Tuesday, June 24, 2008

Proving the deity - Or Bayesian Bollocks?

Gil Gaudia (‘God: Lies, Damned Lies and Statistics’, The American Atheist, Sept. 2007) merits kudos for lacerating the specious premise of Stephen Unwin’s pompous book ‘The Probability of God: A Simple Calculation that Proves the Ultimate Truth’ .

The core issue, as Dr. Gaudia observes, is the: “assumed values- or assumptions – created by the person using the theorem to support a preconceived viewpoint.”

In many ways it would be analogous to my bright, seven year old niece offering that a “50% probability exists” for a pipe-smoking, invisible (except to her) elf to be sitting in her basement fireplace ten hours out of twenty-four.

In a more sophisticated sense, it would be analogous to Tulane physicist Frank Tipler asserting that a “50-50” chance exists for his “Omega Point” to really be the “First Person of the Trinity”. (E.g. see his balmy tome: ‘The Physics of Christianity’ in which he purports to “prove” the Trinity, the Virgin Birth, the Resurrection of Jesus. and a host of other unadulterated, supernatural codswallop via quantum physics and general relativity)

The trouble is that all of these lack a reasonably sound ab initio epistemology that would warrant a credible statistical assumption. It is precisely the epistemology (if such exists) that would allow us to take the assumption seriously.

For example, I might make the claim that a 50% probability exists for a monster solar flare capable of frying every electric transformer in the U.S. This sounds way out, but we must recall the monster flare of March, 1989 which fairly brought down the Ottawa power grid and indeed, “fried” coils and wires subject to enormous induction currents of 10^ 6 A and more. Hence, given what we know of the magneto-hydrodynamics of flares, and how magnetic substorms are generated on Earth – there is an epistemology for the assignment of probability.

Moreoever, it is one that can be subject to tests. For example at the University of Alaska-Fairbanks now there are ways of generating numerical models of magnetospheric substorms which can duplicate conditions arising from powerful flares.

Can Unwin do the same for his “God”? Can he even differentiate clearly its nature – say from the Hindu Brahmin, or the Islamic Allah, the Jews’ Yahweh or even Tipler’s “Omega Point”? In other words, WHICH God is it that he's “proving” with Bayesian stats, and HOW does he KNOW? What are the discriminating attributes that uniquely distinguish this entity from any other? If this can’t even be specified, it would be like me claiming a “50 % probability” of a monster flare occurrence exists, without even estimating the likely x-ray class, and optical class. I would be hooted out of any solar physics’ seminar!

One would have thought that proving an entity as momentous as “God” would constitute much more than Bayesian smoke and mirrors a posteriori tricks and statistics.

It's all very well to claim an a posteriori "proof" can be rendered for “God” (via Bayesian statistics) but this won't do in serious discussion until epistemological prerequisites are met. It is interesting that whenever assorted God provers like Unwin invoke their proofs, they never provide a discriminating epistemology – which begins with a basic definition of the entity.. They merely use the G-word as if we already know what the entity is they are “proving” (The lone claimed exception to this which I have seen is physicist Frank Tipler in his book, ‘The Physics of Christianity’ – but in order to arrive at it, Tipler must not only resort to a posteriori assumptions and stats, but also mangle all the normative concepts of quantum mechanics!)

So the question persists: WHICH God is it that is being proven?

Is the entity based on the deist (“create it and leave it be”) model? The pantheist Spinozan (“all is God”) model? The personal but "infinite" Christian model? The Christian Trinity? The Hindu impersonal Brahmin model? Until this is resolved in a uniform, operational sense – skeptics and scientists are justified in taking any uses of the G-word with the proverbial grain of salt.

As I have found on numerous forums, and in dozens of venues wherein I've debated these folks (e.g. Salon, AARP forums, Amazon book reviewes etc.) when you press them for an epistemological basis to their deity the best they can do is a generic “designer God”. In other words, as the Discovery Institute and others would warrant, it is the God of “intelligent design” which will answer to the Christian “God” name.

Once you have them offering an “epistemology” – in this case one that posits a “design” presumes a “God” – you have them by the cojones.

They will insist there must be “design” because the odds of overcoming remote probabilities are otherwise too low. But how does one separate naturally low probabilities from outcomes based on the intervention of a special agent? What criteria allow such separation?For example - we know that any one supposed single fusion of (proton) nuclei in the Sun is expected to occur on average once every 14 billion years.However, no designer is required to explain how the Sun shines (from fusion reactions) rather quantum mechanical tunneling. There exists a finite probability that any given proton will surmount the Coulomb repulsive barrier and fuse with another proton. Yes, this is a rare even, but there are trillions of protons in the Sun so it emerges as not so “rare” as one may first believe.

Another question to challenge the epistemology: Why doesn't the "designer" insinuate itself into the domains of other worlds in the solar system to create ("design") life? Why isn't Mercury inhabited, or Venus? Or Jupiter? IF the designer is also omnipotent it ought to be able to design outside of purely natural (or terrestrial) norms and limits. (Thus an organism on Venus, for example, that can live off sulphuric acid, CO2 in the atmosphere and an atmospheric pressure of 90 atm.)If the designer is not omnipotent, and indeed doesn't exist in the first place - it makes more sense that life will only occur on certain planets within habitable temperature zones and containing the elements (oxygen, nitrogen, water etc.) needed for life. In such cases, it isn't "design" at work but a long, gradual process of chemical evolution that eventually leads to life forms. Thus, the only real emergent reason for a designer in the first place would be that it possesses ubiquitous power to design ANYWHERE! If it can't do that, or is limited by conditions already in place - we simply don't need it, it's redundant.

The final challenge to toss at the God-claimant is predicting what the designer or design agent can do for another planet other than Earth. Provide him with ‘x’ attributes of the unknown planet (also give also the spectral class of the star) and let him specify the nature of the “designed” life that would emerge. If he knows his designer deity as well as he claims, he ought to be able to do this and display flexibility irrespective of the parameters thrown at him.

Failing that, force him to offer tests of falsification. Make him tell you what observation – if anything – will falsify his claim. Ideally, force him to relate this to chromosomal observations for various genomes on Earth.(Say using the cytochrome-c gene in one or more species, to determine their interrelatedness, say of the human to the chimp, or the human to a lowly yeast species).

Tipler, at least, has done that – and interestingly, one finds that most of his proffered “tests” have already been falsified. For example, his claim that there can be no “heat death” of the universe. However, the current acceleration of the cosmos via dark energy shows indeed that all the bodies of the cosmos will indeed one day cool to absolute zero temperature. No humanity, no life, no consciousness. It is but a matter of time.

Unwin, Ken Ham and others get away with their nonsense and specious “proofs” largely because skeptics and Atheists fail to call them on their baloney. Demanding that they provide a real epistemological basis for their lunatic claims, as opposed to Bayesian smoke and mirrors. Alas, we often have such a volume of baloney to rebut, that it is often difficult to prioritize!

No comments: