circular. See my previous responses for further explanation.

    November 14, 2011 — 13:31
  • Clayton: As I argued for in my previous responses we are sufficiently justified to accept that it is impossible to eliminate the possibility that God exist, since, as I argued for in my previous responses as well, the definition of God as ‘personal first cause’ is not logically contradictory, so that there cannot be an a priori proof for the proposition that God doesn’t exist. Moreover, it is not possible to eliminate the possibiliy that God exists by direct intuition, empirical observation or testimony either.
    Further, it is in fact possible to know that Schmod does not exist. Take a possible world in which God (being the ‘first cause’, that is, the origin or ground of that world) exists and in which God decides to create nothing else. In that world God can be taken to know that Schmod does not exist.

    November 15, 2011 — 1:54
  • Clayton Littlejohn

    Emmanuel,
    You wrote:
    “As I argued for in my previous responses we are sufficiently justified to accept that it is impossible to eliminate the possibility that God exist, since, as I argued for in my previous responses as well, the definition of God as ‘personal first cause’ is not logically contradictory”
    I guess I wasn’t persuaded by the argument. You seem to think this principle is true:
    (P) If p isn’t logically contradictory, there can’t be an apriori proof that satisfies Cartesian standards for knowledge that shows that ~p.
    Objection: I take it that neither (P) nor (~P) is logically contradictory, but (a) an omniscient being would know by Cartesian standards whether (P) or (~P) and know whether (P) or (~P) admits of counterexamples. Either you disagree and there’s no omniscient beings at all, or you agree and agree that (P) is false. Since your argument depends upon (P) and the truth of (P) seems to rule out beings that know (P) by Cartesian standards, I don’t think you’ve given us any reason to think that there couldn’t be an omniscient being distinct from God that knows metaphysical truths whose negations aren’t logical contradictions.
    In addition to what Schmod could know apriori, I don’t see any reason to accept what you say about Schmod’s aposteriori knowledge about God’s non-existence when you write:
    “Moreover, it is not possible to eliminate the possibiliy that God exists by direct intuition, empirical observation or testimony either.”
    Just as God could know by Cartesian standards what evils there are in the world and what it takes to justify them, so could Schmod. I say that Schmod knows by Cartesian standards that there are evils that God wouldn’t allow. Your claim to the contrary is an assertion, not an argument.
    You wrote, “Further, it is in fact possible to know that Schmod does not exist. Take a possible world in which God (being the ‘first cause’, that is, the origin or ground of that world) exists and in which God decides to create nothing else. In that world God can be taken to know that Schmod does not exist.”
    I guess I don’t find that persuasive, either. You’re assuming that Schmod isn’t a necessary being and that Schmod depends upon something else for its existence. That’s not something you’re in a position to say. As I showed in my previous response, your principle commits you to the claim that “Schmod doesn’t exist” is necessarily false. (Schmod isn’t logically contradictory. You can’t say (without assuming things you haven’t shown about God’s mind) that anyone can know that Schmod doesn’t exist, etc…). You can no more prove (to Cartesian standards) that Schmod doesn’t exist than I can prove (to Cartesian standards) that God doesn’t exist by saying that Schmod is (if anything) a first cause that is omniscient but not omnibenevolent and then saying that there’s a world where Schmod decides not to create God.

    November 15, 2011 — 11:49
  • Clayton: Suppose for reductio that (~P) is not logically contradictory. Now, (~P) asserts that there is a proposition, say p, that isn’t logically contradictory and for which it is true that the laws of logic entail ~p. But then ~p is a tautology, which completes the reductio since p, being the negation of a tautology, is in fact logically contradictory after all. Hence, we should not, as you do, take it that neither (P) nor (~P) is logically contradictory.
    You also say that Schmod knows by Cartesian standards that there are evils that God wouldn’t allow. But this, I would argue, is not the case. For, no agent can eliminate by logical proof or direct intuition the possibility that God has one or more morally sufficient reasons for permitting some particular evil or set of evils.
    Further, omniscience nor omnibenevolence is part of the definition of God employed for my argument. The argument entails a personal first cause, not an omniscient (or omnibenevolent) personal first cause. Moreover, for my argument I do not need to assume that ‘omniscient personal first cause’ (or ‘omnibenevolent personal first cause’) is a coherent concept. So, for all we know, this concept might be incoherent, and if so, my metaphysical principle cannot be used to infer that there is an omniscient (or omnibenevolent) personal first cause.

    November 15, 2011 — 14:23
  • Clayton Littlejohn

    Emmanuel,
    You write:
    “Suppose for reductio that (~P) is not logically contradictory. Now, (~P) asserts that there is a proposition, say p, that isn’t logically contradictory and for which it is true that the laws of logic entail ~p. But then ~p is a tautology, which completes the reductio since p, being the negation of a tautology, is in fact logically contradictory after all. Hence, we should not, as you do, take it that neither (P) nor (~P) is logically contradictory.”
    Hmmm… (~P) asserts that there’s a proof that satisfies Cartesian standards for knowledge of ~p even if p isn’t a logical contradiction. There’s no contradiction there. Such a proof would have to be deductive, but the premises of the proof needn’t be logical truths. The flaw in your reductio is just this: you fail to appreciate that there can be proofs that transmit Cartesian knowledge where the premises of the proof aren’t logical truths. (To say otherwise, you’d have to say that it’s a logical contradiction that there are propositions that aren’t logical truths that can be known with Cartesian certainty. Certainly _that_ isn’t a logical truth! Moreover, you’d want to say that an omniscient being like God knows lots of truths that aren’t logical truths with Cartesian certainty.)
    “But this, I would argue, is not the case. For, no agent can eliminate by logical proof or direct intuition the possibility that God has one or more morally sufficient reasons for permitting some particular evil or set of evils.”
    Yes, but again, we’re trying to see what’s required for your original proof to work. Surely if God could know by Cartesian standards what it would take to justify an evil, what evils there are, and what justifications there are, so would Schmod. And, so Schmod could know whether whether there’s a sound argument from evil and know this by Cartesian standards.
    You wrote:
    “Further, omniscience nor omnibenevolence is part of the definition of God employed for my argument. The argument entails a personal first cause, not an omniscient (or omnibenevolent) personal first cause. Moreover, for my argument I do not need to assume that ‘omniscient personal first cause’ (or ‘omnibenevolent personal first cause’) is a coherent concept. So, for all we know, this concept might be incoherent, and if so, my metaphysical principle cannot be used to infer that there is an omniscient (or omnibenevolent) personal first cause.”
    You’re “missing the bullet”, so to speak. You ran a proof using your principle that (allegedly) shows that there’s a personal first cause. That’s consistent with my running a parallel proof to show that there’s an omniscient first cause, an omnibenevolent first cause, a left-handed first cause, etc. If, in the course of so doing, I’ve found a being with godlike epistemic powers who can know with Cartesian certainty everything there is to know about metaphysics, you can’t say that there’s no being (besides God) that could know with Cartesian certainty whether there’s a God. Schmod’s existence seems to be guaranteed by your principles and Schmod isn’t God. Remember way back when, you said that it’s just obvious that there could be an omniscient being. I thought I was agreeing with you.

    November 15, 2011 — 15:52
  • Clayton: Of course there can be proofs that transmit Cartesian knowledge where the premises of the proof aren’t logical truths. That’s obvious. For, those premises could be known, on the Cartesian view, by either direct intuition or incorrigible experience.
    Now, my point was that, on the Cartesian view, the only way to eliminate the possibility that God exists would be to show by mere logical means that the concept of God, i.e. personal first cause, is contradictory, since direct intuition or incorrigible experience could help to establish mathematical facts or facts such as ‘I exist’ and ‘I feel cold’, but they will surely not help to conclusively establish (modulo logical deduction) that God does not exist.
    Further, as I argued before, for my argument I take it that we are justified to hold that the basic concept of ‘personal first cause’ is coherent, since we have reasonably a sufficient understanding of the notion of both ‘person’ and ’cause’, and because these two notions seem to be mutually independent, non-redundant or non-overlapping.
    But then no person can know that God, understood as personal first cause, does not exist. And this would be true for your ‘Schmod’ as well.
    Yet I would not affirm that the more complex concepts of ‘omniscient personal first cause’ and ‘omnibenevolent personal first cause’ are coherent as well, precisely because the notions of ‘omniscience’ and ‘omnibenevolence’, contrary to that of ‘person’ and ’cause’, are surely much more controversial in the contemporary debate. Arguments have been developed for the claim that for example omniscience is logically impossible, and refutations of those arguments have been proposed as well.
    But then I would say we are not in the position to use the premises of my argument to conclude that there is an ‘omniscient personal first cause’ or that there is an ‘omnibenevolent personal first cause’. After all, if these concepts would indeed be contradictory then it is not impossible to know that there isn’t an omnibenevolent (or omniscient) personal first cause, although it would in that case still follow that there is a non-omnibenevolent (or non-omniscient) personal first cause.

    November 16, 2011 — 5:57
  • Clayton Littlejohn

    Hi Emmanuel,
    “Of course there can be proofs that transmit Cartesian knowledge where the premises of the proof aren’t logical truths. That’s obvious.”
    I didn’t think it was obvious. Your reductio from earlier seemed to suggest you thought otherwise.
    You wrote:
    “Now, my point was that, on the Cartesian view, the only way to eliminate the possibility that God exists would be to show by mere logical means that the concept of God, i.e. personal first cause, is contradictory, since direct intuition or incorrigible experience could help to establish mathematical facts or facts such as ‘I exist’ and ‘I feel cold’, but they will surely not help to conclusively establish (modulo logical deduction) that God does not exist.”
    That assumes that Schmod doesn’t know some metaphysical principle of other that rules out God’s existence or that such metaphysical principles couldn’t be known with Cartesian certainty. That assumption seems problematic (to me, at any rate). Perhaps such principles would show that even a coherently conceivable state of affairs isn’t metaphysically possible.
    Now, I had assumed that the possibility of an omniscient God played an important role in your proof for a number of reasons. First, in trying to generate an apparent counterexample using the decimal expansion of pi, you needed an omniscient being to block my attempt at generating a contradiction. Second, you had to assume that there’s some being who could know with Cartesian certainty that there’s a God. For you, ‘God’ seems to be a descriptive name, one that doesn’t pick out an omnicompetent being, but merely a personal first cause. One worry about this approach is that while it’s plausible that if there’s a personal first cause, it could know that it exists with Cartesian certainty, it doesn’t follow that it could know that God exists with Cartesian certainty.
    To see why this second point matters, consider Descartes. Descartes was the author of the Meditations. While the narrator could know with Cartesian certainty that he/she himself/herself existed, it doesn’t follow that the narrator knew with Cartesian certainty that the author of the Meditations existed. If the PFC doesn’t know with Cartesian certainty that the PFC exists (under that description), then doesn’t your response to objection #3 fail (Here’s what you wrote:
    “My response would be that even if someone could always think that he or she is dreaming, hallucinating or being deceived, it still does not follow that it is impossible to know that God exists. For, take a possible world in which God exists. In this possible world there is a subject that knows that God exists, namely God. Indeed, in that world God knows that God exists. So, it is not impossible to know that God exists.”
    I say that if you can’t build in godlike epistemic capacities to your PFC, you haven’t demonstrated that there’s some possible world in which a being knows with Cartesian certainty that there’s a PFC. You’ll at least have to say that there’s some possible world in which the PFC is omniscient if you want there to be some possible world in which someone knows with Cartesian certainty that the definite description ‘the personal first cause’ denotes something.)
    So, with all due respect, I don’t think you’ve yet appreciated the epistemic difficulties that your argument faces.
    There’s a further worry as well. In elaborating the details of your argument, you seem to think this is a good line of reasoning:
    (1) The concept of a PFC is coherent.
    (2) If the concept is coherent, nobody can know with Cartesian certainty that the description ‘the PFC’ isn’t satisfied (No, not even the mighty Schmod!)
    (3) Thus, nobody can know with Cartesian certainty that the PFC doesn’t exist.
    (4) Thus, it’s necessarily false that the PFC doesn’t exist.
    I still think your susceptible to parity arguments. Isn’t the idea of a necessarily existing world that has no independent cause a coherent concept? I see no reason to think that it’s not. And so, I give you this:
    (1′) The concept of the uncaused world (UW) is coherent.
    (2′) If the concept is coherent, nobody can know with Cartesian certainty that the description ‘the UW’ isn’t satisfied.
    (3′) Thus, nobody can know with Cartesian certainty that the UW doesn’t exist.
    (4′) Thus, it’s necessarily false that the UW doesn’t exist.
    The UW and the PFC aren’t compossible. If the idea of a necessary uncaused person that’s a cause of the universe and is coherent, I submit that the idea of a necessary uncaused being that’s not a person is also coherent.

    November 16, 2011 — 17:49
  • Clayton: My reductio from earlier did not suggest otherwise. For, as I argued above, in the case of trying to conclusively eliminate the existence of God direct intuition or incorrigible experience doesn’t help, so that’s why I explicitly mentioned that in that specific case the proof must be based on logic alone. But that surely doesn’t entail that all Cartesian knowledge is a matter of logical truth.
    Further, I do not need the possibility of an omniscient God to generate an apparent counterexample using the decimal expansion of pi. For, a possible world in which God knows all mathematical truths (instead of all truths simpliciter) by direct intuition will do. And, yes, I would indeed claim that in those possible worlds in which God exists, God knows that God is God. Who could be in a better epistemic situation regarding its own nature than that world’s God? I definitely would say that it is sufficiently plausbile to maintain that, if anyone, at least God knows that God is God. And for this we don’t need God to be omniscient either.
    Moreover, it is not the case that nobody could know that the UW doesn’t exist. For, again, consider a possible world in which God exists. In that possible world God, being the personal first cause of the world, knows that the world (except for God, who is first cause and thus uncaused) isn’t uncaused. So, in general, the fact that something is logically consistent doesn’t imply that it couldn’t be false (and known to be false) in some worlds.

    November 17, 2011 — 3:25
  • Clayton Littlejohn

    Emmanuel,
    “For, a possible world in which God knows all mathematical truths (instead of all truths simpliciter) by direct intuition will do.”
    But, you seemed to be moving towards the view that God is a descriptive name with its reference fixed by the personal first cause (PFC). It’s not obvious why the PFC would have such mathematical abilities or could have such mathematical abilities. You need, however, for there to be a being that has such abilities to save your principle. Surely reasonable doubts we have about the coherence and possibility of an omniscient being would extend to a PFC with such amazing mathematical abilities.
    “And, yes, I would indeed claim that in those possible worlds in which God exists, God knows that God is God. Who could be in a better epistemic situation regarding its own nature than that world’s God? I definitely would say that it is sufficiently plausbile to maintain that, if anyone, at least God knows that God is God. And for this we don’t need God to be omniscient either.”
    Even if nobody could be in a better epistemic position than God to know that God is God, it doesn’t follow that God could know with Cartesian certainty that God exists. (After all, it might be that nobody can meet the Cartesian standards for knowing that.) Notice that I didn’t deny that if God thought to itself ‘I think I exist, so I must exist’, God could thereby know with Cartesian certainty that it exists. What I questioned is what entitled you to this assumption: the PFC could know with Cartesian certainty that it was the personal cause of the world. For most of the things that we do, we cannot know with Cartesian certainty that we’ve done them. So, unless the PFC is endowed with some pretty extraordinary epistemic abilities, I don’t see how the PFC could eliminate the possibility that the world wasn’t caused, that some other being caused the world and tricked it into thinking that it was the world’s creator, that the world had really been created three seconds earlier with the false memory of creating the world intact, that the PFC is living in the Matrix, etc.
    “Moreover, it is not the case that nobody could know that the UW doesn’t exist. For, again, consider a possible world in which God exists. In that possible world God, being the personal first cause of the world, knows that the world (except for God, who is first cause and thus uncaused) isn’t uncaused. So, in general, the fact that something is logically consistent doesn’t imply that it couldn’t be false (and known to be false) in some worlds.”
    That’s rich. You can’t ask the atheist to consider the possible world in which God exists and creates the universe. Isn’t that just the very issue at issue? If (4′) is true, there’s no such possible world.
    I claim that my (1′)-(4′) are on par with your (1)-(4), however, (4′) entails that the description ‘the PFC’ isn’t a description that any object satisfies in any possible world. Since you haven’t shown how to block the derivation of (4) and (4′) from your principle and some ancillary assumptions you needed to try to prove that there’s a PFC (apart from the transparently question begging move of assuming that there’s a PW in which (4′) is false), I think I’m justified in concluding that your principle commits you to (4) and (4′). Any such principle has to be false. I fear that your argument is unsound.

    November 17, 2011 — 8:17
  • Clayton: The issue is whether God actually exists, not whether God possibly exists. Now, as discussed above at various occasions, I indeed hold that we are sufficiently justified to claim that God at least possibly exists, so that, as I pointed out in my previous response, your (2′) and (3′) are false. If the atheist wants to reasonably deny the mere metaphysical possibility of God’s existence, then he or she would have to argue that the existence of God is metaphysically entirely impossible, which, as I argued before as well, is something that cannot plausibly be shown. And, I would add, there being a possible world within which God exists and within which God has the aforementioned mathematical capabilities is, for the same reasons, sufficiently plausible as well (as is the claim that there are possible worlds in which God exists, but in which God does not have those capabilities). Moreover, note that ‘necessary existence’ is not part of my definition of God. Indeed, as I demonstrated before, the alternatively rendered version of my argument for the actual existence of God is perfectly compatible with there being possible worlds in which God doesn’t exist.

    November 17, 2011 — 15:39
  • Clayton Littlejohn

    Hi Emmanuel,
    Your argument for the actual existence of God, if it assumes the possible existence of God, rests on a highly contentious assumption. Moreover, as I’ve demonstrated above, your principle supports equally well the proposition that it’s necessarily the case that the world is uncaused and the conclusion that there must be a first cause. Your only response to this has been to assert on no grounds whatever that God is possible. Okay, well, I assert on the same grounds that the uncaused world is possible. Stalemate. It’s quite simply amazing that you don’t see this as a problem for your argument. (None of my objections, by the way, assumed that necessary existence was part of God’s definition. It’s a point you’ve raised a number of times and it doesn’t accurately describe anything I’ve ever said in response to your argument.)

    November 18, 2011 — 1:52
  • Clayton: Stalemate? Not at all. For, my “highly contentious” assumption that God possibly exists is perfectly compatible with there being uncaused (yet not necessarily uncaused) possible worlds. Notice, again, that my alternatively rendered argument entails that God actually exists, and this surely leaves open the possibility of there being possible worlds in which God doesn’t exist, such as your uncaused (yet not necessarily uncaused) worlds. Now, I believe our extensive bilateral discussion has reached its natural end. If you want to continue, then please do so on my own blog (http://bit.ly/oplzD8).

    November 18, 2011 — 7:26
  • Clayton Littlejohn

    Yes, well, the assumption that God possibly exists is highly contentious (as almost everyone concedes) and is flatly incompatible with the claim that ‘The UW doesn’t exist’ is necessarily false. Since the necessary existence of the personal first cause and the necessary existence of the uncaused world is what’s at issue and you won’t even concede that much, I think I can agree that this discussion has reached its end.

    November 18, 2011 — 8:30
  • Dianelos Georgoudis

    Clayton,
    “If you think trying to convince the atheist is a fool’s errand, shouldn’t you take that point up with the person who expressed an interest in convincing an atheist?”
    One reason why trying to convince others is often a fool’s errand is that there are many worldviews which are unfalsifiable from the point of view of those who hold them, such as the scientific naturalism I described above, young Earth creationism, skeptical theism, the evil demon hypothesis, the computer simulation hypothesis, solipsism, etc.
    Now all arguments work within some given epistemic commitments. When Emmanuel asks whether his argument is convincing I think he means whether his argument is thought to be sound by those who share his epistemic commitments. And, given the nature of his argument, I think it is clear that his epistemic commitments entail that reality is metaphysically intelligible. (By “intelligible” I mean “amenable to understanding” or “knowable”.) Indeed, much of theistic argumentation, and certainly the type of argument that the scholastics developed, is based on the idea that what is metaphysically ultimate is intelligible. In other words these theistic arguments have the form: If the metaphysically ultimate is intelligible then theism is true. By chance I’ve just found this quote by Berkeley: “As to what is said of the absolute existence of unthinking things without any relation to their being perceived, that seems perfectly unintelligible”. Which, Berkeley implies, is reason enough for one to reject belief in such things.
    In conclusion, what seems to characterize theistic thought is precisely the commitment to intelligibility, a commitment that I find is in some sense a necessary one. For if the world is ultimately not intelligible then why think about metaphysics? Thus it seems to me that when you ask “Why should a principle that purports to uncover what’s metaphysically possible or impossible depend upon what’s rationally believable?” you are implicitly rejecting the idea that what’s metaphysically ultimate is intelligible. But then you are placing yourself outside of Emmanuel’s basic epistemic commitments and thus fail to engage with him.
    “This doesn’t address the general worry that you can’t derive metaphysical claims from epistemic ones.”
    I think I understand what you mean and I agree. On the other hand one must embrace some epistemic claim before arriving to any ontological claim. Thus, whether we like it or not, all ontological claims are ultimately based on epistemic commitments.
    Let me clarify what I’m saying. Clearly, ontology beats epistemology, in the sense that the right epistemology is right because it fits the right ontology, and not the other way around. On the other hand we must use some kind of epistemology before arriving to any beliefs about ontology. It seems to me that this implicit circularity causes much of the difficulty of metaphysics. If this picture is right then there is no such thing as non question-begging metaphysical argumentation, for by choosing any epistemology as the basis of one’s reasoning one is excluding those ontological hypotheses which do not fit with the epistemology chosen. Therefore metaphysical argumentation should always have the following form “If one commits oneself to X epistemology then belief in Y ontology is reasonable”.
    Here’s a case in point: If scientific naturalism is true and our cognitive capacities have been produced by a blindly mechanical process driven by biological adaptation then we should *not* expect them to be capable of understanding what’s metaphysically ultimate (for such understanding offers no adaptive advantage whatsoever). Take something as basic as the naturalistic interpretation of quantum mechanical phenomena (which, by the way, are factual and independent from the actual state of the theory of quantum mechanics). When one points out how absurd such interpretations are, the naturalist sensibly points out that what strikes us as absurd is irrelevant; what is “rationally believable” is a property of the evolutionary history of our brain, and not of the truth. I think the naturalist is right. Therefore, if in an argument against scientific naturalism one uses an epistemology which entails that the metaphysically ultimate is intelligible then one is implicitly excluding scientific naturalism and thus begging the question.
    You ask for the reasoning behind the following claim: “The naturalist who claims that rationality is not dependable when one considers metaphysics (and perhaps even explains why this is so) moves herself into a self-referentially incoherent position.”
    I suggest that the following are self-referentially incoherent: “I reason that my reasoning is not dependable”, or, more explicitly, “I have reasons to believe that reality is such that my reasoning about it is not dependable”, or more explicitly still, “I believe in scientific naturalism because I trust in my metaphysical reasoning, even though scientific naturalism entails that my metaphysical reasoning is probably not trustworthy”.

    November 22, 2011 — 7:47
  • Mancis

    Hi Emanuel,
    1. Given your definition of “God”, it is surely logically possible that God doesn’t know (in the Cartesian sense) that God exists.
    2. God defined as “personal first cause” CAN “always think” that he “is dreaming, hallucinating or being deceived” and therefore CANNOT rule out the possibility that he is mistaken in his belief that he is “the direct or indirect originating cause of all other concrete particulars”.

    December 17, 2011 — 18:17
  • Mancis: God is in fact quite special in the sense that God is the unconditional origin or ultimate ground of reality. And so, as being the absolute first cause of everything that exists, God is in an ideal epistemic situation with respect to God’s identity. God’s belief that God is God is therefore sufficiently incorrigible or basic. Indeed, God’s belief that God is God is surely not less warranted than John’s belief that John is John, or Mary’s belief that Mary is Mary. So God’s belief that God is God counts as knowledge, even under the Cartesian view.

    December 20, 2011 — 17:15
  • The same way one can prove the existence of anything. Mermaids, unicorns, phlogiston, ether, Zeus, and the spaghetti monster.

    April 9, 2012 — 2:41
  • Eric T. Schellekens

    I state:Little red dwarfs live inside black holes
    Physics does not allow us any detailed information of what goes on inside black holes.
    Hence what goes on inside black holes is
    unknowable.
    This does certainly not mean that it is
    untrue that little red dwarfs live inside black holes.
    I challenge you to prove it untrue.
    As this is the basic assumption behind your god exists proof,I cannot accept it (lest you can convince me that
    little red dwarfs live not inside black holes)

    April 15, 2012 — 18:53
  • Marcin

    Is it necessarily true that if x is “a personal first cause”, x knows that x is “a personal first cause”? I don’t think so. And equivocating with the word “God” is not going to help your case at all, Emmanuel 🙂

    April 16, 2012 — 8:38
  • Mihai: ‘Spaghetti monster’-type of objections are already addressed and refuted in the discussion above, and in my paper on gjerutten.nl
    Eric: In the actual world it might for us, human beings, be unknowable what goes on inside black holes. But from that it doesn’t follow that there is no metaphysically possible world at all within which some agent knows what goes on inside black holes. Therefore your parody objection does not go through.
    Marcin: I’m not saying at all that it is necessarily true that if x is “a personal first cause”, x knows that x is “a personal first cause”. I only assert that there is at least one metaphysical possible world within which a personal first cause x exists, and in which x does know that x is the personal first cause.

    April 16, 2012 — 9:50
  • Marcin

    Well, if it is possible that x is God, but x doesn’t know that x is God, “Possibly, God exists” can surely be true without “Possibly, some subject knows that God exists” being true. And I see no good reason to think that a personal first cause can attain Cartesian certainty and know (in the Cartesian sense) that there is a personal first cause.

    April 16, 2012 — 11:24
  • Tim Reynders

    Hi Emanuel,
    If I’m correct your proof goes as follows:
    P1: ‘If it is impossible to know that p, then p is necessarily false’
    P2: ‘It is impossible to know that God does not exist’
    Hence ‘God does not exists’ is false
    Therefore ‘God exists’
    I think your proof is at least incomplete. P1 presupposes that ‘It is possible to know P1’; for if it be impossible to know P1, than P1 is false and your proof reduces itself to an “ex falso quodlibet”. Therefore you first need to proof that ‘It is possible to know P1’ or need to add another premise out of which it can be concluded that ‘It is possible to know P1’
    It does not suffice to add: P3: ‘It is possible to know P1’
    For this presupposes that ‘It is possible to know P3’ (otherwise both P1 and P3 are false, and the proof reduces again to an “ex falso quodlibet”.
    Greetings,
    Tim

    April 16, 2012 — 13:18
  • Koen Verstrepen

    The first principle is untrue.
    Let’s call the first principle, the principle K: ‘If it is impossible to know that p, then p is necessarily false’.
    Consider now a second principle, the principle L: “principle K is true”. Now principle L, although reasonable, is impossible to know (for sure).
    Finally, replace the p in principle K by L and it follows that L is untrue and therefore K is untrue. Hence, if K is true, K is necessarily untrue. A paradox.
    K being true is as reasonable but impossible to know as god being non existant.
    rgds,
    Koen Verstrepen

    April 16, 2012 — 14:38
  • Marcin: You say that you do not see a good reason to think that a personal first cause can know that there is a personal first cause. Now, in the above discussion I have already responded multiple times to this objection. But see also my refutation of objection #8 in http://bit.ly/GKZIDl

    April 16, 2012 — 15:31
  • Tim: Why would it not be metaphysically possible to know that P1? In fact, it seems to me that it is indeed possible to know that P1. For, we can conceive a possible world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood. Moreover, P1 is not a c-proposition (see above discussion), so that your worry is not a problem for the alternative rendering of my argument.

    April 16, 2012 — 15:43
  • Koen: You claim that it is metaphysically impossible to know L. Yet, you do not provide a good supporting reason for this claim. But then we do not have to accept it. Moreover, L is not a c-proposition (see above discussion). Therefore, your worry, if real, is not a problem for the alternative rendering of my argument.

    April 16, 2012 — 15:51
  • Tim Reynders

    Hi Emanuel,
    It is not for me to show to you that ‘P1 is possible to know’. In order for your proof to hold, you have to prove that ‘it is possible to know P1’ holds. “Seeming” does not suffice, because the proof fails if P1 is false.
    Furthermore I don’t think that ‘we can conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’; because this phrase can be used for every falsity, that hasn’t proven to be false yet.
    If P1 is false, than it is impossible to ‘conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’
    If you state ‘we can conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’ as an assumption than you also state the underlying assumption ‘it is possible to know that we can conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’, and so on. You still need an infinity of assumptions for your proof to hold.
    For if one of the assumptions is impossible to know, the whole proofs reduces itself to an “ex falso quodlibet’.
    You can postulate that P1 is true, and you also can postulate that ‘it is possible to know P1’ is true, but this does not suffice.
    Therefore the proof is incomplete unless it is soundly proven that ‘P1 is possible to know’, or if a descent assumption P3 can be created out of which can be concluded that both P1 and P3 are possible to know; otherwise your proof requires an infinite set of assumptions.
    Your second principle ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’, has the same problem:
    Let F be the phrase: ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’, then we have the c-proposition Q: F is true. Your second principle also presupposes that ‘it is possible to know Q’, for if it be impossible to know Q, then there is no world in which Q is known, hence there is no K-world in which Q is true, hence F is false; therefore your second proof is also incomplete.
    It is incomplete unless it is soundly proven that ‘F is possible to know’, or if a descent assumption R can be created out of which can be concluded that both R and F are possible to know; otherwise your second proof also requires an infinite set of assumptions.
    Kind regards,
    Tim

    April 16, 2012 — 19:26
  • Tim Reynders

    Hi Emanuel,
    It is not for me to show to you that ‘P1 is possible to know’. In order for your proof to hold, you have to prove that ‘it is possible to know P1’ holds. “Seeming” does not suffice, because the proof fails if P1 is false.
    Furthermore I don’t think that ‘we can conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’; because this phrase can be used for every falsity, that hasn’t proven to be false yet.
    If P1 is false, than it is impossible to ‘conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’
    If you state ‘we can conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’ as an assumption than you also state the underlying assumption ‘it is possible to know that we can conceive a world in which God exists and in which God has sufficient modal-epistemic intuitions to know that the ultimate ground of metaphysical unknowability is necessary falsehood’, and so on. You still need an infinity of assumptions for your proof to hold.
    For if one of the assumptions is impossible to know, the whole proofs reduces itself to an “ex falso quodlibet’.
    You can postulate that P1 is true, and you also can postulate that ‘it is possible to know P1’ is true, but this does not suffice.
    Therefore the proof is incomplete unless it is soundly proven that ‘P1 is possible to know’, or if a descent assumption P3 can be created out of which can be concluded that both P1 and P3 are possible to know; otherwise your proof requires an infinite set of assumptions.
    Your second principle ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’, has the same problem:
    Let F be the phrase: ‘If p is a c-proposition that is true in at least one K-world, then there is a world in which p is known’, then we have the c-proposition Q: F is true. Your second principle also presupposes that ‘it is possible to know Q’, for if it be impossible to know Q, then there is no world in which Q is known, hence there is no K-world in which Q is true, hence F is false; therefore your second proof is also incomplete.
    It is incomplete unless it is soundly proven that ‘F is possible to know’, or if a descent assumption R can be created out of which can be concluded that both R and F are possible to know; otherwise your second proof also requires an infinite set of assumptions.
    Kind regards,
    Tim

    April 16, 2012 — 19:28
  • Tim: First, my argument is not a proof. I claim that both premises of my argument are sufficiently plausible and not that they can be proven. Now, the claim that P1 is knowable is sufficiently plausible as well since we can conceive of a metaphysically possible world in which P1 is known. Second, you say that the conceivability criterion fails since it can be used to establish every falsehood. But this is clearly not the case. Take the falsehood ‘a is not a’. We cannot conceive a possible world in which a is not a. And therefore conceivability can not be used to establish every falsehood. Third, the alternative rendering of your objection fails as well because both F and Q are not c-propositions.

    April 17, 2012 — 1:29
  • peter

    ‘If it is impossible to know that p, then p is necessarily false’. This principle seems cogent.
    Really? it just seems fake logic.
    For, if a given proposition p could be true, then, plausibly, there is some possible world in which some subject knows that p is true. In other words, if in *all* possible worlds *all* subjects do not know that some proposition is true, then, plausibly, that is because that very proposition cannot in fact be true.
    Why ‘plausibly’? There isn’t anything plausible here. One might as well argue that if it is impossible to know that p is true, there is no possible world where anyone can know whether p is true or not. So p in not ‘necessarily’ false: it is impossible to know whether p is or not.
    That first premise does not stand on any feet. And any argument falls down with it.

    April 25, 2012 — 10:01
  • Peter: The first premise has it that all truths are possibly known, either in the actual world or in at least some other (modal near or distant) possible world. Now, of course, the principle that all truths are knowable is not a formal theorem of logic. It is a synthetic claim about the nature of reality, namely the claim that reality is ultimately intelligible. And in fact a lot of traditional and contemporary metaphysical positions are committed to this claim for their relevant domain of discourse, such as (neo-)Aristotelianism and internal realism.

    April 28, 2012 — 2:42