Comments on the Craig-Rosenberg debate, part 2: the so-called “logical argument from evil”

I’m going to dedicate most of this post to Rosenberg’s opening statement, but first let me say “here here” to something from Jason Rosenhouse’s comments on the debate:

Craig proceeded to unload eight “arguments” for God’s existence. I use the sneer quotes because in most cases he was just making assertions, and not arguments.


Okay, now Rosenhouse’s comments on Rosenberg’s opening statement:

From my perspective as an atheist, Rosenberg was pretty disappointing. Stylistically Craig was the clear winner. He is a much more polished and confident speaker. Worse, however, was that substantively Rosenberg was also not so impressive. He made some good points, and I can certainly understand his frustration at having to respond to Craig’s mostly foolish arguments, but the fact remains that his responses were frequently unconvincing, or at least needed to be expressed more clearly to be convincing. His biggest blunder was in presenting the logical argument from evil, in which it is asserted that there is a flat-out contradiction between the existence of a loving God and the existence of evil. This is as opposed to the evidential argument, which states merely that the prevalence of evil is strong evidence against God.

I agree with what Rosenhouse says about style. I got the impression that Rosenberg did his homework on Craig and knew what he’d be up against; what was missing was Rosenberg rehearsing his opening statement 20 times in front of a mirror. But I want to take issue with what Rosenhouse (and many other people) have been saying about Rosenberg’s alleged use of the “logical” problem of evil.

In general, the insistence of people who follow these issues on classifying versions of the problem of evil as either “logical” or “evidential” is weird. It isn’t something you see with any other kind of argument in philosophy. What we care about with deductive arguments is first whether they are valid, and second whether the premises are true, whether people can agree that they’re true, whether people should agree that they’re true, etc.

If there’s agreement that an argument is deductively valid and the premises are true, it doesn’t matter if the premises are logical truths, if they’re necessary or contingent, or if they’re a priori or a posteriori. It matters somewhat whether we’re certain the premises are true, or whether we just think they’re just probably true, but thinking an arguments premises are only probably true doesn’t turn the argument into an “evidential” or “probabilistic” argument.

(Note: when I started writing this post, I vaguely remembered Theodore Drange making the same points I made in the above two paragraphs in his book Nonbelief & EvilBut on checking the book, Drange’s points were only kinda similar. Is there any chance I was thinking of someone else?)

As far as I can tell, Rosenberg only used the word “logical” twice in his opening statement in reference to the problem of evil. He called the argument a “logical deduction,” but that could be said of any deductive argument, and so doesn’t tell us anything about the content of Rosenberg’s argument. And he also said that, “it’s not enough to fob it off on the mystery of God’s plan or on the mere logical compatibility of these two views,” suggesting that he isn’t worried about the logical status of the argument’s premises.

Based on this, it seems to me that dismissing Rosenberg for “presenting the logical problem of evil” is wrong. What is true is that the sort of very simple argument from evil that Rosenberg gave is not popular among atheist philosophy of religion specialists. But it’s not like there’s a philosophical consensus against those arguments, because they’re quite popular among atheist philosophers not specializing in PoR. I could give many examples: Massimo Pigliucci, David Lewis, some of my undergrad professors, etc.

What religious apologists would like to say about this is that you should trust the people who specialize in PoR because they know the issues best. Therefore Lewis, Pigliucci, Rosenberg et al. all don’t know what they’re talking about. But there’s another explanation: that the non-PoR folks are more or less aware of what the PoR folks are saying and just unimpressed, in fact so unimpressed that they see no point in trying to contribute to PoR. I won’t try to argue for one explanation or the other here here, except to note that the situation is a bit odd. After all, evolutionary biologists have had no such trouble convincing the other biologists of the claims of evolutionary biology.

That said, I do have a worry about Rosenberg’s argument. Rosenberg claimed that if an all-powerful, benevolent God existed, he’d prevent all suffering. What exactly does that mean? If it just means any kind of pain at all, it seems easy to make a “greater goods” defense work there. BDSM aside, there can be something satisfying about, say, being sore the day after a good workout.

Now, to my ear “suffering” has much stronger connotations than “pain,” so maybe Rosenberg meant something more than just “pain.” But he would have been on firmer ground if he had explicitly restricted himself to talking about things like the Holocaust. (And had he done that, IMHO, there would’ve been no need to do anything fancy with the structure of the argument.)

  • Verbose Stoic

    The pain/suffering thing is a big conflation in a lot of the arguments. Even without the “satisfaction” argument, pain is really nothing more than a really strong aversive stimuus, which means that it stops you from doing certain things or lets you know that something is wrong. Interpreted that way, pain is actually GOOD, and if God eliminated it then He’d have to invent something to be as aversive … which would then be essentially the same thing as pain. So if you want to base a “Problem of Suffering” on pain, it’s easy to see how pain is actually good, but if you don’t simply use pain then a lot of arguments — like that of animals and their suffering — are a lot harder to pull off.

  • Epicus Montaigne

    I’m currently re-reading van Inwagen’s Problem of Evil and he agrees with you, on the near-meaningless and largely unhelpful distinction of a “logical” and “evidential” problem of evil. When you talk about restricting his comments to things like the Holocaust, its a “local problem of evil” and thats in opposition to the “global problem of evil”. Its a pretty good work all things considered. Also, he’s not strictly PoR, but he likes some of them.

    • Chris Hallquist

      Yes, van Inwagen’s position there is pretty similar to my own (though re-skimming his book just now, I don’t see that he gives a reason for his position).

  • miller

    You seem to be saying that the distinction between “evidential” and “logical” arguments is merely one of whether the premises are certainly true, or just very likely true. But I think the distinction has to do with the type of inference (ie deduction vs induction), not the type of premise.

    And I believe that this distinction matters. For example, I would say that most ontological arguments are logical, and most teleological arguments are evidential. And I think this makes a difference, because when the ontological argument fails, it fails utterly (contra Plantinga, who claims that it at least shows the reasonableness of god-belief). But I think it is arguable that the teleological argument only fails in degree. That is, it slightly improves the probability of gods, but not appreciably.

  • MNb

    “Interpreted that way, pain is actually GOOD”
    Great, VS. Now please apply this excellent logic to the concrete case of Elisabeth Fritzl – something apologists always fail to do and something CH mentions in his last alinea.

  • Steven Carr

    I think Christians like to split the problem of evil up into ‘logical’ and ‘evidential’ arguments, so they can respond to the ‘logical’ argument by saying that only an evidential argument can work, and can respond to the ‘evidential’ argument by saying it is not a logically sound argument.

  • Carneades-Skeptic Griggsy

    Fr. Meslier’s the problem of Heaven trumps all defenses and theodicies. PLantinga depends on the argument from ignorance with the greater good and the unknown reason defense arguments.

    • Patrick

      It doesn’t actually trump all defenses and theodices.

      Look. Arguing with a theist about the logical coherence of the theist’s theology is a fool’s game. Its like arguing with Calvin over the rules of Calvinball. They can just make stuff up, so you can’t get any ground.

      Take the idea of middle knowledge. There are tons of people out there who consider middle knowledge to be an accurate description of god’s way of viewing the world. But its not like they figured this out, I dunno, building a giant telescope and watching god to see if he behaved like he had middle knowledge. Instead, skeptics accused theists of holding incoherent beliefs. Middle knowledge seemed like a good way of getting out of that incoherence. So they decided it was true.

      They could have come up with a different way out. There are probably innumerable ways they could have avoided the problem by writing new theological beliefs into their religion. When you can make stuff up, the sky’s the limit, right? But middle knowledge was what they picked. It was capable of getting them out of a theological jam, so they decided it was real.

      See the problem? The best you’re ever going to accomplish in this argument is to highlight the theistic willingness to make stuff up. Ironically, you can’t demonstrate logical incoherence amongst beliefs that are descriptively incoherent. But that’s what you’re dealing with, like it or not.

      • Steven Carr

        ‘Take the idea of middle knowledge. There are tons of people out there who consider middle knowledge to be an accurate description of god’s way of viewing the world.’

        Middle knowledge?

        That is Craig’s and Plantinga’s firmly held belief , that although Batman, the Joker and Gotham City are fictional characters, their hypothetical god really does know exactly what Batman and the Joker would freely choose to do as they battle each other in Gotham City, if their god had decided to create them.

        Err, guys, these characters are fictional. Your hypothetical god can’t know what they really would choose to do if they existed.

        Because they don’t, you know, exist…..

  • Jeffery Jay Lowder

    I could be wrong, but I’m pretty sure the person who made the point about the relative insignificance of the logical vs. evidential distinction is Daniel Howard-Snyder in his introduction to his anthology, The Evidential Argument from Evil, p. xvi:

    “While we may easily draw this distinction [between logical and evidential arguments from evil], we are hard pressed to defend its significance. I’m afraid I don’t have much to say about the matter, except for the obvious. I mean, whenever one meets what its author purports to be an argument from evil (in contrast with, say, the sort of response to horrendous evil that you might find in Albert Camus’s The Plague or Eli Wiesel’s Night), one ought to consider whether the author intends to assert that facts about evil known with certainty are incompatible with theism. If she does, then one should query why the argument is not dubious for the same reason that Mackie’s argument from evil is dubious. If she does not–that is, if she intends to say that facts about evil which are known with certainty make theism significantly unlikely, or that theism is incompatible with certain facts about evil that are themselves quite likely–then one needs to think hard about the sorts of issues this book is about.

    Independent of whether the logical vs. evidential distinction is significant, I have a problem with it from a ‘naming convention’ perspective. Logic is commonly divided into two branches: inductive and deductive. Based on that, what is a student supposed to think when an author refers to a “logical” argument from evil? Are they supposed to say, “Oh, so your argument is, like, really logical?” And if “logical arguments from evil” are, well, logical, then does that mean that “evidential arguments from evil” are illogical?!?
    I think a less confusing set of labels are “logical incompatibility arguments” and “comparative improbability” arguments. According to arguments of the first type, some fact about evil is–you guessed it–logically incompatible with theism. Arguments of the second type claim that some fact about evil, while logically compatible with theism, is less probable on theism than it is on naturalism. These proposed alternative labels are not as concise as the “standard” terminology, but at least it’s clear what they mean!

    • Chris Hallquist

      Maybe that’s what I was thinking about? I did read that book several years ago.

  • Pingback: Thoughts on the “Logical vs. Evidential” Distinction