In an earlier post, I argued that the definition of omniscience as knowing every truth and believing nothing but truths is insufficient for omniscience because an omniscient being would also be certain, and knowledge of every truth does not guarantee certainty of every truth.
Here’s another thing that the definition leaves out. Normally, when we say that someone knows or believes p, we are talking about non-occurrent knowledge. We say things like: “Alice knows the atomic number of carbon”, even while Alice is not thinking about carbon. However, I think an omniscient being—one that enjoys the perfection of knowledge—will need to have occurrent knowledge of all truths. Moreover, the omniscient being will need to always be attending to every piece of that knowledge to a maximal degree. (It is not a perfection in us to attend to everything we think maximally, because for us attending to one thing often excludes attending to another. But that’s due to our imperfection.)
2 comments:
Hello!
I have been recently contemplating on Milne's paradox (see Milne's 2007 paper "Omniscient beings are Dialetheists", and it seems very worrying for classical theism where we want to preserve omniscience. It argues that:
Sentence S: No omniscient being knows S
It argues this sentence (or the proposition expressed, more precisely) has a truth value, since, under atheism, it is clearly true (there are no omniscient beings under atheism). So, it has a truth value. However, under theism, if it is true, theism is false, (God is not omniscient) and if the sentence is false, then some omniscient being knows S, but since knowledge is factive, S is then true. So, theism is going to have a contradiction. The trouble is, when you try to deny it has truth-value, we can always argue since atheism is logically possible, it is logically possible it has a truth value, and it does not seem that anything would make a sentence's (or proposition's) truth value change across worlds.
This has been quite a starking and worrying issue I do not know how to deal with without being a dialetheist, which I deny, so I'd appreciate your thoughts!
I haven't read the paper.
I used to worry a lot about S: "God believes S is not true."
But while it's a tough thing technically to resolve, it's not very different from the good old "contingent liar". You erase a board, and write in the middle: "The first sentence on this board isn't true." Well, since that's the first sentence, it's true iff it's not true, it seems, a contradiction. But it also seems to have a truth value, since without changing the sentence at all, it would be true if you wrote "2+2=5" at the top of the board.
I think the lesson of all such things is deep: Whether a sentence expresses a proposition depends on external circumstances. Whether "No omniscient being knows this" expresses a proposition depends on whether there is an omniscient being and whether "The first sentence on this board isn't true" expresses a proposition depends on what else if anything is on the board.
Here's another way to see basically the same point. Compare:
S: No omniscient being knows S.
S': Either there is no omniscient being or S' is not true.
Now a bit of logic shows that for any T, "No omniscient being knows T" holds if and only if "There is no omniscient being or T is not true" holds. Thus, there shouldn't be any significant difference in what you can argumentatively accomplish with S or with S'. But if S' is a good argument against an omniscient being, then the following is a good argument against evolution;
S'': Either there is no evolution or S'' is not true.
But S'' is not a good argument against evolution.
Post a Comment