Friday, July 26, 2013

Fine Tuning and Probability

An interesting issue which illustrates the perils of probability theory – and, strange to say, Keynesian radical uncertainty – is the so-called fine tuning argument from design.

Essentially, this is the view that the universe’s physical and chemical laws appear to be fine tuned, in the sense that if only slight changes occurred in one of the various constants, the universe could not support life. Since it is deemed extremely improbable that the universe would be so finely tuned for life, this is (supposedly) evidence of a designed universe created by some kind of intelligent designer.

It is no surprise that various theists and theistically-inclined philosophers and scientists are partial to this argument.

There are many problems with it. First, there is the counterargument that the universe does not appear to be optimally fined tuned for organic life at all, but perhaps only minimally fined tuned. After all, the vast majority of the universe consists of the mind-bogglingly vast vacuum of intergalactic and interstellar space: this is filled with electromagnetic and particle radiation that is lethal to carbon-based life (except perhaps for the most robust microorganisms).

Secondly, it is far from clear that the emergence of self-replicating molecules and the subsequent evolution of complex carbon-based life is either easy or particularly common in our universe.

Thirdly, even if we could grant that there might be some convincing inductive argument in support of an intelligent designer, why would that designer have to be supernatural? Why not a being or beings who were non-supernatural and who merely arose by Darwinian evolution in another universe? (for example, perhaps such beings may have designed our universe to create stars and black holes, and life was a highly improbable accident, and so on).

But, even if all these concerns are put aside, the overwhelming problem with the fine tuning argument from design is, quite simply, the probability issue. How can anyone ever prove that the current universe with its physical laws is in fact extremely improbable in an objective sense?

This can be stated as follows:
(1) In order to calculate an a priori probability one would need a list of the complete set of possible universes (in a sample space) that might arise from the antecedent conditions before the Big Bang to the point when laws are formed. And not just that, but we must know that all outcomes are equiprobable as well. However, there is not a shred of evidence that anyone can provide either of these things: it is clear that you cannot construct a priori probabilities.

(2) what about relative frequencies? It is utterly impossible to conduct empirical tests of relative frequencies of outcomes in trials because that would require experiments in which we create the universe over and over again to see the types of universes created, and then to obtain a set of such outcomes and ascertain whether stable long run relative frequencies for outcomes occur. Hence no objective relative frequency probability can be constructed for the present universe either.
In short, the very idea that the present universe’s laws are highly improbable is unproven speculation. For all we know, our present universe might be highly probable; or might have a probability of 1; or only be moderately improbable, and so on. At the moment, however, we cannot really say.

The same thing can be said of the question:
“Why is there something instead of nothing?”
This very question presupposes that it is highly probable there should be nothing instead of something. But we have not proven that assumption at all. Perhaps, after all, it is extremely probable that there should be something like the universe instead of nothing (for example, once we get into the strange world of quantum mechanics, with its virtual particles and quantum fluctuations, strange things appear to happen).

Of course, many philosophers and skeptics are well aware of what I have said above, and many years ago the analytic philosopher C. D. Broad made the same point in a critique of the philosophical theologian F. R. Tennant:
“The Design Argument really makes two uses of the notion of antecedent probability. It has to contend both that it is antecedently improbable that the world should be such as it is without being the product of the design, and that the existence of a world-designer has an appreciable antecedent probability. Now, as regards the first point, I cannot see that Dr. Tennant has answered the objection by his distinction between ‘mathematical’ and non-mathematical probability. Is there any sense of probability, mathematical or ‘alogical,’ in which a meaning can be attached to the statement that the antecedent probability of one constitution of the world as a whole is greater than or equal to or less than that of any other? I very much doubt if there is.” (Broad 1930: 479).
These days sophisticated proponents of the fine-tuning argument from design are quick to admit that their probability estimates are not objective, but merely epistemic probabilities derived from inductive reasoning.

But even an epistemic probability depends on the amount or degree of relevant knowledge (Skyrms 2000: 23), and if this isn’t an example of an issue where we face something close to radical uncertainty (in the Keynesian sense) – where the weight of evidence must be very small (perhaps close to 0) – then I don’t know what is.

BIBLIOGRAPHY
Broad, C. D. 1930. “Philosophical Theology, Vol. II.: The World, the Soul, and God, by F. R. Tennant,” (Review) Mind n.s. 39.156: 476–484.

Shotwell, David A. 1987. “Is the Universe Improbable?,” Skeptical Inquirer 11: 376–382.

Skyrms, B. 2000. Choice and Chance (4th edn.). Wadsworth, Belmont, CA.

12 comments:

  1. There is a more fundamental counterargument. In The Anthropic Principle Does Not Support Supernaturalism, Michael Ikeda and Bill Jefferys argue that the fine tuning argument confuses prior and posterior probability.

    Ikeda and Jefferys argue that fine tuning is an argument against a designer. There are more "supernatural" creators who could create life without creating life-friendly physical constants because a supernatural being is by definition not bound by physical law. Thus, the observation that would definitively distinguish a supernatural creator from a purely natural universe would be the observation that the physical universe was not life-friendly, which is, of course, opposite of the observations actually made.

    ReplyDelete
  2. Oooo. Not aware of this debate. But let me play Devil's advocate for fun (or should that be God's advocate in this case?). Okay, point by point then.

    (1) This would not be evidence against a Creator. Christianity, for example, is founded on the idea that the Universe is imperfect after the Fall. Indeed, such imperfection would actually be a good case for a Christian Creator. (I would imagine that other religions also see the universe as imperfect too, but I don't know enough to substantiate).

    We could go further on this too. We could take the Gnostic position and say that such is evidence of the impurity of physical matter.

    (2) There really is no evidence either way. Also, a theist would just say "Well, we exist, right? That's enough".

    (3) I don't think that makes much sense. Where would these beings have come from? Who created them? And so on. It is an infinite regress argument.

    As to the probability issues, I don't really know enough about this. But the consensus among physicists appears pretty extensive on this issue. Of course, as you know, I'm partial to the argument that these physicists are actually just thinking in abstractions and they are really dealing with metaphysical issues when they talk about such things. But I've been saying that all along. ;-)

    ReplyDelete
    Replies
    1. "But the consensus among physicists appears pretty extensive on this issue. "

      But it is not a consensus about objective probabilities. It is about the parameters or universal constants and how changing them would mean stars could not exist and so on: e.g., if the strong nuclear force was weaker than it is, then no element except hydrogen would exist in our universe.

      But this tells nothing about the objective probability of the current universe.

      Delete
    2. It sounds to me like a metaphysical argument then. Basically they're looking at the fine balance of the universe -- and come on, even I've thought in the past that there is a fine balance and an order that is rather impressive -- and then saying that this is pretty awe-inspiring. That's what the metaphysical argument has always been: "Look at the order in the universe. There must be some force behind this." And so on. But they'e just couching this awe in the language of probability. I.e. rather than saying that the order is awe-inspiring they are saying that it is highly improbable.

      Delete
    3. This would not be evidence against a Creator. Christianity, for example, is founded on the idea that the Universe is imperfect after the Fall.

      This is a bad counterargument because it renders the theistic position completely unfalsifiable. Any logically possible observation is compatible with creation+fall. Yes, theories need to change according to the evidence, but good science demands that the changes make the theories narrower, not broader.

      Delete
  3. 1) Our claim isn't that this universe contains the most amount of life that you could fit in it. We're saying that if things were slightly different, there wouldn't be any life at all. Saying how horrible things have been and will become doesn't change the precision of the fine-tuning. There are many Bible verses saying bad things will happen before the end. Luke 21:11, Luke 21:25, Matt 24:29, Matt 24:7, Luke 17:29, Mark 13:25, Rev 6:12-13, 2 Peter 3:10-12.

    2) http://www.is-there-a-god.info/clues/designfacts.shtml

    3)Physicist Stephen Hawking says, "It appears that the fundamental numbers, and even the form, of the apparent laws of nature are not demanded by logic or physical principle.” Also you would have to believe that a life-prohibiting universe is physically impossible, but a life-prohibiting universe is logically possible and there is no evidence suggesting otherwise. Saying the universe had to take a form suitable for life is ridiculous, which is why physical necessity has few, if any supporters.

    ReplyDelete
    Replies
    1. "Our claim isn't that this universe contains the most amount of life that you could fit in it. We're saying that if things were slightly different, there wouldn't be any life at all"

      And there is still no adequate evidence that the present universe is highly improbable, which is the point you have clearly missed.

      Delete
    2. Our claim isn't that this universe contains the most amount of life that you could fit in it. We're saying that if things were slightly different, there wouldn't be any life at all.

      This is too weak of a claim for a supernatural and perfect creator.

      Also you would have to believe that a life-prohibiting universe is physically impossible, but a life-prohibiting universe is logically possible and there is no evidence suggesting otherwise.

      This does not follow from naturalism. According to LK's position, saying that that probabilities are epistemic rather than ontological is only a reason to be skeptical of drawing any unusual conclusions about the universe from probability theory. To use probability theory at all to prove anything, we must at least have some knowledge of a real distribution with various features. By definition, we do not have any knowledge of the distribution of our ignorance.

      And, of course, as Ikeda and Jefferys show, even if we were to grant most of the the premises demanded by the FTA, then given the posterior observation of life existing, it is more likely that the universe is natural.

      Not only, as LK points out, do we lack sufficient evidence to take the FTA seriously, it is actually fundamentally logically invalid, or requires inference rules fundamentally different from those of standard probability theory.

      Delete
    3. The Ikeda/Jefferys argument is criticized at this link--

      https://letterstonature.wordpress.com/2010/11/05/what-do-you-know-a-fine-tuned-critique-of-ikeda-and-jefferys-part-2/

      Delete
    4. That person (and presumably you as well) does not understand probability at all. I would try to explain the difference between prior and posterior probabilities, but since I've never seen an apologist who's actually interested in things like math and truth, the effort would probably (pun intended) go to waste.

      Delete
  4. Why is there something instead of nothing?

    This very question presupposes that it is highly probable there should be nothing instead of something....


    The more serious - and surely misguided - presupposition is that there is something, "instead of nothing". :-)

    ReplyDelete
  5. You don't understand the argument - that's the problem... Read Feser or Aquinas.

    ReplyDelete