Glenn Peoples on the Evil God Challenge

I have been having an exchange with Glenn Peoples on his blog about the Evil God Challenge. Glenn thinks the problem of evil (and reverse problem of good) is neutralized by the theodicies (and reverse theodicies). Hence there's no reason provided by the vast quantities of evil/good we observe to conclude that belief in a good or evil god is unreasonable. So all Glenn has to do to show that belief in a good god is quite reasonable is, he thinks, to come up with e.g. a fairly good moral argument for the existence of God. So here's my latest comment...

Let me explain how things look from my end.

I give you what appears to be overwhelming empirical evidence against the existence of your particular God - the evidential problem of evil (e.g. hundreds of millions of years of horror before humans show up, a million plus generations of children around half of which are killed through disease and/or starvation before they reach the age of 5 before Jesus shows up, etc. etc.)

You appear to respond, in effect, by saying: (i) but we theists have all sorts of explanations for all this evil (theodicies), which I think are quite good explanations (ii) even if they are not that good, they can be supplemented by sceptical theism which I don’t rule out, so (iii) the onus is on you to show all these theodicies collectively fail and that sceptical theism is untenable, before you can say that you have provided good evidence against the existence of my God.

But the thing about the theodicies, Glenn, is that they are what Popper calls ad hoc. They lead to now new tests. Or, if they do, but the further test fails, there’s always another gerrymandered explanation for the failure that can be cooked up. Similarly, appeals to God’s mysterious ways and facts-beyond-our-ken are ad hoc. There’s no way empirically to test the claim that such facts-beyond-our-ken is indeed the correct explanation for why there’s so much evil.

Much the same intellectual strategy that you are employing to defend theism is also employed by Young Earth Creationists (YEC), conspiracy theorists, Erich von Daniken style alternative historians (aliens built the pyramids) and countless other wackos to convince themselves and their followers that what they believe cannot be so silly after all.

For of course, if I present a series of evidence-based arguments against YEC, its proponents can say, “Ah, but we have some, we think, quite good explanations of the order of the fossil record, for light from distant stars, etc. - hundreds of such explanations in fact” (explanations cooked up at the Institute for Creation Research and other multi-million dollar funded “research” institutions), and (ii) in any case, God might have his mysterious reasons for arranging the fossils, etc. like that, so (iii) the onus is on you to show all these YEC-type explanations collectively fail and that such appeals to God’s mysterious reasons is untenable, before you can say that the facts to which you point provide good evidence against YEC.

Of course, when we then try to show the failings of the YEC explanations offered, the proponent of YEC can always gerrymander up yet more explanations, and then even more, thereby continuing to make their theory “fit” the evidence. They thus render their theory empirically unfalsifiable (this is the strategy I call “But it Fits!” in my book Believing Bullshit).

But that is, indeed, all bullshit, isn’t it? The fact is, YEC IS pretty straightforwardly falsified by the available empirical evidence, notwithstanding the possibility of endlessly explaining that evidence away by ad hoc means and/or appeals to mystery. Most of us can see that straightaway (those of us whose minds have not been captured by YEC, that is). The endless ad hoc-ery and mystery-mongering is just a smokescreen.

The onus is clearly not on us to refute all the explanations on offer by the YECs. In fact that’s an impossible task given the ad hoc character of their explanations and the fact they're prepared to keep constructing them ad nauseum. It’s entirely reasonable for us to insist that the available empirical evidence DOES indeed very effectively undermine YEC, and that it does so precisely because the YECs’ method of explaining it away is so hopelessly ad hoc.

This is why, before we are presented with any argument FOR classical theism or YEC that might be furnished to save or support the theory, it is indeed entirely reasonable to conclude, on the basis of the kind of observational evidence outlined, that classical theism/YEC is false.

POSTSCRIPT. Glenn has responded with three points, to which I've responded. Here's the points with my responses...
PPS 18th Dec. I have now expanded the explanations below because they were too sketchy.

Hi Glenn

my quick response to your three comments.

First, here's what an ad hoc hypothesis actually is (as Popper and I use the term). It's a hypothesis introduced to save a theory from refutation, a hypothesis that is not independently testable.

Illustration. The Aristotelean cosmology said the heavenly bodies are perfectly spherical. Galileo observed mountains on the moon through his telescope. One Aristotelean attempted to save his theory by insisting there was an invisible substance on the moon that covered the mountains, making it perfectly spherical. This theory-saving hypothesis was ad hoc because (at the time) it was untestable.

Not all theory-saving hypotheses are ad hoc. Newton's theory of universal gravitation predicted a smooth orbit for Uranus. Uranus was observed to have a wobbly orbit. To save Newton's theory, scientists introduced the hypothesis that there was a further planet tugging Uranus out of orbit. This new hypothesis was not ad hoc as it led to new tests - astronomers looked at where the mystery planet would have to be, and found it - that's how Neptune was discovered.

Even when individual theory saving hypotheses are not individually ad hoc, they can be collectively rendered ad hoc if the defender of the theory is prepared endlessly to cook up new hypotheses to save the theory. Or appeals to mystery, of course, which are also, in effect, ad hoc. This is the strategy I call "But it Fits!" in the book Believing Bullshit.

Now to Glenn's response. He says...

GLENN: 1) I’m not even close to being persuaded that the plausibility of theodicies is anything like the plausibility of explanations for why we should believe in a young universe.

ME: What you’re persuaded of is irrelevant. I have pointed out why your method of dealing with the problem of evil is essentially similar to that employed by Young Earth Creationists to deal with counter-evidence.

GLENN: 2) Theodicies don’t strike me as ad hoc. Things like the free will defence or the soul building defence (etc) are generalisable. E.g. the might be stated something like “For any perfectly good and all powerful being, it would still be conceivable that they allow X provided it has some outcome that is compatible with their good character, such as Y.” Ad hoc explanations are really one-off explanations of a sort that are just made up to explain one very specific situation by appealing to principles that are of no use otherwise. So it’s not ad hoc at all.

ME: That’s not what ad hoc means, Glenn. Ad hoc explanations lead to no new tests. The theodicies are ad hoc, by Popper’s definition (he coined the phrase). Look it up. Or, when the theodicies are not ad hoc, and the further test is failed, they are salvaged by yet another defensive manouevre, just as in the case of YEC, thereby rendering the theory unfalsifiable (or an appeal to mystery, of course). Nutters who believe dogs are spies from the planet Venus, etc. employ the exact same strategy.

Ad hoc hoc defences CAN be generalizable. For example, to defend my theory that the Earth is ruled by alien lizards, I can deal with an apparent counter-evidence by saying: "Ah, but that evidence was of course planted there by the alien lizards to fool us." That's a great general, blanket immunizing strategy. it's not one off.

GLENN: 3) Even if things were different and theodicies were ad hoc, they are intended as explanations for why a person might do or allow something that you didn’t expect them to. If anything is allowed to be ad hoc, surely it’s something about why so-and-so might do something. If you rejected the explanation because it was ad hoc, you’d be effectively stacking the deck against any explanation in terms of a person’s intentions, which would be unfair in this case, to put it mildly. But this is moot, since theodicies aren’t ad hoc in any important sense anyway.

ME: The theodicies are indeed ad hoc in Popper's sense. They lead to no new tests (either that, or further explaining away is done ad nauseum to deal with further explanatory failures, or they're supplemented by appeals to mystery). This is NOT like when someone does something out of character and we say, ah, but they probably had this reason for doing it. Often, we can test our hypothesis. So the suggestion is not ad hoc at all. And the occasional ad hoc explanation for anomolies is in any case acceptable (even Popper thought so). However, when there’s considerable evidence against a theory and it’s all dealt with by ad hoc means (and/or appeals to mystery), then that counter-evidence is NOT neutralized.

You’re strategy is, in short, very much like a wife who, when presented with a husband who very often acts in seemingly cruel and vicious way, beating her and her children, maintains he is nevertheless entirely noble and virtuous. She simply explains all the bad stuff away in a manner that is entirely ad hoc (or, when her excuses and explanations for his behaviour clearly fail, just constructs yet more explanations ad nauseum, and/or appeals to his having mysterious unknown reasons).

You, Glenn, say: "If you rejected the explanation because it was ad hoc, you’d be effectively stacking the deck against any explanation in terms of a person’s intentions". This is just false. You have misunderstood what "ad hoc" actually means, as I and Popper use the term. Explanations in terms of people's intentions usually aren't ad hoc, as it's usually possible to test the explanation. E.g. We believe Tom is kind and non-violent. We discover he has killed someone with a knife. We postulate that he killed in self-defence. That it was a case of self-defence is something that can be investigated and indeed potentially shown to be false. It's not ad hoc. But even if it were, it would acceptable if it's a one off example. What's not acceptable is to rely almost entirely on ad hoc means to save your theory from refutation. That's what you are doing, Glenn.

To return to the beaten wife - the wife is being irrational if she insists there’s no prima facie good evidence that her husband is NOT entirely noble and good. She’s deluded. You seem, to me, are a similar case.

Now of course, the wife might insist she has these other very good reasons for thinking her husband really is noble after all. Perhaps she has. But, as things stand, her husband’s horrific behaviour really is excellent evidence that he’s not entirely noble and good, notwithstanding the wife’s endless supply of untestable excuses and explanations.

That’s right, I am suggesting you’re deluded, Glenn. Not very gracious of me, but it’s what I think. Clearly, when we are both so very confident of the reasonableness of our respective, but mutually exclusive, positions, one of us very probably is pretty deluded. The above considerations suggest it’s you.

PPPS. That this is the fundamental problem with the theodicies (and skeptical theism, actually), a problem that the EGC brings out at an intuitive level, is something I'm writing up as an academic paper.

Postscript 20 Dec. Glenn has responded again. Here's my (slightly edited) reply (quoting him):

Glenn you say: "you've got to insist that even explanations that are compatible with all the facts an are true will be discarded by your method of labelling explanations as ad hoc, basically ensuring that no explanation at all will get through your filter,"

Of course true explanations can quite rightly and justifiably be rejected. Happens all the time. But in any case you haven't shown your explanations are true, at this point (considering just the evidential problem of evil prior to considerations favouring theism). You are just assuming they are, at this point!

"No explanations at all will get through your filter." Not sure what this means. Non ad hoc explanations of counter-evidence are fine. Even the occasional ad hoc explanation is acceptable. The only thing I am ruling out is a theory defended against seemingly very powerful counter-evidence more or less entirely by ad-hoc means (plus mystery-mongering). I'm saying, very sensibly, that that does NOT neutralize the counter-evidence! This must, by now, be blindingly obvious to you.

But the key point, Glenn, is, once explaining away all counter-evidence by more or less entirely ad hoc means (plus mystery mongering) is allowed - and that IS your strategy, as you seem finally to have realized - EVERY NUTTY THEORY BECOMES ENTIRELY IMMUNE TO COUNTER EVIDENCE. Indeed, this is the preferred method of dealing with counter-evidence by nutcases the world over.

I can now quite reasonably believe the world is ruled by evil, shape-shifting alien lizards. A wife can quite reasonably believe the husband who beats her and her children is wholly noble and good. Any counter-evidence can quite reasonably be endlessly explained away by ad hoc means (supplemented, if required, by mystery-mongering). Our absurd beliefs will be just as reasonable as yours. And yours as reasonable as ours.

Fumbling Feser

Edward Feser, Catholic philosopher and big fan of Aquinas, wrote a post a year ago explaining why he thinks the evil God challenge doesn't apply to his sort of non-personal, classical God-of-the-philosophers. This is because Aquinas et al demonstrated that anything that's God must be good, given the medieval background metaphysics. Hence an evil god is impossible.

I pointed out that showing an evil God is impossible is irrelevant (I also pointed this out in the paper "The Evil God Challenge" which Feser has read). Even if there were a conceptual problem with the idea of an evil God (and there may also be similar problems with the notion of a good God, actually, but let's set that worry to one side), that does not prevent the evil God challenge from being run. Feser still can't understand why, but here's the reason.

Assume an evil God is conceptually impossible. Nevertheless, there might also be powerful empirical evidence against an evil God. In fact there is - far too much good in the world. And if that empirical evidence is sufficient to rule an evil God out beyond reasonable doubt (at least until some very good counter-argument etc. is forthcoming), why then isn't the evil we see sufficient to rule a good god out beyond reasonable doubt(at least until some very good counter-argument etc. is forthcoming)?

Some have objected that if something is ruled out conceptually, then it makes no sense to suppose there could also be empirical evidence against it. But it seems there can.

For example, William Lane Craig's cosmological argument relies on the thought that an infinitely old universe is a conceptual impossibility. Yet Craig also thinks there's also good empirical evidence that the universe is not infinitely old (i.e. evidence for a Big Bang). So he, for one, accepts that something that's ruled out conceptually might also be reasonably ruled out inductively, on the basis of empirical observation.

Anyhoo, Feser has responded again, in a post called "Broken Law", repeating the same old points. So I thought I'd call this post "Fumbling Feser" in reply. As Feser continues to fumble and drop the ball on the evil God challenge. He continues to maintain it just "doesn't apply" to his classical God. In fact, it does apply. Which is not to say it cannot be met (perhaps even by the construction of a cogent demonstrative proof of the existence of a good God, whether personal or non-personal, perhaps even in the style of Aquinas).

So here's my latest comment on his "Broken Law" post...

Hi Edward

You say:

"Suppose, finally, that you also think there are demonstrative (as opposed to merely inductive or evidential) arguments for the existence of the God of classical theism -- that you endorse an Aristotelian argument from motion to a purely actual Unmoved Mover, say, or Aquinas’s “existence argument” in On Being and Essence for something that is subsistent being itself, or a Neo-Platonic argument for a source of the world that is an absolute unity. If such arguments work at all, then given the background metaphysics, they prove conclusively (and not merely with some degree of probability) that there is a God who cannot in principle be anything less than perfectly good."

Perhaps you have such an argument. In setting out the challenge, I don't claim you don't. This is the bit you still don't understand. The evil God challenge is a challenge. Perhaps the above meets it. Perhaps not. But the above does not show that that the challenge does not apply, i.e. because if correct it shows an evil god is impossible.

Showing an evil God is impossible is irrelevant, for the reasons I explained and which you still don't get.

Demonstrating there's a good god IS relevant, but it is simply a way of meeting the challenge, rather than showing it "does not apply".

You may think this is a fine distinction that doesn't matter much, but it does matter. Because it leaves the very powerfully formulated version of the problem of evil set up via the evil god challenge still on the table, rather than just swept aside on the grounds it "doesn't apply".

Do your medieval "demonstrations" that few philosophers find persuasive really carry much weight against my otherwise overwhelming empirical evidence that your God does not exist?

We'd need to examine them and find out. Though, as I say, the verdict of the philosophical community is already in.

I know you think that's because most philosophers don't really understand them. Actually, I am pretty familiar with arguments of your sort. I've even read your book on Aquinas.

In effect, your response to the evil god challenge is just to say *the evidential problem of evil* "doesn't apply" to your classical God. Because you can demonstrate your God exists.

As a response to the problem of evil - and my evil god challenge - that's obviously hopelessly question-begging.

POSTSCRIPT: The moral is, if you think you can ignore the evil god challenge because you think you can show a priori that an evil god is impossible, think again. That does NOT deal with the challenge. Feser's objection is just an illustration of this more general error.

The "It's hopelessly impressionistic!" response to the evidential problem of evil

The suggestion that the evidential problem of evil is pretty useless as an argument against the existence of God because it is based on claims about suffering, etc. that are hopelessly "impressionistic" has come up several times recently.

David Hart made the move in response to my contribution to 50 Voices of Disbelief (see this post). And now "Fun with Formal Ideas" runs the same move in a comment on the preceding post.

So it's worth dealing with. Here's the comment on the preceding post (nb by "Eth" commentator means the God of Eth):

"Eth is impressionistic, Stephen, it is not founded in any facts, data or evidence where these may be considered synonymous or statistically significant and relies finally upon a appeal to common sense."

David Hart said:

"Nicholas Everitt and Stephen Law recycle the old (and incorrigibly impressionistic) argument that claims of God’s omnipotence seem incompatible with claims of his goodness."

{POST SCRIPT: as TAM's transcript of the radio prog reveals, Denis Alexander also hints at the "It's hopelessly impressionistic" move. Re. huge amounts of suffering, Alexander says: "we simply are not in a position to measure those kind of things, we can measure certain things in science and so forth but..."}

The objection seems to be that assessments of how much good or evil exists are so subjective and unreliable as to be pretty worthless as evidence. Pain and suffering, for example, cannot be given numerical values, or reliably measured using a calibrated instrument in the way, say, mass or velocity can. Ditto happiness, etc.

Does this really deal the evidential problem of evil a fatal blow?

I don't see why.

The evidential problem of evil (and mirror problem in the God of Eth) is based on the empirical observation that the universe is filled with immense amounts of seemingly pointless suffering. This provides us with excellent evidence there is no all-powerful all-good God, in the same way immense amounts of (from an evil God's point of view) seemingly pointless good is excellent evidence there's no evil God either.

Consider a different scenario - a school run by an all-powerful headmaster whom we never see. But we can see how his school is run. Many pupils are beaten senseless, forced to eat shit, and left physically and psychologically crippled by their experiences. Others have wonderful gifts bestowed on them - great food, education, etc. The distribution of these goods and evils appears to be pretty random.

Now consider two hypotheses - that the school is run by a supremely wicked headmaster, and that the school is run by a supremely good headmaster. Both are pretty decisively ruled out by what we observe of the way the school is run.

Notice that the fact that we cannot give numerical values to the pleasures and pains we observe being dealt out, or place them on some sort of objective, calibrated weighing scale, is largely irrelevant. It does not follow that the argument against each hypothesis is based on evidence that is "hopelessly impressionistic".

If that *did* follow, no conclusions about the moral properties of any individual could ever be drawn on the basis of the pleasures or pains they knowingly inflicted. For the reasoning we used would similarly be based on assessments that were "hopelessly impressionistic"!

The "hopelessly impressionistic" response to the evidential problem of evil is, in short, mere smokescreen.

For the evidential problem of evil and my version of it go here. http://journals.cambridge.org/repo_A72V8TEm

What complicates things is that the "hopelessly impressionistic" move is often mixed together with a "no-see-um" move (I think Hart is running a combination of the two - though it's hard to be sure).

The Evil God Challenge

My Paper "The Evil God Challenge" is now available online at the CUP journals page http://journals.cambridge.org/repo_A72V8TEm

This is the final, published version, appearing in Religious Studies shortly.

On Faith - Washington Post

Great writing on Haiti and the problem of evil at the Washington Post. Including a very sharp Dawkins. Go here.

My thanks to Erroll Treslan.

Thought for the Day, 15 January 2010

Some more on Haiti and the problem of evil, this time from The Rev. Dr Giles Fraser:

The word "theodicy" describes the intellectual attempt to justify the existence of God in the face of human suffering. Coined by Leibniz at the beginning of the eighteenth century, he argued that out of the various possible worlds that God could have created, he might have created the best of these, a world containing less suffering than all the other options available. With this suggestion, Leibniz sought to explain how it's at least logically possible that a merciful God could create a world with the suffering that it has.

And then, in 1755, some years after Leibniz published his famous argument, a massive earthquake hit Lisbon on the morning of the first of November, the popular feast day of All Saints. A 15ft crack opened down the middle of the street. Locals watched the tide disappear only to return as a huge wave that drowned most of the city. 30-40 thousand people were killed.

It was in the face of this terrible disaster that Voltaire came to mount his celebrated attack upon Leibniz in Candide. Voltaire cast Leibniz as the foolish Dr Pangloss, ready to trot out the absurd idea that this is the best of all possible worlds whatever misfortune befell him. The satire was biting. He was claiming that all theologians seem to care about in the face of human misery is getting God off the hook. Theodicy, Voltaire insists, is a moral disgrace and a sick joke.

Well, I have no answer to the question of how God can allow so many innocent people to die in natural disasters, like the earthquakes of Lisbon or Haiti. And indeed, I can quite understand that many will regard these events as proof positive that religious people are living a foolish dream like the idiotic Dr Pangloss.

And yet, I still believe. For there exists a place in me - deeper than my rational self - that compels me to respond to tragedies like Haiti not with argument but with prayer. On a very basic level, what people find in religion is not so much the answers, but a means of responding to and living with life's hardest questions. And this is why a tragedy like this doesn't, on the whole, make believers suddenly wake up to the foolishness of their faith. On the contrary, it mostly tends to deepen our sense of a need for God.

What many believers mean by faith is not that we have a firm foundation in rational justification. Those, like Leibniz, who try to claim this are, I believe, rationalizing something that properly exists on another level. Which is why, at a moment like this, I'd prefer to leave the arguments to others. For me, this is a time quietly to light a candle for the people of Haiti and to offer them up to God in my prayers. May the souls of the departed rest in peace.

Mr Deity and the evil

Another theistic explanation for natural disasters...

Haiti: Problem of evil and Sentamu vs Robertson

Letter on NSS Newsline here.

I awoke this morning (Thursday 14 January) to the struggled musings of John Sentamu, Archbishop of York, on the Today programme. He was attempting to reconcile the supposed existence of a loving, omnipotent God with the tragic events in Haiti. The only appropriate response he was able to muster was that he could offer "nothing which makes sense of the horrors". Had he paused at that point, then he would, at least, have been intellectually honest. But, no, he meandered onwards to utter several irrelevant, rambling, and borderline absurd opinions... read on.

Programme available for a short time here. I share the frustration of the author of the letter - what on Earth is Archbishop Sentamu on about (process theodicy?)?

Pat Roberston has a different view on Haiti: "True story...". At least Robertson's explanation is clear.