I’ve been reading a lot about metaphysics and modality, especially in line with Graham Oppy’s work. One of his key premises behind preferring naturalism over theism is that when it comes to global causal reality, naturalism has the same explanatory power as theism but is simpler.

For example, in regards to fine tuning, even if it is incredibly improbable for the constants that lead to life to be the way they are, it may be just as improbable (if not moreso) for God to have tuned the world in the way that He did given His omnipotence. Why did He choose to do X instead of Y? As such, the improbability remains very low, but naturalism is simpler. If one posits that God has an inbuilt preference for creating life in this way, then one can also posit that naturalism has an inbuilt law like preference for creating the constants in the way they are. Given no independent evidence of a God fine tuning constants, one can still consider that some naturalistic law does the same “explanatory work” without any independent evidence as well. Since naturalism posits fewer kinds of entities, it then remains preferred.

What’s interesting to me about this strategy is that the same response could be made to a voice in the clouds saying “Hi, I am God.” As long as it can’t be ruled out that mindless processes couldn’t create that voice, one can use the same line of reasoning to state, “Well, this voice may be very improbable, but it is just as improbable given God’s free choice and omnipotence. And if God has particular preferences, perhaps nature has those same preferences as well. But naturalism posits fewer kinds of entities; thus, it should be preferred.”

Something seems very counter intuitive about this reasoning and yet I also can’t find anything wrong with it. Would this reasoning apply?

  • 1
    I would never blithely accept a voice in the clouds proclaiming itself to be God. Even from a theistic angle, I would be reluctant to accept that on its face, for I would be wary of the possibility of demonic deception. Or I might wonder about a technological production of the phenomenon (by secretive humans or offworld aliens, etc.). Now, if the voice made various threats and promises that were reliably carried out, with no apparent chance at resistance, my trust in the voice (as the voice of a sufficiently powerful being to be God) would increase, but might never reach 100% certainty. Dec 9, 2023 at 18:51
  • 1
    There was(!) a sudden voice in the clouds, according to the gospel of S. Mark, Chap. 1,10-11 after Jesus' baptizing: "And straightway coming up out of the water, he saw the heavens opened, and the Spirit like a dove descending upon him. And there came a voice from heaven, saying, Thou art my beloved Son, in whom I am well pleased." - The theist and the atheist assess totally different the historical value of this passage .
    – Jo Wehler
    Dec 9, 2023 at 19:16

5 Answers 5


Don't trust a voice from the clouds: that is trivially simulated with a balloon and an MP3 player.


If there was a sudden voice in the clouds saying “I am God”, would that be evidence?

Yes. But with only that evidence, a number of different stories can be created that are equally likely. The less evidence there is, the more fantastic the explanation can be:

  1. God
  2. Extra terrestrials
  3. Donald Trump
  4. CIA
  5. Prankster
  6. Swamp gas
  7. Weather Baloon
  8. etc.

The length of this list is beyond my ability to imagine.

  • Little evidence for #3, not sure if that is good or bad.
    – Scott Rowe
    Dec 10, 2023 at 13:24
  • 2
    @ScottRowe There is some overlap between 3, 5 and 6. ;-) Dec 10, 2023 at 18:22

On many, or even most, philosophically refined conceptions of God, a sound in the sky to the effect that it sounds like something is saying, "I am God," would not be especially trustworthy. It might not be untrustworthy by default, either; but at any rate, see the SEP article on divine revelation for various more promising theories of communication "from on high." As it stands, the sentence seems parathetic, like, "I am false," or, "I am unprovable," etc., and anyway unbecoming of something as awesome (and confusing!) as the simplicity of the divine infinity.

A better sentence to be enunciated from the clouds would be, "I am a sentence whose existence is caused specifically by God, to express God's will," but we should wish for this sentence's existence to be followed by further displays of intentional power, to corroborate the sentence's yet-self-referential manifestorium of the divine nature. For example, if the sentence were a sort of "Boltzmann statement," it could congeal out of the sonic-entropic ether "just like that," to be followed by no compelling evidence that it was anything but such a congealment.

Or imagine a sentence booming from above, "I exist," or, "God exists." How should such sound effects be actual evidence of God's existence? In the first case, we would not be immediately given the "I" in play, in the second case, we might imagine that some preacher has gotten a hold of some ultrapowerful megaphone, etc.

I would be wary of claiming that naturalism posits fewer "kinds" of entities than theism, however. Firstly, if kinds occur on various levels, we should ask: does naturalism have fewer levels? Per level, does it have less kinds, or are there as many kinds on all levels as there would be on theism? Suppose that naturalism has countably many kinds of beings to its name, and theism one more, God: yet since ℵ0 + 1 = ℵ0 anyway, both naturalism and theism would turn out to have the same cardinal amount of kinds of beings! It is not clear enough, then, that qualitative parsimony is attained on the non-theistic picture, vs. the theistic one (the situation is similar to the dialectic of ante rem and in re realism about mathematical referents: intuitionism, for example, seems to end up with infinitely many intuitions, and why is it acceptable to have so many intuitions but not so many objects of said intuition? or why is the ante rem level "going too far" when we have already gone so far as the quantum microphysical domain, with all its own "weirdness"?).


If God Creates by Speaking then Nature is (already) The Sound of God's voice. Think about it.

  • 1
    In the beginning was the Word. And the Word was with God. And the Word was God. The same was in the beginning with God. All things were made by Him, and without Him was not anything made that was made. - Pretty conclusive.
    – Scott Rowe
    Dec 9, 2023 at 21:55
  • 1
    In the beginning, when God created the heavens and the earth, the earth was a formless wasteland and darkness covered the abyss. Then God said, "Let there be light ... and there was light." - Genesis. Dec 10, 2023 at 1:44
  • Such beautiful language! Probably thanks to Tyndale and James I. People say that OM / AUM is the sound of creation, being and ending. Sometimes people hear it. Too bad experiments have never detected such (yet!)
    – Scott Rowe
    Dec 10, 2023 at 13:20

If you have one model that predicts more booming voices making divine proclamations from the clouds given the other measured conditions, and another model that predicts fewer booming voices from the clouds given the same measured conditions, and you measure a booming voice from the clouds, it's evidence in favor of the model that predicts more.

Evidence is not a synonym for conclusive evidence all by itself.

As usual we can express this question more usefully by defining a game with simple rules, instead of nebulous models that only give the vaguest statistical predictions. Suppose we have have two well-shuffled decks of 100 playing cards. One of them has 90 red cards and 10 black cards. The other has 90 black cards and 10 red cards. One of the decks is selected at random and placed face down before the player. The player is given the opportunity to bet which deck he has, then he draws one card and is given the opportunity to bet again. What are the worst odds the player should accept for the first bet? What are the worst odds the player should accept for the second bet?

  • I'm not quite seeing the connection of the cards example with the voice question. How many instances of booming voices have there been? Or was the cards one a projective thought experiment?
    – Scott Rowe
    Dec 10, 2023 at 13:23
  • The player has a model that predicts black nine times out of ten and a model that predicts black one time out of ten. This is the first time the player has played the game, so there is no history of instances, only two models which both correspond to what the player knows about reality before playing the game.
    – g s
    Dec 10, 2023 at 17:06
  • In the game, the two models are exhaustive; the models' stochastic predictions are easily and exactly computable; the difference between the models' stochastic predictions is large; and we have a known finite space of possible measurements. None of these are true about the OP's scenario, which obfuscates the main point, wanting to know how does evidence work when there are multiple explanations that predict a certain event.
    – g s
    Dec 10, 2023 at 17:09

You must log in to answer this question.