A Case Against Omniscience: Fallibility

Objection 10: If Thinker X is omniscient and thus knows it, premise (5), on which your conclusion depends, cannot be true. You don't offer any epistemically relevant reasons for doubting that 'Thinker X is omniscient' in the case where Thinker X is the God of classical theism.

Reply: My response here is an amplification of my previous responses to similar objections, namely Objections 2, 3 and 7 in § 5.

Of course, the purpose of my supporting evidence for my premise (5) cannot be to show that the God of classical theism has epistemically relevant reasons for doubting that he is omniscient. That would entail trying to demonstrate a contradiction to be true—a fool's errand. I draw readers to a key stipulation of my argument: that the variable 'Thinker X' ranges over thinkers that think itself God (i.e., that, in part, think itself omniscient). In the formal symbolic version of my argument in Section 4, I state this formally as the domain of discourse for variable 'x' being 'the set of thinkers that thinks itself God'. A mistake is to jump to the conclusion that this set of thinkers that think itself God actually contains a member who is God (i.e., who is omniscient). My premises (1) to (5) do not declare either way whether God is or is not an actual member of the set. Premise (5) only refers to thinkers that think itself God. Whether that set of thinkers includes as a member an actual God is determined by inference later in the argument. So, for me to establish the truth of premise (5), I do not need to show that God has epistemically relevant reasons for doubting that he is omniscient.

Let me illustrate my response here with this example that may help to get across the structure and sequence of my argument. First, recall my formulation of my premises (4) to (6) in the informal natural language version of my argument.

1. If Thinker X knows that 'Thinker X is omniscient', then Thinker X has no epistemically relevant reasons for doubting that 'Thinker X is omniscient'.
2. For Thinker X, there are epistemically relevant reasons for doubting that 'Thinker X is omniscient'.
3. It is false that Thinker X knows that 'Thinker X is omniscient'.

Now, just as there are people who think of themself as God, there are people who think of themself as a real witch with supernatural powers. I can repurpose the same structure and sequence of my argument against omniscience to demonstrate that it's impossible for there to be real witches. That argument is as follows:

1. If Witch-hopeful X is a real witch, then there are no reasons why Witch-hopeful X cannot cast real spells.
2. For Witch-hopeful X, there are reasons why Witch-hopeful X cannot cast real spells.
3. It is false that Witch-hopeful X is a real witch.

As with my argument against omniscience, premise (6′) is entailed by (4′) and (5′) using the inference rule, modus tollens. Also, as with my argument against omniscience, I can support premise (5′) inductively. Here, I support premise (5′) with some commonly accepted facts about how we know the causes of disease, the causes of crop failure, the causes of trips and falls, the causes of romantic liaisons and of all of the other events posited to be caused sometimes by witch-believers. I add as further evidence our established knowledge of how witch stories originated, what we know about the Pygmalion Effect and so on. All of this constitutes not logically conclusive evidence, but, none the less, convincing inductive evidence in support of premise (5′).

I now show this argument to superstitious Pauline (with some introductory premises and a formal conclusion). Pauline objects that my premise (5′) is false because my premise ignores the case where Witch-hopeful X is a real witch (let's call her 'Sarah'). If my argument took into account Sarah, she explains, then I would see that for a real witch (Sarah), there are no reasons why that Witch-hopeful X cannot cast real spells. How convincing do you find Pauline's excuse that real witches are possible in spite of my argument against witches really existing?

As with objectors to my argument against omniscience, Pauline misunderstands the logical structure of my argument against witches. Pauline assumes from the start that there is a real witch (Sarah) that is a member of the set of Witch-hopefuls. But the question about the existence of a real witch is what rationally ought to be decided after reviewing the evidence for premise (5′) and not before.

In the face of my explanation, Pauline could revise her objection to declare premise (5′) false simply on the basis of the possibility of real witches existing. But this ignores the fact that possibly real witches that are not actual are not members of the set of Witch-hopefuls. Only real Witch-hopefuls (whether a real witch or not) are candidates for members of the set.

Pauline could go on further to object that the inductive evidence I offer in support for premise (5′) can't be accepted as true simply because it's possible that real witches exist. But that is to beg the question—supporting the falsity of the proposed evidence solely with the bald denial of the evidence.

The reasons for why these strategies fail to refute the argument against witches mirror the reasons for why these moves do not succeed in disproving my argument against omniscience. Both strategies mistake how members are allocated to the class defined by the respective domains of discourse. If one rejects the force of Pauline's objections to the argument against witches, then, likewise, one ought to dismiss the veracity of these kinds of objection to my argument against omniscience.

Copyright © 2022 Leslie Allan