Alvin Plantinga's review of Dawkins' God Delusion

[Cheerful Charlie]

No, if sound the argument shows that atheists but not theists can't consistently think their position justified (i.e. likely true because of the arguments they've come up with).


There is no sound theist argument. Theism has made many a claim about the natural world over millenia and theism has a perfect record, always wrong. The world is not flat, is not the center of the universe and the sun does not orbit the earth. Animals werenotcreated afterman ala Genesis 2, nor the sun and stars after Earth ala Genesis.

The omni-everything creator god described by theology is impossible due to internal contradictions created by that god's claimed attributes.

That god is claimed to be omnipotentent all without any comptetent explanation as to how.

The Easter Bunny argument: "The Easter Bunny waves her paw and magic happens!".

How does god do all the things his claimed to do? Easter Bunny argument. he just does is all. Not really an argument. Special pleading at best.

Cheerful Charlie

[Von Smith]

No, if sound the argument shows that atheists but not theists can't consistently think their position justified (i.e. likely true because of the arguments they've come up with).


Dawkins isn't saying that only naturlists' brains evolved, or that only atheists' minds are the product of the neurological activity of their brains; he is saying that *all* of our minds are this way. If he is correct, then it doesn't matter if your a theist or an atheist; our minds are what they are, their limitations are what they are, and our ability to justify our beliefs are what they are. Having religious beliefs that deny the human condition does not enable theists to overcome said condition.

Whether Dawkins is correct or incorrect is an issue of fact, not an issue of what would be cooler and assure us greater certainty if only it were true. And that is essentially all Plantinga has said; his objection to Dawkins' naturalism and materialism boils down to: "It would really suck if our brains were just adaptive machines; but wouldn't it be really cool if instead there were ghosts in those machines that could magically discern The Truth(tm)?" Even if you agree with AP's premise, it doesn't lead you to any real argument against materialism.

And at any rate, his premise is probably false. Positing that our minds were created by God doesn't warrant trusting their access to truth any more than positing that they evolved naturally. Plantinga has to appeal to arbitrary accessory assumptions about what God's intentions were. But if he can help himself to the bare-assed assertion that God "just did" make our minds reliable, then so can an atheist, with equal warrant and consistency, claim that our brains "just did" evolve to be reliable.

Actually, I would go further to suggest that the theist actually loses out in this regard. There is, contrary to what Plantinga suggests, at least some evidential basis for supposing that adaptive belief-forming systems will tend to form accurate beliefs about the world. I doubt Plantinga would suggest with a straight face that military operations conducted with poor intelligence are as likely to succeed as those conducted with the benefit of ample and accurate intelligence. There is no comparable *evidential* basis for expecting reliability of a mind created by a God.

[Antiplastic]

I doubt Plantinga would suggest with a straight face that military operations conducted with poor intelligence are as likely to succeed as those conducted with the benefit of ample and accurate intelligence.

Your doubts are misplaced. Apologetics are un-underestimable.

[Tears In The Rain]

I'd like to see Plantinga's strategies for scoring well on the SAT's without reading the questions. There must be some strategies that do just as well as knowing how to answer the questions, right?

[breezanne]

I'm thinkin' that last question would be more for the anti-theist than the theist.

[Zebulon]

And at any rate, his premise is probably false. Positing that our minds were created by God doesn't warrant trusting their access to truth any more than positing that they evolved naturally. Plantinga has to appeal to arbitrary accessory assumptions about what God's intentions were.

Exactly. There is no way to demonstrate that the likelihood of a benign god is greater than the likelihood of a malicious or trickster god. Fervently desiring it to be true doesn't make it so.

[Thomas Ash]

Dawkins isn't saying that only naturlists' brains evolved, or that only atheists' minds are the product of the neurological activity of their brains; he is saying that *all* of our minds are this way. If he is correct, then it doesn't matter if your a theist or an atheist; our minds are what they are, their limitations are what they are, and our ability to justify our beliefs are what they are.

No, on the theistic hypothesis there's a benevolent God who guides evolution or just creates us so that we're truth-sensitive. Yes, God's being benevolent & caring about our truth-sensitivity is a "bare-assed assertion", but as was discussed earlier in this thread, that's what hypotheses are.

[Zebulon]

Yes, God's being benevolent & caring about our truth-sensitivity is a "bare-assed assertion", but as was discussed earlier in this thread, that's what hypotheses are.

The difference being that science tests its hypotheses. There is no way to test the assertion that a god is benevolent and caring.

[Von Smith]

No, on the theistic hypothesis there's a benevolent God who guides evolution or just creates us so that we're truth-sensitive. Yes, God's being benevolent & caring about our truth-sensitivity is a "bare-assed assertion", but as was discussed earlier in this thread, that's what hypotheses are.

An hypothesis is proposed to explain some fact in evidence, to be tested against further facts in evidence. The claim that God cares about our access to the truth does not function that way in Plantinga's objection to naturalism. Plantinga is not faulting naturalism for not being able to explain an alleged *fact* about our access to the truth, but for undermining the claim that we have any to begin with. IOOW, his objection boils down to the complaint that it would really suck if Dawkins were correct.

It wouldn't help his position much to say that the former *was* Plantinga's point. Saying "because God wanted it that way" is a vacuous sort of explanation, a virtus dormitiva explanation. Why is the sky blue? Because God made it blue. Why should we think our minds are reliable? Because God made them reliable.

I notice you haven't addressed my other major points. To summarize:

1) If one can arbitrarily claim that God "simply does" care about our access to the truth, then one can with equal justification claim that adaptive brains "simply do" have access to said truth.

2) One can adduce actual evidence to support the idea that some degree of accuracy in cognition is necessary for it to be adaptive in the natural world. No such evidence can be adduced to support the idea that a Supreme Being would care about our access to the truth.

[Thomas Ash]

The difference being that science tests its hypotheses. There is no way to test the assertion that a god is benevolent and caring.

Sure there is - you look at the world He's allegedly created (and perhaps see a problem of evil).

[Thomas Ash]

An hypothesis is proposed to explain some fact in evidence, to be tested against further facts in evidence. The claim that God cares about our access to the truth does not function that way in Plantinga's objection to naturalism. Plantinga is not faulting naturalism for not being able to explain an alleged *fact* about our access to the truth, but for undermining the claim that we have any to begin with. IOOW, his objection boils down to the complaint that it would really suck if Dawkins were correct.

True, Plantinga's not giving evidence for the theistic hypothesis, but he's doing more than fallaciously saying "naturalism sucks, so it must be false": I suppose he's saying something like "naturalism, unlike theism, does not allow our beliefs to be justified (because it would give us no reason to think them reliable), so we can't have a justified true belief in naturalism". This wouldn't mean naturalism must be false.

It wouldn't help his position much to say that the former *was* Plantinga's point. Saying "because God wanted it that way" is a vacuous sort of explanation, a virtus dormitiva explanation. Why is the sky blue? Because God made it blue. Why should we think our minds are reliable? Because God made them reliable.

I disagree that its a virtus dormitiva explanation so long as there is some systematic reason for God to want the things he does (e.g. because they're good) - virtus dormitiva explanations would not offer such systematic, simplifing explanations, and AFAIK that's what's wrong with them. What's wrong with the God explanation?

I notice you haven't addressed my other major points. To summarize:

1) If one can arbitrarily claim that God "simply does" care about our access to the truth, then one can with equal justification claim that adaptive brains "simply do" have access to said truth.

This'd be a legitimate hypothesis, yes. I guess Plantinga would say that adding it to naturalism would make that theory less plausible by decreasing its simplicity and general theoretical virtue. Of course, one could say the same of theism, unless it paid off elsewhere, perhaps by simplifying our theory in other areas.

2) One can adduce actual evidence to support the idea that some degree of accuracy in cognition is necessary for it to be adaptive in the natural world. No such evidence can be adduced to support the idea that a Supreme Being would care about our access to the truth.

I've no argument on the first point; but if the hypothesis is theism, the being would be good and so plausibly care about our access to the truth (so I guess for 'naturalism' we should read 'atheism').

[untermensche]

Sure there is - you look at the world He's allegedly created (and perhaps see a problem of evil).
It cannot be used any further than to say, a benevolent and caring god can possibly use evil to it's purposes.

You take one step towards a serious examination of this thing they call god and you step into quicksand.

It is best to simply dismiss these creatures with a wave of the hand and move on.

[Thomas Ash]

It is best to simply dismiss these creatures with a wave of the hand and move on.

Well, I prefer not to wave my hands :)

[SophistiCat]

I disagree that its a virtus dormitiva explanation so long as there is some systematic reason for God to want the things he does (e.g. because they're good) - virtus dormitiva explanations would not offer such systematic, simplifing explanations, and AFAIK that's what's wrong with them. What's wrong with the God explanation?

Well, 'good' is one word, but that doesn't mean that claiming that something is 'good' is a simplifying explanation, just as 'goddidit', though one word (;)), doesn't make for a non-tautological explanation. You need some ethical theory that amounts to more than a post-hoc rationalization in order to make a meaningful explanation out of this.

[Thomas Ash]

Well, 'good' is one word, but that doesn't mean that claiming that something is 'good' is a simplifying explanation, just as 'goddidit', though one word (;)), doesn't make for a non-tautological explanation. You need some ethical theory that amounts to more than a post-hoc rationalization in order to make a meaningful explanation out of this.

True. I think I already granted to you that God's a poor explanation unless we already have some substantive moral truths independent of Him which can guide His action (so give us some idea of what He'd be more likely to do - otherwise one could just say 'goddidit' of anything, including pointless evil).

[Zebulon]

Sure there is - you look at the world He's allegedly created (and perhaps see a problem of evil).

Setting aside for the moment the question of this god's existence, and its alleged creation of the world, is there any evidence to be seen that such a being may be benevolent? Theodicy seems to suggest either divine impotence, ignorance, disinterest or malevolence. If there is a creator god, he seems more likely to be a bumbling or evil demiurge than the omnipotent, omniscient, omnibenevolent god of the theologians.

[Thomas Ash]

Setting aside for the moment the question of this god's existence, and its alleged creation of the world, is there any evidence to be seen that such a being may be benevolent? Theodicy seems to suggest either divine impotence, ignorance, disinterest or malevolence. If there is a creator god, he seems more likely to be a bumbling or evil demiurge than the omnipotent, omniscient, omnibenevolent god of the theologians.

I agree!

[Thomas Ash]

BTW, Sophisticat, I have another question about our having to calculate P(E) and P(G) in ignorance of E (i.e. this universe supporting life): does this mean that we can't take into account the majority of arguments for and against the existence of God, because they presuppose (or even make it more or less likely than it otherwise would be) that our universe supports life?

[SophistiCat]

BTW, Sophisticat, I have another question about our having to calculate P(E) and P(G) in ignorance of E (i.e. this universe supporting life): does this mean that we can't take into account the majority of arguments for and against the existence of God, because they presuppose (or even make it more or less likely than it otherwise would be) that our universe supports life?

Yes, I think so. This is known as the "problem of old evidence," and formal solution to this problem looks something like this. If P-(G) is our current probability of God (understood as degree of belief), then a successful argument for God should increase that probability, so that it becomes

(1) P+(G) > P-(G).

Let's say that K is all relevant background information (we can just include everything that we know - what isn't relevant will have no effect on the argument). If E is, specifically, evidence of fine-tuning for life, then we'll denote the subset of K that does not include E as K*, so that

K = (K* & E).

Then in order to construct a probabilistic argument for God, we should first determine the prior probability (called ur-probability) that does not depend on E, i.e.

P*(G) = P(G|K*)

Then we update this probability with evidence E, and hopefully find that

(2) P(G) = P(G|K* & E) = P(G|K) > P*(G)

This posterior probability is what our P+(G) is going to be:

P+(G) = P(G)

But now the million dollar question is, will this amount to a successful argument, i.e. will proposition (1) be true? This depends on the relationship between the ur-probability P*(G) and our prior probability P-(G).

As is easy to see, if our prior belief in God in no way depended on knowing E, then the two probabilities will be the same. In that case, as long as proposition (2) holds, i.e. as long as the evidence of fine-tuning raises the ur-probability of God, the argument is successful.

On the other hand, if our belief in God prior to the application of the argument fully depended on the evidence of fine-tuning (or its consequences), then the ur-probability P*(G) will be lower than the prior probability P-(G). Moreover, the argument in (2) will only raise P*(G) up to the level of P-(G), so that

P+(G) = P-(G)

and the argument fails.

The mistake that many proponents of the argument from fine-tuning make is failing to excise the evidence E from the "old" evidence before making the argument, in effect using the same evidence twice: once when forming the prior probability and then again when obtaining the posterior probability.

[Von Smith]

True, Plantinga's not giving evidence for the theistic hypothesis, but he's doing more than fallaciously saying "naturalism sucks, so it must be false": I suppose he's saying something like "naturalism, unlike theism, does not allow our beliefs to be justified (because it would give us no reason to think them reliable), so we can't have a justified true belief in naturalism". This wouldn't mean naturalism must be false.

OK, I see your point, and as I read through "Naturalism Defeated", I can see that Plantinga's argument is more involved that I had thought, so I'll retract my earlier claim.

That being said, I still don't think Plantinga is left with a particularly good argument. Just because one does not assert an a priori justification for believing something does not mean that one doesn't have any justification at all; likewise, simply asserting an a priori justification does not mean that one actually has one (e.g., Calvin's and Aquinas' theological arguments are not really evidence that God has, in fact, made us rational, since they might just have easily made those arguments if He hadn't).

I disagree that its a virtus dormitiva explanation so long as there is some systematic reason for God to want the things he does (e.g. because they're good) - virtus dormitiva explanations would not offer such systematic, simplifing explanations, and AFAIK that's what's wrong with them. What's wrong with the God explanation?

I doubt that there is a non-circular or non-ad-hoc way of establishing such "systematic reasons". For example, how is it objectively "good" for humans to have a particular set of cognitive abilities? Why would those reasons apply to humans in particular? Why would it not be equally good for slime molds to have such abilities, or asteroids?


This'd be a legitimate hypothesis, yes. I guess Plantinga would say that adding it to naturalism would make that theory less plausible by decreasing its simplicity and general theoretical virtue. Of course, one could say the same of theism, unless it paid off elsewhere, perhaps by simplifying our theory in other areas.

How does positing an unevidenced being, and making additional unevidenced and undecidable claims about that being make an explanation simpler?


I've no argument on the first point; but if the hypothesis is theism, the being would be good and so plausibly care about our access to the truth (so I guess for 'naturalism' we should read 'atheism').

Well, again, theism (what Plantinga refers to as "austere theism") does not, in and of itself, entail the notion of a deity that would care about our access to the truth. That is an additional assumption. Not every theist, nor indeed every Christian, agrees with Aquinas or Calvin.

[Thomas Ash]

If E is, specifically, evidence of fine-tuning for life, then we'll denote the subset of K that does not include E as K*, so that

K = (K* & E).

Presumably, you mean 'include E' broadly, so that K* involves only evidence that does not have any bearing on the probability of E? e.g. the universe's obeying regular laws doesn't entail its permitting life, but it is arguably a prerequisite of it.

I suppose an additional requirement will be that when, after factoring in E's effect, we update G on new evidence (E2) which was excluded by K*, we include E in the background evidence? This will let E2 play its proper role in discriminating among the theories which made E relatively probable.

[Cheerful Charlie]

Setting aside for the moment the question of this god's existence, and its alleged creation of the world, is there any evidence to be seen that such a being may be benevolent? Theodicy seems to suggest either divine impotence, ignorance, disinterest or malevolence. If there is a creator god, he seems more likely to be a bumbling or evil demiurge than the omnipotent, omniscient, omnibenevolent god of the theologians.


Is there any evidence of a god at all? We get the notion God is good from revelation that tells us that. But Epicurus's problem of evil was used to show the gods, if they existed were indifferent to us. God cannot be totally evil. Imagine such a world. Mothers would have large litters of children so most could die young of horrible and painful diseases. I will leave it up to you to create, as a thought expeirment what a truely evil, malicious and sadistic god would create as his world.

Now imagine a world created by a truely, conscientious and totally good, merciful and just god as we would imagine such a good god would create. Hurricanes and earthquakes would never kill infants, only the worst sinners in town. Again, I leave this up to you as an exercise.

A tsunami kills 250,000 people because of a large earthquake. A good god would have had 4 or 5 smaller earthquakes so no tsunami would indiscriminately kill innocents including children.

Why is god so very much like nothing at all?

Cheerful Charlie

[SophistiCat]

Presumably, you mean 'include E' broadly, so that K* involves only evidence that does not have any bearing on the probability of E? e.g. the universe's obeying regular laws doesn't entail its permitting life, but it is arguably a prerequisite of it.

I think E would be whatever evidence we specifically want to use as an argument for God. Such as for example the fact that certain constants are within a certain narrow life-permitting range. The requirement for K* is simply that it doesn't include or imply that evidence. But the assumption that the universe is obeying regular laws would be necessary to make a priori for any argument of this sort, either pro or contra. Without this assumption we simply can't calculate any likelihoods. Even God has to be assumed to behave in predictable ways, which I would agree is a problematic notion.

I suppose an additional requirement will be that when, after factoring in E's effect, we update G on new evidence (E2) which was excluded by K*, we include E in the background evidence? This will let E2 play its proper role in discriminating among the theories which made E relatively probable.

Right, it's an iterative process. It's what enables us to hone our theories.

[Thomas Ash]

But the assumption that the universe is obeying regular laws would be necessary to make a priori for any argument of this sort, either pro or contra. Without this assumption we simply can't calculate any likelihoods. Even God has to be assumed to behave in predictable ways, which I would agree is a problematic notion.

Right, I was after some evidence E2 which would have no bearing on E and so could be included in K* - clearly this is hard to find!

[Zebulon]

Is there any evidence of a god at all? . . . Why is god so very much like nothing at all?

Cheerful Charlie

I agree. My point was that even in the hypothetical, the theist cannot simply conjure up the god he wants without making even more assumptions about the nature of a being whose existence itself is completely unsupported.

[Thomas Ash]

I suppose an additional requirement will be that when, after factoring in E's effect, we update G on new evidence (E2) which was excluded by K*, we include E in the background evidence? This will let E2 play its proper role in discriminating among the theories which made E relatively probable.

Hmm, actually since E2 is the sort of evidence we excluded when considering E (because it had some bearing on E's truth, perhaps even entailing E) it seems that by the same rule we can't include E in the background evidence when considering E2? This implies that we can't make a step-by-step case (based on E and then E2) for G or whatever other hypothesis they support: we have to lump E and E2 into our evidence at once.

But even if theism has additional evidence E2 and multiverse theories don't, this E2 still doesn't give theism any additional advantage in fine-tuning arguments (as it at first seems it might do, because theism would then be less 'speculative' than multiverse theories). This is because P(E2|No God & multiverse) is presumably no lower than P(E2|No God). Therefore the impact of E2 will only count as much as it does against atheism anyway.

[SophistiCat]

Hmm, actually since E2 is the sort of evidence we excluded when considering E (because it had some bearing on E's truth, perhaps even entailing E) it seems that by the same rule we can't include E in the background evidence when considering E2?

If they are not statistically independent, yes. But do they have to be?

[Thomas Ash]

If they are not statistically independent, yes. But do they have to be?

Well, most examples of purported evidence for theism presuppose the existence of live in the universe, or are presupposed by it - I can't think of any other examples.