Can God know what it is like to learn? If God is omniscient (all-knowing), then it seems that he would have to know what it is like to learn. However, in order to know what it is like to learn, one must have learned something, which involves moving from a state of not-knowing to a state of knowing. This entails that at one time we were in a state of not-knowing a thing that was learned, then experienced what it is like to learn. But if God is essentially omniscient, he always is and has been omniscient, so was never in a state of not-knowing. Because being in a state of not-knowing is necessary to know what it is like to learn, we would seem to have to say that God does not know what it is like to learn. But this contradicts the original claim that he does know this based on his omniscience. Thus, it seems that God’s omniscience generates a contradiction. Consequently an omniscient God cannot exist.
II. A Formal Presentation of the Argument
The argument above can be rigorously stated through the following reductio ad absurdum:
(P1) An essentially omniscient being, God, exists. (assumption for indirect proof)
(P2) God is and always has been omniscient. (from P1)
(P3) A being’s omniscience entails, among other things, that it has all experiential knowledge. (necessary truth)
(P4) Having all experiential knowledge entails knowing what it is like to learn. (necessary truth)
(P5) God knows and always has known what it is like to learn. (from P2-P4)
(P6) Knowing what it is like to learn entails having learned something. (necessary truth)
(P7) Having learned something entails that one has gone from a state of not-knowing to a state of knowing. (necessary truth)
(P8) God has gone from a state of not-knowing to a state of knowing. (from P5-P7)
(P9) There was a time when God was in a state of not-knowing. (from P8)
(P10) God has not always been omniscient. (from P9)
(P11) God has always been omniscient and has not always been omniscient. (from P2 & P10)
(C) Therefore, God does not exist. (from P1-P11)
III. Possible Objections to the Argument
(1) The first possible place of contention is P1, where a critic could reject committing to God’s essential omniscience—the notion that, as an essential property, omniscience necessarily (and thus always) belongs to God. This imagined critic might think that at one time God was almost omniscient, and shortly thereafter acquired his last two pieces of knowledge—X, and what it is like to learn. However, this strange position has no obvious candidate for X, and in any case does not seem to be a real threat to the argument because P1 is a necessary truth by stipulation of the traditional conception of God as essentially omniscient.
(2) Another possible place of contention is P3—perhaps experiential knowledge (knowing what it is like to experience something) is not included in omniscience. For instance, perhaps omniscience only entails having all propositional knowledge. In that case omniscience only entails that a being knows all true propositions and believes no false ones—not that it has all experiential knowledge. However, this objection questionably assumes that experiential knowledge does not count as a genuine kind of knowledge, contrary to our common linguistic practices. For instance, assertions like “I know what it is like to lose someone close to me,” “I know what it is like to be nauseous,” or “I know what it is like to be an atheist” seem unproblematic. We also quite frequently talk about learning what something is like, which implies acquiring information or knowledge about the world. Moreover, people can sometimes participate in discussions based on similarities in experience, while those lacking the relevant experiences cannot participate in the same way. This suggests that the “experienced” individuals possess genuine information (or knowledge) about the world that the “nonexperienced” individuals do not possess. Thus there is a strong presumption that experiential knowledge counts as genuine knowledge, and this should be taken for granted until good epistemological grounds for rejecting it are provided. Until then, the claim that omniscience can only involve propositional knowledge is unjustified.
A critic might also contest P3 by rejecting a literal definition of “all-knowing.” Perhaps we should impose logical boundaries such that “all-knowing” means “knowing everything that is logically possible for a given being to know.” This revised definition of omniscience entails the following:
(P3′) A being’s omniscience entails, among other things, that it has all the experiential knowledge that it is logically possible for it to have.
But this allows for the falsity of P3: it may be the case that some experiential knowledge is logically impossible for an omniscient being to have, so a being’s omniscience may not entail that it has all experiential knowledge. On the other hand, this revised definition of “all-knowing”—besides being counterintuitive—has a major drawback: it allows for inanimate objects that know nothing, or beings that know almost nothing, to count as omniscient beings. Rocks and grains of sand cannot know anything by definition, so they know everything that is logically possible for them to know—and so would count as omniscient on this revised definition. Or consider the being McIgnorant, who by definition has extremely limited knowledge, yet is still “omniscient” because it knows everything that is logically possible for it to know. Obviously, these are not true examples of omniscience, so the revised definition of omniscience should be rejected in favor of the literal definition, leaving P3 secure.
Furthermore, even if the revised definition of omniscience were correct, making P3 dubious, my argument could be salvaged by introducing the following proposition:
(L) It is logically possible for God to know what it is like to learn.
This proposition, when conjoined with P2 and P3′, entails P5, so the remainder of the argument remains intact. As such, the rejection of P3 based on revising the definition of omniscience is not sufficient to undermine my argument; instead, the rejection of L is also required. However, rejecting L has some serious drawbacks. For starters, if it is logically impossible for God to know what it is like to learn, then he cannot know something that (a) all humans can know and (b) almost all humans do know. But this conflicts with the following premises:
(Q1) An omniscient being knows everything that nonomniscient beings know.
(Q2) God knows everything that humans know.
These premises are not only intuitive, but they have the feel of conceptual truths; it makes little sense to think that nonomniscient beings (like humans) could know something that an omniscient being (like God) would not know. It is much more natural to think that a being who does not know something that is known to a nonomniscient being is itself not omniscient, for there is at least one genuine piece of knowledge that it does not possess, even if it is exponentially more knowledgeable than other nonomniscient beings. What’s more, the denial of L entails that God does not know something that is extremely easy for us humans to know, something that even the least intelligent humans can and do know. It is very strange, to say the least, for an extremely intelligent being like God to be incapable of knowing something that we humans can easily acquire, and that even the least intelligent humans know. Of course, these odd and seemingly senseless consequences of denying L could be true because L (as far as we can tell) could be false; but given such bizarre consequences, it is more plausible to avoid them by accepting L, which in turn leaves my argument intact.
(3) The only other place of contention that I can foresee is P6, despite its strong intuitive appeal and supporting empirical evidence. Conceivably, a critic could argue, a being could know what it is like to learn without having learned something; and perhaps God’s mind contains the experiential knowledge of what it is like to learn without him ever having had to learn something. For if having the knowledge of what it is like to learn is a certain state of mind, then perhaps God’s mind (unlike ours) has this state of mind—a state of knowing what it is like to learn without ever having to learn anything. Thus, P6 is conceivably false.
In response, we might argue that the falsity of P6 is not a genuine possibility because P6 is a conceptual truth. Its antecedent, “knowing what it is like to learn,” refers to having experiential knowledge of learning—i.e., experiential knowledge with the activity of learning as its content. Experiential knowledge, by definition, is knowledge of having or going through an experience; so it requires actually going through an experience. Thus, experiential knowledge of learning requires having the experience of learning; and because learning is an activity in which something is learned, having the experience of learning is nothing more than having the experience of learning something. Moreover, this (genuine) experience of learning something has, quite obviously, learning something as its content. This means that having this experience requires the learning of something. Therefore, the consequent of P6 (“having learned something”) appears to conceptually follow from its antecedent (“knowing what it is like to learn”), which would make P6 a conceptual truth.
This short paper attempts to demonstrate that atheism is true by deriving a contradiction from the assumption that God, an essentially omniscient being, exists. After presenting the argument, I examined possible objections to it and found none of them to be forceful. Thus, until strong argument to the contrary can be offered, I conclude that my argument is sound: not only does God not exist, but it is not even possible for him to exist.
 Restricting God’s omniscience in this fashion might be problematic, but if so it would be no help to me since it does not vindicate P3. So I will not rehearse this kind of argument.
 It should be noted here that knowing what it is like to be an atheist is certainly something that God cannot know, which could provide another argument (bordering on the comical) for God’s nonexistence based on the impossibility of omniscience.
 Vrinda Dalmiya and Linda Alcoff, “Are ‘Old Wives’ Tales’ Justified?” in Feminist Epistemologies ed. Linda Alcoff and Elizabeth Potter (New York, NY: Routledge, 2003), p. 239.
 Dalmiya and Alcoff aptly call the unjustified dismissal of nonpropositional knowledge as genuine knowledge “epistemic discrimination.”
 Patrick Grim, “Impossibility Arguments” in The Cambridge Companion to Atheism ed. Michael Martin (New York, NY: Cambridge University Press, 2007), pp. 207-208
 Michael Martin, Atheism: A Philosophical Justification (Philadelphia, PA: Temple University Press, 1990), p. 289.
 I would like to thank an anonymous reviewer for helpful comments on this paper.
Copyright ©2010 Ryan Stringer. The electronic version is copyright ©2010 by Internet Infidels, Inc. with the written permission of Ryan Stringer. All rights reserved.