The following is an interesting argument presented to me by a friend. The originator of the argument is unknown to me at this time, but seems to be an acquaintance or friend of a friend of a friend. In any case, it is interesting, and I evaluate it below.
D1: God is defined as omniscient, omnipotent and omnibenevolent
D2: Revelation is Gods disclosure of Himself to His Creation
D3: Evidence is propositional knowledge
D4: Revelation is experiential knowledge
D5: Resistant nonbelief is a disposition to incredulity
D6: Nonresistant nonbelief is a disposition to credulity, while being unconvinced
P1: God desires that mankind achieve a filial knowledge of Him
P2: God can be known propositionally and/or experientially
P3: Propositional knowledge does not entail, but may result in, filial knowledge
P4: Experiential knowledge does not entail, but may result in, filial knowledge
P5: If it is the case that God is omniscient then He is aware of states of affairs that would bring about filial knowledge of Him
P6: If it is the case that God is omnipotent then He is capable of actualizing states of affairs that would bring about filial knowledge of Him
P7: If it is the case that God has made His existence propositionally known in the past then such a state of affairs is logically possible
P8: If it is the case that God has made His existence experientially known in the past then such a state of affairs is logically possible
P9: God has actualized states of affairs that have led certain individuals to filial knowledge of Him
P10: God has not actualized states of affairs that would lead certain individuals to a filial knowledge of Him
P11: The degree to which Gods existence is known propositionally is at His discretion
P12: The degree to which Gods existence is known experientially is at His discretion
P13: There is a necessary state of affairs for individuals that would bring about their filial knowledge of God
P14: If it is the case that it is not logically possible for God to actualize a necessary state of affairs to bring about filial knowledge Him, then such a state of affairs is logically impossible
P15: States of affairs, which are logically impossible, do not entail culpability
C1: If nonbelief occurs it is either the case that God has not actualized states of affairs that would bring about filial knowledge, or it is the case that such a state of affairs is not logically possible (D1, D2, D3, D4, D6, P2, P3, P4, P5, P6, P7, P8, P9, P10, P11, P12, P13 and P14)
C2: Nonresistant nonbelief is inculpable (from D6, P2, P3, P4, P5, P6, P7, P8, P9, P11, P12 & P13)
C3: Resistant nonbelief is inculpable (from D5, P14 & P15)
P16: Inculpable nonbelief is incompatible with a God as defined in D1
C4: If it is the case that nonbelief occurs, God does not exist
Accepting all of the definitions, to whatever degree of ambiguity they may be (for instance, what do we mean by experiential knowledge? Is it just any revealing of God’s self?), we can now turn to the premises themselves. We must evaluate each premise in its strongest light (so as to be fair and charitable), and then see if the conclusions follow. What follows will be a discussion of every premise, to know exactly where we stand.
I shall take P1 to mean that “For each and every member of mankind M, ‘M comes to a filial knowledge of God’ (where “filial” shall be taken to mean “saving” in its theological sense) is God’s desire.” In that case, I agree.
P2 is just to state that God can be known and can choose to reveal himself to M. Things can be known about God through truth and reason, and the revelation of God’s Word. I would agree with that as well.
P3 is not as clear. Is it saying that although M can have propositional knowledge of God and not saving knowledge, nevertheless propositional knowledge alone can result in saving knowledge? This is plainly false. For it is not intellectual knowledge, but trust in God (as a matter of the will, not simply the intellect). However, this could simply be stating that propositional knowledge is a necessary, if not sufficient, condition, and that this propositional knowledge, coupled with something else, may eventually lead to a saving faith for M. In this latter case, I agree with P3. For the sake of charity, let us use this meaning for P3.
P4 suffers from a similar lack of clarity, as it is in the same structure, just with “experiential knowledge” (defined as revelation) instead of propositional knowledge. In this case, P4 ought to be taken as the fact that experiential knowledge is a necessary, but not sufficient, condition of saving faith for M, but may, when coupled with other factors, contribute and lead to M’s saving faith. It should also be noted that in order to avoid internal Christian debates, one ought to broaden experiential knowledge as revelation to include things like the Holy Spirit’s drawing M to believe (just to be clear).
P5 is crucial. It assumes, quite rightly (and as the argument later stipulates), that there are states of affairs containing M coming to saving faith (where M is one member). In that case, so long as there is at least one M who has saving faith, I agree. To be fair, it is perhaps the case that the person who created this argument intended that “For every M, if God is omniscient, then he is aware of states of affairs that would bring about the saving knowledge of God for M.” If P5 means the former, then it’s important to note the conclusions of the argument do not follow. So, charitably, we ought to assume it intends to say the modified P5.
However, in this case, I do not agree with P5. I see no reason to think that every M has a state of affairs that would bring about saving knowledge of God. Perhaps it is the case that there are some M’s who have no states of affairs in which they are instantiated where the relevant CCFs are true, sufficiently bringing about M’s free choice to be saved. Obviously, God cannot be aware of something that is not true (that is, he cannot know something as true that is not true). This is only an undercutter of the premise, and not a defeater. But if the arguer expects to convince people who do not already believe the conclusion, he will have to justify this move. So, perhaps he backs off a little and states the more modest
P5* Probably, it is the case that there are such states of affairs containing the relevant CCFs for each and every M to come to a saving knowledge of God, and, necessarily, God knows them.
But how can we know this? Probability is based on background knowledge, and CCFs are based on the closest worlds to this actual world where each M freely chooses to be saved. It’s not even clear one can know very many of his own counterfactuals. Being modest, I would say that at least it’s not clear P5* is true. It certainly appears as nothing more than conjecture.
P6 is also crucial to the debate. Again, its ambiguity seems to indicate that it means something like, “If there are such states of affairs, then if God is omnipotent, he can actualize them.” As we have seen, it may not be the case that there are such states of affairs for every M (but we could happily agree with the interpretation of P6 were this to be our only problem). However, there is another, major problem. It is epistemically possible that CCFs that make up states of affairs make those states of affairs mutually exclusive, as a matter of logic (call this “feasibility”). Suppose, for instance, that John will come to saving faith in circumstances C, and James will come to saving faith in circumstances S. Suppose further that the relevant CCFs are true so that James would not believe in C and John would not believe in S. Finally, suppose S and C are either exhaustive of the states of affairs feasible for God or exhaustive of the types of states of affairs (that is, states of affairs where John and James believe for saving faith). Thus, logically, it cannot be the case that both John and James come to saving faith, because there are no feasible states of affairs in which they both do so. Traditionally, Christians do not believe that God can do the logically impossible, as they believe that is a violation of his nature (as the foundation of truth).
Now one may protest that all I have done is stipulate something to be the case, rather than defend it. True enough. However, I only intend for this to be an undercutter. The same problem of justification and epistemology attends P6 as does P5. We should not accept either of them.
I shall assume P7 means that if God’s existence has been propositionally known in the past, then the state of affairs of some M knowing God propositionally is logically possible. It may be that the arguer intends to say that if God’s existence has been propositionally known by some M’s in the past, then the state of affairs of every M knowing God propositionally is logically possible. Although I am not entirely convinced that follows, I do believe it is at least in principle possible for every M to have such propositional knowledge. So, in any case, it appears as though we should accept P7.
P8 has a similar problem of meaning. It may mean, “if it is the case that God has made himself known experientially by some M’s in the past, then the state of affairs of some M’s knowing God experientially is logically possible.” If that is the case, we can agree: after all, we are merely describing a trivial truth (if X occurred, then X is logically possible). However, it may be that the arguer intends to say “if it is the case that God has made himself known experientially by some M’s in the past, then the state of affairs of certain M’s (or any M’s) knowing God experientially is logically possible.” While I think that is true, in that there is no logical contradiction involved (nor metaphysical absurdity), considerations of feasibility come in. It does not follow that because some M’s are in states of affairs that contain true CCFs for saving faith, that other M’s are also in that state. However, to be charitable, we ought to assume that logical possibility is taken at face value. In that case, I heartily agree with P8.
P9 is pretty straightforward and should be accepted. P10 is somewhat unclear. Is it claiming that “there are states of affairs such that some M, who does not come to saving faith, would come to saving faith, and it is the case that God has not actualized them”? Or is it simply saying that “God has actualized a state of affairs in which some M does not come to saving faith”? The latter makes no comment as to whether or not there are such states of affairs, and it seems almost obvious. The former suffers from the same epistemic-justification problem that plagued P5 and P6. Because the simpler reading of P10 does not yield the conclusions, then plausibly this former meaning is what the arguer intends. In that case, we ought to reject P10 as well.
One might be surprised to note that I will disagree with P11. If doxastic voluntarism is true, or at least true with respect to belief in God, then the degree to which God is known, propositionally, is not (at least in part) up to him. It depends, in part, on the willingness to believe. Of course, on the other hand, if all M’s really do know God exists, then P11 is true (but only at the cost of the arguer admitting he knows God exists—which is counterproductive for his purposes). As far as P12 goes, this all depends on meaning. If we mean that the extent to which God reveals himself to M’s is in the full control of God, then that is correct. I think this is the most charitable interpretation of P12, and so I will affirm it.
P13 is a little confusing. Taken in a straightforward way, it seems to indicate that there is a state of affairs for each individual M that would result (in a CCF way) in saving faith, and that this state of affairs is necessary. In that case, everyone has saving faith! But, plausibly, this is not what the arguer meant. Perhaps he meant that, necessarily, there is at least one state of affairs such that for every M, M would freely come to a saving knowledge of God. But does that mean for every possible M? For every actual M? If we take this interpretation, it suffers the same problem of P5, P6, and P10, and so ought to be rejected. Finally, perhaps it means there is a state of affairs for each individual M that functions as a necessary, though not sufficient, condition for M to come to saving knowledge of God. But in that case, the same problem attends. How do we know the relevant CCFs are true for each M, so that even if the necessary conditions are in place, the sufficient will follow? I see no reason to think that if the sufficient conditions are not true, the necessary must be. I am not sure which of these two is more charitable or plausible in terms of intention, but in either case I would reject P13.
Because of prior ambiguity, it is difficult to know what, precisely, P14 is asserting. It may be saying that if it is the case that a conjunctive state of affairs with the relevant CCFs true such that for every M, M freely comes to a saving knowledge of God is not logically possible for God to actualize, then it is not logically possible at all. Here considerations of feasibility ought to be employed, as well as a discussion of necessity, and accidental necessity. I am inclined to say that there is a logically possible world PW, such that in PW every M (that is, every M in the actual world) freely chooses to be saved. In this technical sense, we can happily agree to P14. But that is a far cry from claiming that there is such a world FPW, where every M would freely come to a saving knowledge of God. Perhaps there are no such worlds.
“Aha!” says someone out there. “So that means it’s not logically possible for those people to be saved after all!” There are two problems with this statement. First, it commits the modal fallacy of distributing the necessity of an entire state of affairs to some individual parts. Just because John is only saved in C and James in S, it does not follow that it is impossible for James to be saved, were C to be actual. The second problem is that it does not account for the necessity of a statement. If something is logically impossible, it is so in all possible worlds. But, as is quite plausible, it is not the case that any one M exists in all possible worlds. Thus, at best, what one can argue is that, for certain M’s, it is a necessity de dicto that they do not believe in any possible world.
For P15, it seems to be stating that if some action X is a logical impossibility for M, then M cannot be held responsible for not doing X. It is the “ought implies can” principle, and I agree with it wholeheartedly. Therefore, I endorse P15.
Notice now that either the conclusion C1 does not follow, or it is wholly innocuous. For under our interpretation, we deny P5, P6, P10, P11, P13, and possibly P14 (though we will accept it with the notes above). Ostensibly, all of these are needed in order to yield C1 (and certainly the denial of five of them will be sufficient to avoid C1). It is basically a false dilemma. The third option is that if non-belief occurs, it is because the relevant CCFs that need to be true for M to come to a saving knowledge of God are not true in any world in which M is instantiated. On the other hand, we can decide to take other interpretations of the disputed premises, and all that follows is that God has actualized a state of affairs in which some M does not come to saving faith, which, of course, makes absolutely no comment on whether or not God could have done so, whether there are any such states of affairs, which CCFs are true, etc.
Because of denied premises, C2 also fails. C3 does not follow, either because we have rejected P14 (because of the modal fallacy), or because P14 is true but P14 and P15 do not yield C3, or because of the fact that necessary conditions do not entail sufficient conditions, and/or because the relevant CCFs are not true precisely because of M’s choice or would-be choice (or some combination of these). It all depends on the meaning of P14, but in any case, the intent of C3 can be avoided.
P16 can be regarded as true, in that if the Christian God exists, then whomever withholds belief (in the saving sense) is culpable for having done so. However, since other premises have been denied (and in turn their respective conclusions), C4 does not follow. We have examined each of these premises in this interesting argument. We have attempted to be as charitable as possible in interpreting them. We have shown that many of them are fundamentally flawed, or require further explanation or some strengthening in order to be taken as true. Because of this, I would say this argument against God’s existence can be regarded to have failed.
 If one doubts this, she can plug in this “fleshed-out” definition into P5 and run the argument.
 I am also assuming, for clarity’s sake, that “would bring about” is not causal, but more akin to Molinist counterfactuals of creaturely freedom (CCF). If “would bring about” is essentially describing causal mechanisms, we must reject even this modified P5, for God desires not automatons, but willing participants in the relationship.
 It’s worth pointing out that if God “made his existence propositionally known,” it’s difficult to see how this is different than “revelation,” or experiential knowledge (according to D2 and D4), and thus P7 does not differ in content from P8.
 Indeed, it may be the case that every M does have such knowledge. See my article, “Do Atheists Know God Exists?”
 In this case, what is logically impossible is not the individual state of affairs of some M coming to saving faith, but rather the conjunctive state of affairs of competing CCFs (as discussed earlier). No logical impossibility should be distributed to the individual conjuncts.
 Obvious, that is, except to universalists.
 I once had an atheist insist he was absolved from believing in God because belief in God is necessary for salvation. He argued that he should have at least that. My response is that if someone shoves you out of a plane (without a parachute) from 35,000 feet, assuming you will die, will it become better if you are shoved out at 30,000 feet? No. The same result will happen, even though the necessary condition (not being shoved out of a plane from 35,000 feet) for survival in that case will have been accomplished.
 Notice that even this does not follow from the objection. For perhaps it is the case that there are worlds in which M believes, but that in PW (or FPW) the relevant CCFs are not true for M.
 There can be variants to this third option, so that really there are more than three.