Wednesday, September 29, 2010

Theistic argument run-down

In a recent blog post, Larry Moran raised the question: what are the best arguments for God? The subtext here is that apologists and "faitheists" are always claiming that there are sophisticated arguments for God out there, but rarely seem to be able to describe them.

One commenter, Martin, has risen to the challenge:

A few off the top of my head:

1. Kalam cosmological argument
2. Argument from contingency
3. Plantinga's modal ontological argument
4. Maydole's modal perfection ontological argument
5. Fine-tuning arguments
6. Argument from reason
7. Evolutionary argument against naturalism
8. Moral arguments
9. is loaded with arguments

Not bad, and kudos to Martin for actually coming up with some concrete claims. Ideally he would have said which of these arguments he finds most convincing, but that's a different issue. Let's have a look down the list.

1. The Kalam cosmological argument

This is a variant of the standard fine-tuning argument, employed by apologist William Lane Craig. The standard version takes as its premise that anything that exists has a prior cause. This lays it open to the reductio ad absurdum that, since God allegedly exists, He too must have a prior cause (a meta-God), who must also have had a cause (a meta-meta-God) and so on. Not a terribly elegant explanation.

The Kalam version simply states that everything that has a beginning has a prior cause. Since God is conjectured not to have had a beginning, He therefore doesn't need a cause. Since the universe is conjectured to have had a beginning, it does need a cause, and God fits nicely into the gap.

This argument has attracted a range of criticisms. In particular, it has been argued that things do appear without prior cause all the time in quantum mechanics. Various effects in QM only make sense if "virtual particles" are continually appearing from nothing (in pairs so as not to violate principles of symmetry) and vanishing again.

Craig's defence is that these events don't really count because the particles aren't coming from nothing; they're arising from fluctuations in the background energy field of the universe. This is technically accurate. But it means it's Craig's assumption is impossible to test. If anything within the universe can take the universe as its "prior cause", what exactly would count as a refutation?

There are of course other criticisms, but I only have the one evening. At the very least, though, this argument cannot be considered a "proof" in the same league as mathematical proofs or scientific theories. Like most extant philosophical proofs it is linguistically fuzzy, and it relies heavily on intuitions that are of dubious worth in a non-Euclidean universe.

2. Argument from contingency

A big topic in philosophy has been the study of the distinction between "necessary" truths (1+1=2 can't not be true) and "contingent" truths ("Lifewish is sitting in front of a keyboard" is true but could be false). There is an entire system of logic designed for drawing these distinctions.

The theistic argument from contingency (aka the modal cosmological argument) basically points out that it is really really difficult to reason from the existence of necessary truths to the existence of contingent truths. Basically, how did the first universe-like thing (whether that be this universe, another universe, or something more exotic like an M-brane) come about?

Since the universe is contingent, it is argued, there must have been something to start it off. God, who is assumed to be necessary, is considered a good candidate for this.

I haven't even looked into this one and already I can see a few refutations. For example, under what circumstances can necessary things give rise to contingent things? If it is easy for this to happen, could there not be a necessary thing other than God that's capable of the task? If it is hard, how come God can do it? Intuitively it seems odd that a necessary God could give rise to a contingent universe; how would He know which universe to create out of the various options?

It's also open to question in what sense the universe is contingent. Even if it is contingent, maybe it's part of a multiverse that is in some sense necessary. There are various other options to be considered, any one of which is enough to screw up the chain of logic.

Again, this argument is a nice philosophical tetherball to play with, but it is not terribly convincing to anyone who isn't already convinced. It exhibits basically the same cracks as all the other regress arguments, albeit with a slightly fancier wallpaper covering.

3. Plantinga's modal ontological argument

The classic ontological argument is: imagine the greatest God it's possible to imagine. If that God were real, He would be even greater. But then He wouldn't be the greatest God it's possible to imagine. For consistency with our initial premise, we have to assume that this God is real.

Plantinga's variant attempts to translate the fuzzy concept of "imagine" into more rigorous modal logic (the logic of "possible worlds"). First, assume that, in some possible world, there is a God that is "maximally excellent" (I love that phrase, it's very Bill and Ted).

But a maximally excellent God would presumably have the power to reach into other possible universes than their own (otherwise you can imagine an even more bodacious deity). So they'd be maximally excellent in all possible worlds, including our one.

This argument seems to have been absolutely slated by every philosopher who looks at it, including Plantinga himself. (See here for discussion.) Plantinga argues that, since the premises are "rational" (by which I think he means "not obviously daft"), the conclusion must be rational.

All I can say to that is: rational? You keep using that word. I don't think it means what you think it means. To me, and to most other skeptics, "rational" has a higher meaning than just "too complicated to understand, let alone critique". Security through obscurity is a bad principle to build a worldview on.

4. Maydole's modal perfection ontological argument

This is a relatively recent argument, and I hadn't come across it before. The original paper is behind a paywall, so I'm forced to rely on this forum post.

On investigation, this proof is like every other modal "proof" I've read: it smuggles in its conclusion via a complicated statement that sounds plausible until you think about what it actually means.

The statement: "If it's possible that there exists an x that is an F, then there exists an x so that it's possible that x is an F."

How it's being used: "If supremacy is possible then there must exist something that's potentially supreme."

Using this starting point, Maydole pulls a cunning trick. There is a standard theorem in modal logic that if something is possibly impossible then it's impossible. This makes more sense if you translate it differently: "If, from the viewpoint of some possible world, X is false in every possible world, then X is false".

By defining supremacy in terms of impossibilities, Maydole uses this theorem to create a sort of "potential world contagion" in a similar way to Plantinga's argument. Once the walls between possible universes are broken down, the possibility of God becomes the proof of God.

How to criticise this argument? Difficult without spending a full blog post on it - as with Plantinga's variant, the argument's main defence is its length (and I still only have the one evening). To start with, though, I would note that, in the version I linked to, statement P3 does not actually follow from statement M2.

I suspect that this error is a side-effect of the forum member's attempt to boil the argument down. But it demonstrates how fragile the argument is as a whole... and how little progress I can make without access to the original paper. I may research it in more depth at a later date if anyone is remotely interested.

5. Fine-tuning arguments

These have been refuted to death. The basic form is:

1) The universe (or some part of it) has a certain property with a certain value.
2) Human life wouldn't be able to tolerate a different value of that property.
3) Therefore God fine-tuned that property, because He cares about us so much.

Classic refutations:

1) Human life might not exist if the property were different... but some other wildly different lifeform might. In a universe with four dimensions, I wouldn't be typing this blog post. But a fifty-armed intelligent plasma cloud might be.

(See also Douglas Adams' puddle analogy).

2) Even if the property's value was an amazing slam-dunk, how many shots were there at the hoop? Many modern physical theories require there to be a ridiculously large number of alternate universes, each with slightly different attributes. Maybe we're just in the one of them that happened to be habitable?

(Note that there is no extant physical theory that requires God to exist...)

3) If God cares so much about getting the environment right for us, why is 99.9999etc percent of the universe uninhabitable vacuum? Why is 99.86% of the solar system's mass stuck inside the Sun, rather than being used to make more habitable planets? Why is 71% of the Earth's surface covered by relatively unhelpful water rather than fertile soil? God does not seem to be going out of His way to make things easy for us; why should the "fine-tuned" property be His one exception?

6. Argument from reason

Here I assume that Martin is referring to the transcendental argument. Roughly stated: in a Godless universe there's no obvious reason to think we'd be able to come to accurate conclusions. So, by arguing about anything (including God's existence), we are implicitly conceding that He exists.

I like to think of this as the Argument from It's My Ball And You Can't Play With It.

There are two really insanely obvious counters to this. Firstly, the ability to accurately model the behaviour of things confers a strong survival advantage. So we'd expect it to crop up occasionally in evolved species.

Such a creature might not be capable of discovering Ultimate Truths. But there's a pretty strong argument that we don't do that either.

Secondly, even if there were no Godless cause for rationality, this argument doesn't explain why God would be any more likely to create thinking organisms. Why would God not create mad creatures? Only by positing a very very specific God - a God that's naturally inclined to create sane humans - can we get to the conclusion. But then we might as well just posit a universe that is inclined to give rise to sane humans, and save ourselves the trouble of coming up with a proper causal explanation. This would make biology classes a lot shorter...

(In fact, many prominent believers have claimed that reason is in some way inferior. Why would God sully His hands with such an invention?)

Without dealing with both these counterarguments, the "argument from reason" is dead in the water.

7. Evolutionary argument against naturalism

This is just Plantinga's rehash of argument #6, focusing on the question of whether evolution can give rise to "true" rationality. Given his complete lack of understanding of evolutionary biology, I'm going with "no"...

See here for more discussion, if you're really keen. If you're a veteran of the Darwin Wars, or you're familiar with the classical transcendental argument, there's not a lot of interest here.

8. Moral arguments

In brief: people do good stuff, therefore God.

We're really scraping the bottom of the barrel here, I'm afraid. The evolution of morality has been studied in truly obscene depth (see for example Matt Ridley's book The Origin of Virtue). It turns out that, in a vast range of situations, moral behaviour follows directly from evolutionary premises.

For example, around some coral reefs there are big fish who get food stuck in their teeth. There are smaller fish who clean their teeth for them (thus getting a free meal). But what's to stop the feeder fish from eating the cleaner fish that groom them?

It turns out this does occasionally happen. But the cleaner fish have got very good at distinguishing individual feeder fish, and can recognise a back-stabber they've seen previously. So any feeder fish that goes rogue soon runs out of targets. It also doesn't get its teeth cleaned, which is not good if you've got a hot date in the evening.

It's harder to account for "heroic" (read: suicidal) morality - people throwing themselves on grenades and so on. However, I'd like to draw your attention to the principle of overcommitment. This says: I will do what I've promised to do, and I will do it in a really over-the-top fashion.

This can have useful effects. For example, if you're the kind of person who goes psycho and beats the crap out of people if they spill your beer, no-one is going to come near your glass. The "mutually-assured destruction" of nuclear war is another example. Paradoxically, by gaining a reputation as a nutter, you can

The converse is also true. By becoming the very avatar of all that is nice and friendly, you can get people to give you more slack than you'd get any other way. Think of the respect people have for those who spend their lives and money on charitable causes. It's an excellent way to get girls.

The problem with this is that you have to keep doing it. The moment you throw a comrade on the hand grenade rather than jumping on it yourself, your cover is blown (even if you aren't...).

9. is loaded with arguments

Maybe so. But at this point I'll draw your attention to a famous quote by Einstein. On being informed that the Nazis had published a booklet called "100 Scientists against Relativity", he commented "if they were right, one would have been enough".

Similarly, if God existed, one truly solid demonstration would be worth a hundred of the half-baked "proofs" I've dissected above.


John A. Davison said...

I am pleased to find someone who is impressed with my friend Martin. Martin has always been a vocal supporter of my Prescribed Evolutionary Hypothesis.

Lifewish said...

Hi John,

I think I came across your PEH a while back. That's the idea that new patterns of gene behaviour don't result from random mutation, but rather from the expression of previously-unused genetic material. Right?

It's an interesting idea, but it's not really compatible with biology as I understand it. Of course, it's entirely possible I've misunderstood the PEH (or biology). A few queries in particular:

1) Bacterial genomes are mostly crammed very tightly with genetic material. Where is there room for your hypothesised palæo-genes?

2) Genomes are demonstrably prone to random mutation in varying degrees. If a given chunk of genetic data isn't being used, it will tend to decay into meaningless; the more specific it is, the faster it will lose function. How would the palæo-genes survive this process?

3) The field of computational biology includes some very sophisticated ways of spotting functional genetic material (e.g. by folding pattern, by binding site motif, by comparison to other sequences). Do these methods detect palæo-genes; if not, why not?

4) If genes can remain inactive for long periods before regaining function, why does cladistics work so well when applied to genetics? For example, primates possess the same gene for vitamin C production as other mammals, but it is inactive (hence scurvy). Why would this gene not be able to reactivate itself as needed, e.g. in 18th-century fruit-deprived sailors?

5) What concrete, testable, unique predictions does the PEH make? (This is a general question that I ask about everything.)

John A. Davison said...

Lifewish,whoever that is,

You misinterpret my science. I recommend you visit my website where you will find my papers both published and unpublished. You will also find extensive discussions of my work at "brainstorms" forum.

Then if you have questions ask them at my website where I am always happy to answer questions, preferably from those who are willing to use their real names.

Anonymous said...

'God' in the singular is cause but cause of itself does not have to presuppose a single entity as cause.This carries some post revelatory retrospective cast. However, with the development of mathematics the 'infinity dissonance' of odd and even instinctive aestheticism would be gratified by 'One'; numbers after all are a duplication.
The extrapolation of the cause and effect of the particular to the totality of particulars is so plausible that only the effects post revelatory dogma would find the need to oppose it.
Anyway,repudiation of the purported evidence for a transmaterial cause of this world,if it were to be true, wouldn't of itself dismiss the possibility of such an entity(ies)
with no causal relationship to it.

Lifewish said...

Anonymous, thanks for your interesting comment. I had to read it a couple of times to fully grasp your meaning; apologies in advance if I've misunderstood you.

However, with the development of mathematics the 'infinity dissonance' of odd and even instinctive aestheticism would be gratified by 'One'; numbers after all are a duplication.

Reread your Bertrand Russell. Any sufficiently interesting mathematical system can generate objects equivalent to the integers.

Sorry to be terse, but as a maths grad myself I can't help but get annoyed by this use of mathematical aestheticism to "disprove" polytheism. It is itself mathematically inelegant.

The extrapolation of the cause and effect of the particular to the totality of particulars is so plausible that only the effects post revelatory dogma would find the need to oppose it.

Plausible =/= correct. If by "effects post revelatory dogma" you mean modern skepticism, damn right we find the need to oppose it.

It is possible to construct an entirely plausible-sounding philosophical argument for almost any position; that is the triumph and the price of philosophy. This is why, when confronted with a "plausible" argument, I prefer to look at the real-world consequences of that argument and its converse.

To quote the author Tom Clancy: "The difference between truth and fiction is that fiction has to make sense."

Anyway,repudiation of the purported evidence for a transmaterial cause of this world,if it were to be true, wouldn't of itself dismiss the possibility of such an entity(ies)
with no causal relationship to it.

This is true. However it would:

a) render such entities epistemologically unnecessary (Laplace's "sir, I had no need of that hypothesis");

b) imply that the very large number of people who feel they personally have strong evidence for such an entity are very badly mistaken.

air max 90 homme said...

It's an interesting idea, but it's not really compatible with biology as I understand it. Of course, it's entirely possible I've misunderstood the PEH (or biology).