A Probabilistic Reverse Argument from Fine Tuning

I was reading the thread “A Reverse Fine-Tuning Argument” which makes the argument that Theism is logically inconsistent with the FTA; that is: if fine tuning is true theism is false. A distinct but related version I have often heard and would like to discuss is:

Evidence that the universe is fine tuned improves the probability that naturalism is true and reduces the odds that theism is true. I will address a few of the standard objections after presenting the argument.

Please note that this argument is completely independent on any theory about how the universe came into being under either the theist or non-theist options. Posts on the merits of various current or past cosmology theories are off topic. The views of cosmologists are also largely off-point. The universe is here. If we rule out God did it or Nature did it prior to examining the proof we are begging the question. One may believe God does not exist or that nature cannot account for life but that must be suspended as one looks at the proof or why bother having the discussion.

P1: If God as traditionally defined in Christianity exists then God is all-powerful (i.e. he can do anything that is not logically impossible)

P2: This God could create any of an infinite number of life sustaining universes (e.g. the world could be a flat disk, with four pillars that hold up the firmament which has water above it, the universe could be centered around the earth with the planets and stars on fixed spheres, there could be only our solar system, there could be life on all planets, …… Since God is all-powerful this list has an exceedingly large number of options)

P3: The chance that God would select this particular way of instantiating the universe is very small (note this is positioned as the parallel claim to, “since there are lots of potential values of the fundamental constants the chance of the set we see is very small”, in FTA)

P4: Under naturalism the only universes that are likely to support life are very old and very large and would have few places where life could be found

P5: Our world is very old and very large and has few places where life could be found

P6: Since P5 is true, it is more likely we are under naturalism where this (a very old, very large universe with certain values for cosmological constraints) had to be the case than under Theism where this is only one of many options.

The rationale behind P6 can be shown by analogy of a two jar experiment.  Assume we have two jars the first has thousands of balls one black the rest white; the second jar has only a single black ball. Each ball black or white represents a life enabling universe. We are told a black ball has been drawn and asked which jar it most likely came from. It is clear that the evidence of having drawn a black ball favors that the ball was drawn from the second jar.

The first objection to this proof is God could have made the universe this way. While this is true unless you have some evidence as to why that would be the case it is irrelevant as that fact was already considered in the argument (i.e. the one black ball in the first jar)

The second objection is God may have had a very good reason that we are not aware of for making the universe this way. Again, true but irrelevant unless you have evidence to support this claim. Stating that the universe is this way so God must have had a reason for doing so is begging the question. Saying God works in mysterious ways does not affect the argument.

The third objection is God did not want to interfere with free-will (or some other reason) so he remains hidden by picking the single model for the universe that naturalism would have allowed for. First this implies that naturalism can account for this universe which negates the FTA used to demonstrate God. Second It make the naturalist and theist positions equivalent on the evidence of our universe but makes God’s role in determining the fine tuning susceptible to Occam’s razor arguments.

The forth argument is that is the world was God’s only (or few) logical option(s) since if the universe did not behave in a rational (naturalistic) and predictable way we could not trust our judgments and have reasons to accept God. Note this is a modern argument as it would not apply to the majority world-view though out pre-modern history. This objection is just a special case of the third objection above since we can easily have consistent local behavior in any of the potential universes described in P2 above. It also indicates that miracles would somehow undermine our ability to accept God which is inconsistent with claims from the Bible including the resurrection.

MorleyMcMorson:

Thank you for your thoughts on this

1a. “The range of constants doesn't change given either naturalism or theism”

Actually this is only true if you accept our “natural laws” of physics are impossible for God to change. I see no reason to assume that God is limited to universes that manifest Einstein’s general relativity or Maxwell’s equations.  If God can create many other systems of “natural laws” then the range and even the particular constants required can and would change under Theism. But I do not feel this is central to the proof as it seems to hold even if the range of constants is the same for both.

1b. “you'd need to show that God would have just as much reason…”

I claim ignorance about any of God’s reasons but this type of ignorance (where I have no reason to assume any particular outcome is more likely than another) is usually dealt with through the use of a uniform distribution function over the sample space which results in any particular outcome being equally likely (unlikely).

2. “It's very questionable whether the probabilities you're using are the relevant ones for discussion. Why wouldn't they be P(L)/T vs. P(L)/N?  Or something else? “

This particular argument is about the impact of the fact that fine-tuning is observed on the T hypothesis and the N hypothesis. So I think they are the right probabilities to discuss. One could look at P(L)/(T) vs P(L)/(N) but that would be a different proof and discussion.

1b. “you'd need to show that God would have just as much reason…”

I claim ignorance about any of God’s reasons but this type of ignorance (where I have no reason to assume any particular outcome is more likely than another) is usually dealt with through the use of a uniform distribution function over the sample space which results in any particular outcome being equally likely (unlikely)