Here is a well-known ad hominem argument in the philosophy of religion. Take a philosopher who both employs the problem of evil to argue against theism and also endorses an account of morality that explains it away in one way or another. Such a philosopher is rightly the object of philosophical ridicule.
Here's perhaps another.
I was teaching Mackie's stuff on miracles, and thought of the following combination of views. Suppose a philosopher buys Mackie's arguments, and also wants to use the problem of hiddenness as an argument against the existence of a loving God. An ad hominem argument against such a philosopher may be appropriate here as well.
Here's why. If you buy the hiddenness problem, you are complaining that the evidence isn't clear enough. What is one asking for here? Presumably not that God turn one of the classical proofs into a good argument when it in fact isn't. And if you buy Mackie's Humean stance against miracles, you can't be asking for a miracle to convince you of theism. So all that remains, it seems to me, is some natural phenomenon that would convincingly display the existence of God. But that isn't a promising line, given the plausibility of the idea that science is at least methodologically naturalistic. Given this idea, any natural phenomenon should be treated as fodder for explanation by the natural order, and failure as only a sign that we need better natural theories rather than as a sign of something supernatural. Failure to do so is just to resort to a "God-of-the-gaps" theory, and those are disreputable.
I'll close with a request for input. If you buy the hiddenness objection as well as buying the argument that it can never be rational to believe that a miracle has occurred, how could this combination of views be coherent?
How's this?:
"God could create me such that it is self-evident to me that he exists. I would simply find myself having a basic belief in him. Why didn't he do this, at least in my case?"
Even if he takes the view that it would *be* a miracle for God to create a human being with theistic basic belief, he wouldn't be asking for his theistic belief to be inferred from that miracle. In this context, the objection from hiddenness is not that God has failed to provide clear evidence by which we might infer his existence, but that God hasn't made his existence self-evident, when this would be trivially easy for him to do.
So he might continue to hold that it's never rational to believe a miracle has occurred. He's still complaining about divine hiddenness, but not in a way that undermines the Humean perspective.
Is *this* combination of the hiddenness and Humean theses incoherent?
There are two suggestions here, one that it be self-evident that God exists and the other that it be a basic belief that God exists. On the former, that's like asking that a proof be made good when it is bad. The epistemic principles that specify self-evidence are necessarily true if true at all, so if God's existence could be self-evident, then it already is. On the latter, merely having a basic belief--one not inferred from other beliefs--won't do the trick. I doubt the problem of hiddenness would disappear if God zapped people into believing in him without argument. We'd then just have meta-incoherence when the objector reflected on the epistemic status of such a belief, being unable to explain how such a belief is rational. Such meta-incoherence would happen as in your paragraph about the belief being miraculous. You'd ask yourself where this belief came from, conclude it could only be miraculous while also recognizing that it can never be rational to believe that a miracle has occurred. It would be better then to ask for a *properly* basic belief, one for which one could find good reasons on reflection. That won't get off the ground any quicker than the self-evident possibility though, I don't think.
If you buy the hiddenness objection as well as buying the argument that it can never be rational to believe that a miracle has occurred, how could this combination of views be coherent?
Suppose that these together entail that God is necessarily hidden (or necessarily rationally believed so). Can we conclude that the complaint that God is hidden is inappropriate? I'm not so sure. Certainly many have complained that "God would necessarily exist, were he to exist at all" (just as God would be more easily known) along with "some worlds are uninhabitable by God" (just as no miracles are possible or rationally believed so). Does the latter fact (assuming it is one) make the former complaint inappropriate? Can't you complain that God would necessarily exist (were he to exist at all) though, as a matter of necessity, God could not exist in every world? Is that complaint inappropriate?
"The epistemic principles that specify self-evidence are necessarily true if true at all, so if God's existence could be self-evident, then it already is."
I understand this response, but doesn't it come at the price of having to say that externalism about self-evidence is incoherent? What if someone's account of self-evidence wasn't in terms of the proposition's conceptual content (strictly speaking), but rather that someone simply sees a self-evident claim to be true upon grasping or understanding it (Plantinga, "Reason and Belief in God," 56)? On this view, whether a claim's truth is self-evident to us might depend upon our design plan (perhaps upon God's decision to make us a certain way), in which case, it's a contingent matter, and not necessary.
I'm not arguing this view, just pointing out that the alleged incoherence of hiddenness + Humeanism on miracles now crucially depends upon externalism about self-evidence being incoherent (and not just false). That's a pretty significant cost of this response, correct?
"I doubt the problem of hiddenness would disappear if God zapped people into believing in him without argument. We'd then just have meta-incoherence when the objector reflected on the epistemic status of such a belief, being unable to explain how such a belief is rational."
Well, I don't see why we need to be able to *explain how* a basic belief is rational, in order for its basicality to confer *prima facie* warrant on it. Let's apply this to mathematical beliefs. Do I really need to be able to explain how my belief that "2+2=4" is rational, in order to avoid epistemological "meta-incoherence"? If not, why then for basic theistic belief?
The idea here is that, according to my imaginary objector in the posted comment above, if God gave us basic belief in himself such that it would strike us as self-evident that he existed, at the very least theistic belief and (simple) mathematical belief would be on an epistemic par. So the objector says: why didn't God do this? I'm not saying, of course, that this is a good position to take, just that this version of the hiddenness objection can be coherently conjoined with the Humean epistemological stance on miracles. Am I wrong in continuing to think this?
Well, the comment implies that externalism about self-evidence is false, but I don't think incoherence is required. Holding an externalist theory of self-evidence plus endorsing the self-evident character of a basic belief doesn't make it so.
On the math example, I didn't say we needed to be able to explain how in order to avoid meta-incoherence. Meta-incoherence results when you try to explain but reach the conclusion that it can't be explained. So what I was claiming is that if we are reflecting philosophically on the status of a belief, and hold that it is a properly basic one but that there is no good explanation of propriety here, that's incoherent. Coherence could be salvaged by making the reflection less complete, of course: only reflect to the point of endorsing proper basicality, but don't reflect enough to reach the conclusion that propriety can't be explained. Such a rescue is possible, but shouldn't provide much intellectual comfort to the hiddenness objector who also denies the possibility of rationally believing in miracles.
Mike, I'm not sure the complaint would be inappropriate by such a person. A person can hold an incoherent position, and yet appropriately claim all sorts of things involved in the position. Appropriateness of assertion, for me anyway, is a function of justified belief, and I was careful not to treat the incoherence here as implying anything about what was justified and what wasn't.
Jon,
No problem. Let me rephrase the question. Do you think the latter objection is incoherent? I'm not sure it is. It seems to be claiming that the existence of God entails an inconsistency. So God does not exist. Similarly with the former objection. If God exists, then it's true both that he would be more easily known and could not be known.
Dear Jon:
It's fun to respond to old stuff...
1. Your point on hiddenness is very clever and seems basically right. The only objection I can think of is that one might believe that (a) if God exists, God exists necessarily; and (b) all necessary truths have finite proofs from self-evident premises. If (a) and (b) hold, then if God exists, there is a proof of the existence of God. The problem of hiddenness, then, turns into the question of why God hasn't revealed the proof.
2. My own solution to the problem of hiddenness is to maintain that there are sound cosmological arguments from premises that it is epistemically perverse to deny. :-)
3. "Take a philosopher who both employs the problem of evil to argue against theism and also endorses an account of morality that explains it away in one way or another. Such a philosopher is rightly the object of philosophical ridicule."
Graham Oppy in Arguing about Gods defends the coherence of being such a philosopher. The idea is that instead of positing the existence of evil in the argument, one posits the existence of that which the theistic opponent would call evil.
Alternately, one might use conditionals. "If there is such a thing as objective morality, then murder is an evil. If God exists, there is such a thing as objective morality (being objectively morally good is one of God's attributes). Therefore, if God exists, then murder is an evil." Now, if E is evidence against H, so is (if H, then E), and at least to the same extent (unless my quick Bayes' theorem calculation is off).
What do you think?
For clarity: the second suggestion under point 3 is mine, not Graham's if memory serves.
Hi Alex, the Oppy response confuses me a bit. At first, I thought it was just the idea of aiming for a reductio, as in your conditional argument later. But it isn't that, is it? So, what would a theistic opponent call evil? Maybe nothing! Maybe the works of Michaelangelo. . . So maybe instead we should say, "the stuff people like Hume and Russell and Rowe talk about as a problem for theism." And then treat the description as fixing reference, so that we're asking about the evidential relationship between *that stuff* and theism. I don't see how this will help unless the stuff in question is bad and conflicts in some way with the goodness of God.
Better, I think, to try your way of tying good and evil to God's existence and then trying for a reductio.
But there are problems here too. Reductios of the usual sort would require that the deductive argument from evil works. But what is an inductive reductio? What you'd get from such an argument is that there is epistemic tension in the theist's position. To which one should probably say that there's *a lot* of tension in any position that denies the existence of evil!
OK, just another ad hominem, I know! About the conditionals argument: I'm inclined to think that fundamental moral claims are necessarily true if true at all, and that makes conditional reasoning about it more complicated that usual conditional reasoning. Especially for bayesian calculations!
But intuitively, if we think of the conditionals as contingent ones, and we treat them as indicative conditionals rather than material conditionals, we get further problems. Suppose I'm at the zoo and suppose H=that is a mule, and E=I am seeming to see a zebra. Do we think that E is evidence for ~H (evidence against H)? Seems right. How about the claims that if that thing is a mule, then I'm seeming to see a zebra. That seems wrong, though. It would be a strange response to answer your son's request about your evidence that it's not a mule by saying, "well, if that thing is a mule, then I'm seeming to see a zebra."
So we need to interpret the conditional as a material conditional if we have any hope here. Then we'll get the same result as if we use the claim that either that's not a mule, or I'm seeming to see a zebra. Since each disjunct is evidence against H, the disjunction is.
But if this is how the argument is supposed to go, then the theist can respond as follows. "I agree with you nontheists that there is no evil. That was an invention by the powerful to control the weak. There are only states of affairs and preferences. So it is true that God prefers some things and doesn't prefer other things, and that distinction lines up fairly well with the classes of things we've been calling good and evil. But there's no problem of what God doesn't prefer like there might have been thought to be a problem of evil. God's preferences sometimes conflict (wanting us to be happy and also wanting us to chart our own course, for example, or wanting no suffering but also wanting knowledge to be possible) and that's all we need to cite to explain why some things happen that God doesn't prefer. The existence of things that a deity like the one classical theists conceive wouldn't prefer isn't any sort of decent evidence at all there isn't such a deity."
Dear Jon,
Happy Easter!
Yes, the "what the theistic opponent would call evil" solution is problematic. It would seem to be like arguing that Jones is looking at what he would call a circle, and hence he is not looking at a square. That argument is only persuasive if there is good reason to think that Jones correctly identifies circles, but the irrealist about good and evil cannot make this move.
I did mean the conditional to be a material conditional. The response you suggest on behalf of the theist seems to only work for theists willing to reduce good and evil to divine preferences. But this reduction seems implausible since it would mean that God's being praiseworthy for being good is identical with God's being praiseworthy for being such as he prefers himself to be. But even if there is a kind of praise that someone deserves for being at peace with his preferences, that is not the kind of praise that God deserves for being good.
Maybe, though, you want to argue that any reason to reject this kind of theism is also a reason to reject the non-theist's moral irrealism. That seems right.
(By the way, do you think suffering is needed for knowledge? It seems that any role that suffering plays in gaining us knowledge could be played by quasi-memories of unreal sufferings. God could implant in us such quasi-memories, and could even bring it about that each time we re-experience these quasi-memories we have a veridical religious experience of God reminding us that the quasi-memories are non-veridical.)
Hi Alex, I didn't really mean what I wrote! I don't think suffering is necessary for knowledge. I meant to be referring to the existence of natural evil, where once the causal structure of the world is in place, divine interventions that occur too often would rob us of the capacity to understand the natural order of things. So too many miracles, too little knowledge. Perhaps there are ways to design the system so that no suffering or evil ever occurred in it (no death, for example), and hence no conflict between evil and knowledge, but I don't know.
On the praiseworthy point: if there is no such thing as good or evil, I would think all claims about praiseworthiness would be false as well. I agree this is extravagantly revisionary, but if the nontheist can defend the a-axiological point about good and evil, this is what we should conclude that the language of praiseworthiness, to the extent that we want to salvage it, needs to be reinterpreted in terms of attitudes of approval on our part, or some such antirealist approach. Not satisfying at all, I think, but I don't see that there is an alternative here.
Dear Jon,
I was wrong to set this up in terms of a material conditional.
But maybe the way to reconstruct the dialectical situation is this. Our atheist, call him "Makkie", has some arguments against the existence of good and evil. However, at least some of these arguments are not particularly persuasive if theism is true. For instance, the argument from naturalism against axiology is unsound if theism is true, and we may suppose that this argument is one of Makkie's main motivators. Moreover, evidence for theism is evidence for the existence of something objectively praiseworthy, and hence is evidence against axiological irrealism.
Thus, Makkie believes:
Moreover, he believes the third claim not just as a material conditional (in that sense, it follows from the first), but as a bona fide ordinary language indicative conditional.
Moreover, Makkie believes:
Again, he believes it not just as a material conditional.
This is not an obviously unreasonable stance. For instance like Peter van Inwagen, I believe there are no artifacts. But I also believe (and I expect so does Peter), non-vacuously, that if there are artifacts, then there are laptops and sculptures.
So now the question is whether Makkie could be reasonable in believing that the existence of evil provides evidence against the existence of God? Well, Makkie could reason as follows. Now, if Makkie further accepts a deductive argument from evil, then Makkie can say:
If there are no true axiological claims, then there is no God (by (C)).
This does not strike me as an unreasonable argument, if, contrary to fact, the deductive argument from evil works. Moreover, it seems to me that something of this sort should also work even if the argument from evil is non-deductive.
Moreover, Makkie can then turn his reasoning against those theists who are more strongly committed to axiological realism than to the existence of God. There surely are many such theists. (On reflection, I don't think I am one of them.)
Yes, if the deductive argument from evil works, this is surely correct. I'm not convinced yet about the non-deductive case (which I think is the only interesting case, anyway). But I think we agree on the structure of the debate, since if you'll give up theism before axiological realism, you may be in trouble. But when the atheist starts from the assumption that ax-realism is false, he can't sneak in the claim that a theist is committed to ax-realism.
I don't object, of course, to the power of the argument you give to make reasonable a certain view of things. But it can do that and yet be a bad argument! :-)
I believe there are no artifacts. But I also believe (and I expect so does Peter), non-vacuously, that if there are artifacts, then there are laptops and sculptures
But if for PVI reasons of compositionality you believe there are no artifacts, then you should conclude that necessarily there are no artifacts. In that case your conditional is vacuously true. There are no worlds in which artifacts are not composite (excepting, I guess, gunk worlds). Is that where you get the contingency of 'there are no artifacts'?
Mike:
I think there can be non-vacuous per impossibile conditionals, both indicative and subjunctive. I don't have a good account of them, but neither do I (nor anyone else as far as I know) have a good account of non-per-impossibile conditionals.
For an example of a per impossibile conditional of the sort I'm using here, suppose that Kripke is right about essentiality of origins. Nonetheless, the following conditional is more than just trivially true: "If George Bush is the biological son of Golda Meir, he is Jewish in rabbinical law", whereas the following conditional is true only in the trivial sense: "If George Bush is the biological son of Winston Churchill, he is Jewish in rabbinical law."
I think there can be non-vacuous per impossibile conditionals, both indicative and subjunctive.
Yes, Alex, I've heard this asserted before. Zagzebski has a paper (I think) in which see urges that subjunctive conditionals with impossible antecedents are (sometimes) non-trivial. All good. But I guess I'd like to see some suggested semantics for these. I'd like to see the corresponding metaphysics. It looks to me like--if what is offered is not some tinkering--the semantics/metaphysics will have to go fictional. Is it worth the tradeoff? I doubt it.
I heard David Lewis say something along those lines a couple months before he died. I don't know if it was ever a part of his formal semantics for counterfactuals or if this just occurred to him at the end of his life. He didn't have a lot of time to put it into a paper after I heard him say this, and it wasn't part of some formal presentation, so I had no idea if it had even occurred to him before.
There is a book by Brandom and Rescher giving a semantics for dealing with impossible worlds, but I must confess I haven't read it.
My own view is that any particular per impossibile counterfactual contextually encodes a meaningful claim, but I am not sure a general semantics is possible.