Wednesday, January 18, 2023

Probability, belief and open theism

Here are few plausible theses:

  1. A rational being believes anything that they take to have probability bigger than 1 − (1/10100) given their evidence.

  2. Necessarily, God is rational.

  3. Necessarily, none of God’s beliefs ever turn out false.

These three theses, together with some auxiliary assumptions, yield a serious problem for open theism.

Consider worlds created by God that contain four hundred people, each of whom has an independent 1/2 chance of freely choosing to eat an orange tomorrow (they love their oranges). Let p be the proposition that at least one of these 400 people will freely choose to eat an orange tomorrow. The chance of not-p in any such world will be (1/2)400 < 1/10100. Assuming open theism, so God doesn’t just directly know whether p is true or not, God will take the probability of p in any such world to be bigger than 1 − (1/10100) and by (1) God will believe p in these worlds. But in some of these worlds, that belief will turn out to be false—no one will freely eat the orange. And this violates (3).

I suppose the best way out is for the open theist to deny (1).


IanS said...

I’d reject (1) in any case. Perfectly rational beings would have no use for belief. They would have consistent credences (which could be 0 or 1) for everything. On this view, beliefs are a shortcut that we, with our limited mental capacity, use to reduce processing load.

The Great Thurible of Darkness said...

Speaking of "God's beliefs" seems about as coherent as discussing what God ate for lunch. He's not a finite creature.

Walter Van den Acker said...


As Ian says, (1) applies to finite creatures, it doesn't apply to an omniscient being. On open theism, God knows everything that can be known, but libertarian free choices cannot be known in advance. If God were to "guess" what I will do tomorrow, of course on open theism, he could turn out to be wrong, but it would be a guess and not a belief.

Alexander R Pruss said...


I am somewhat inclined to say that a high credence just *is* what a belief is.


Like you and Alston, I am inclined to agree that "belief" is at best an awkward way to describe God's cognition. However, I think a major part of the reason why God doesn't fundamentally have beliefs is that there is no room for a gap in God between cognition and reality. Whether an open theist can have the "no gap" view of God depends on the variety of open theism. If the open theist is an open futurist, who thinks there are no facts in reality about future contingents, then they can have the "no gap" view. But an open theist like Swinburne or van Inwagen who thinks there are facts about future contingents and God doesn't know them thinks there is a gap between God's knowledge and reality, and now God is sounding very much like us -- there is stuff that God is uncertain about.