Defending an Argument from Evil Against Naturalism: A Reply to Jeffery Jay Lowder
AS A RELATIVE NEWCOMER to Twitter, I recently tweeted out "An Argument from Evil Against Naturalism," which proceeds as follows:
1. Nature is all that exists.
2. Nature is amoral (neither good nor evil).
3. Evil exists.
...and concludes with this: "For nature to be all that exists, and nature to be non-evil, evil cannot exist." I then added a link to another page where my argument is further explained and the conclusion more tentatively qualified.[1] The basic idea of my argument is that one premise from the (logical) argument from evil against the existence of God – namely: "Evil exists" – actually works just as well, and possibly much better, in the service of theism against naturalism.
To my surprise, this very brief argument was met with an extended series of fairly technical objections, in around two dozen tweets, by Jeffery Jay Lowder from the Secular Outpost blog. Rather than post a series of counter-tweets to match Lowder's own string of tweets, I decided to answer the last bunch of his here, taking care to edit (as charitably as possible) his remarks for typos and clarity along the way. I have taken the time to answer Lowder not only because I felt genuinely challenged by his arguments, but because I have always regarded him as a fair-minded representative of the secular/atheist community. His comments are blocked off in the following ten points:
1. I think there is a lot of confusion, not just with you but with many people, about what it means for something to be an argument from evil. Some people, including you, take the word "literally" to mean any actions or states of affairs which a moral agent is morally blameworthy for doing, not doing, allowing to happen, failing to prevent, etc. Others use the word "evil" in a much looser sense to encompass moral evil, natural evil, dysteleology, imperfections, etc.
The key point of evidential arguments from evil is that there is some (known) fact about the world, for which it is improbable that God has sufficient *moral* justification to allow.
2. Examples of facts to which evidential arguments from evil appeal include:
- pain and pleasure
- flourishing and languishing
- virtue and vice
- triumph and tragedy
- failure of theodicies
- Divine silence in the face of tragedies
3. We need to be careful to distinguish propositions which include a normative or axiological claim from those which don't. Axiological proposition: Suffering is intrinsically morally bad Normative proposition: Torturing an innocent person is morally wrong
Non-normative, non-axiological proposition: Moral agents experience biological pain or pleasure which we do not know to be biologically useful.
4. Arguments from evil, including evidential arguments from evil, do not require the person defending the argument to endorse moral realism or objectivism. They do not contain any premises which presuppose moral realism or objectivism.
5. Arguments from evil do, implicitly or explicitly, appeal to supporting arguments which show that some fact about evil is inconsistent with T, improbable on T, or less probable on T than on N, based on T's entailment of God's moral perfection.
6. Another way in which many apologists incorrectly suggest a false equivalency or parallelism between atheistic arguments from evil and theistic arguments from moral ontology, is that the evidential claims in the two kinds of arguments are qualitatively different. Again, arguments from evil appeal to evidential claims which are non-axiological, non-normative in nature. In contrast, theistic arguments from moral ontology appeal to, well, claims about moral ontology (i.e., "moral goodness" is objective).
7. Your statement "evil exists" is ambiguous. "Evil exists" could include "People with certain terminal diseases die in horrific, gratuitous, pointless suffering." It could also mean "Objective moral DISvalues (like evil) exist."
8. Regarding "complaining," I don't think all instances of alleged "complaining" are actual complaining. For example, when I state an argument from evil, I'm not complaining in your sense. I'm offering evidence that your worldview is false.
Other people undoubtedly do "complain" in your sense. Let C="sense of injustice of unfairness." Your claim is that Pr(C | T) > Pr(C | N). You've stated your claim, but you haven't defended it. I think I can prove that N, conjoined with certain plausible background assumptions like the truth of biological evolution and the evolution of humans, leads us to expect C. I'm not sure about your claim that CT "entails" C, but regardless, CT is a much more specific claim than N, and so is intrinsically far less probable than N.
9. Christian Theism (CT) entails certain kinds of evil in the world, such as original sin, various moral evils committed by humans, etc. CT does NOT entail the specific facts about evil listed in my earlier tweet above.
10. Finally, I don't think you've come anywhere close to defending your claim that N entails that there are no objective moral values (and disvalues, like evil), or that Pr(OMVs | T) > Pr(OMVs | N).
In order for that argument to even get off the ground, we'd first need an argument to show that the existence of OMVs are contingent rather than necessary. That would be hard to do.
If "pain is intrinsically morally bad" is objective true, then it's hard to even imagine how it could be contingently true. But if OMVs are necessary rather than contingent, then Pr(OMVs | N) = 1, not because N entails OMVs, but b/c OMVs are necessary.
Here I will go ahead and answer Jeff's points in the order presented.
As the one allegedly confused here, I would say that (1) is more a point of disagreement than confusion on my part. I am aware that lots of people use "natural evil" to reference certain harsh and painful features of the natural world. But I would maintain, contrary to received wisdom in some philosophical circles, that such a usage is simply invalid, or at best, well, confusing, in the context of theist-naturalist debates. It's one thing to say that painful "natural" phenomena involve strictly non-moral elements, but quite another to say that such phenomena are rightly termed "evil" only because, after all, they would be considered evil when the consequence of moral actors. To put it another way: the very point of my argument from evil against naturalism is that given the specific definitions of natural and evil at issue, the phrase "natural evil" is conceptually incoherent.
To me and lots of other reasonably intelligent people, if a thing is natural (i.e., in the sense consistent with metaphysical naturalism) it is not evil (morally unjustifiable), and vice-versa. That said, anyone who presumes to harness instances of recognizable, identifiable "evil" in arguments against theism should be prepared to face those same instances in arguments against naturalism.
Let us consider the first of the facts supporting evidential arguments in (2). On what possible grounds could pain and pleasure be considered morally unjustifiable facts for God to allow? We should keep in mind that on naturalism, pain and pleasure are no more than sensations experienced by biological organisms. And on the face of it, there is nothing about sensations of pain and pleasure that could possibly indict God or anyone else of anything morally unjustifiable. Lowder, along with Paul Draper and others, argues that pain and pleasure are more probable on naturalism than on theism. But I would counter, as I have argued from the start, that our capacity to recognize the moral significance of pain and pleasure is much more probable on theism, particularly Christian theism, than on naturalism. (Here I would further agree with Otte, who says that Draper's listed observations of biological pain and pleasure are perfectly consistent with various theisms, including Christian theism, and therefore I disagree with Lowder and Draper.[2])
Whereas Lowder's descriptions of axiological, normative, and non-normative/non-axiological propositions in point (3) are useful and acceptable enough, (4) again strikes me as either incoherent or conceptually empty. One cannot argue from an identifiable instance of evil and at the same time from a standpoint of moral subjectivism or anti-realism. One might argue against theism from "sensations of pain" or "experiences of suffering," but then the argument would lose its force. What exactly about sensations of pain or experiences of suffering is inconsistent with theism?
Again the problem with (5) is with recognizing evil as such. Naturalism would not seem to permit such a recognition in the first place, in which case I can just as easily and validly counter that "some fact about evil is inconsistent with N, improbable on N, or less probable on N than on T, based on N's entailment of an amoral universe."
As for (6), I reiterate that no amount of non-axiological and non-normative evidential claims will "add up" to even a single identifiable instance of evil. To "cross over" from observing the fact of pain or suffering in the world to confirming an instance of evil would require the addition of rhetorical descriptors like "horrific," "gratuitous" or "pointless." But these largely subjective, emotive descriptors would themselves not be facts. If, however, at any point it can be truly and objectively said that there exists evil in the form of horrific, gratuitous, etc., suffering in the world, then at precisely that point the evidential argument becomes subject to the same basic form of logic offered in my original argument against naturalism.
I think (7) appears fair enough at a glance, but we should recall that my argument from evil against naturalism is drawn from the same premise used in arguments from evil against theism. If that premise is true, then it's true regardless of who makes the argument or for what view of the world. If that premise is false or ambiguous, then again it's false or ambiguous no matter who presents the argument or to what ends. Besides, the charge of ambiguity cuts both ways: to Christian theists like me, it appears that Lowder would like observable pain and suffering to mean "evil" in reference to theism, in order to demonstrate inconsistency with theism, and "natural" in reference to naturalism, to insulate naturalism from a similar criticism.
Regarding (8), I do think I have defended Pr(C | T) > Pr(C | N), but will restate it here. C, a sense of human moral awareness and accountability, and therefore of injustice or unfairness when it occurs, is a central tenet of theism, again particularly Christian theism. By contrast Lowder says he can prove that C is a rational consequence of N, naturalism, but only when conjoined with a host of various evolutionary assumptions. (C | T) is therefore a much simpler hypothesis than (C | N), and therefore, all other things being equal, Pr(C | T) > Pr(C | N).
As for Christian theism in particular, there is nothing about probability that would render a specific claim less probable than any comparatively broad metaphysical claim. Were that the case, my specific claim that the Red Sox won the 2018 World Series would be intrinsically far less probable than a claim that we are currently living in a computer simulation.
To answer (9), I can only say that Christian theism does entail, or at minimum adequately explains, the facts about evil mentioned earlier. There is no appreciable distinction in Christian theology between natural and moral evil. The sin of Adam led not only to his own death, but to the deaths of his children and all generations to follow, along with a curse unleashed upon the entire planet.
Finally, I don't follow the reasoning behind (10) at all. To my knowledge it's simply not the case that objective truths are necessary rather than contingent truths. It's an objective truth that Austin is the capital of Texas, but there's no principle of logic that necessitates such a state of affairs. (In other words, it would be false to say that there no possible worlds in which the capital of Texas is Houston, or San Antonio.) That clears the way, then, for affirming that theism does indeed entail objective moral values while naturalism does not, and therefore, Pr(OMVs | T) > Pr(OMVs | N).
Even if OMV's were necessary truths, the problem of OMV's facing naturalism would remain, simply by virtue of the logic of the original argument. That is, even if Pr(OMVs | N) = 1, Pr(N | OMVs) << 1, and more importantly, Pr(N | OMVs) << Pr(T | OMVs). So we are left pretty much where we started. Given objective moral values, naturalism is probably false.
[1] Don McIntosh, "An Argument from Evil Against Naturalism," OnFaith, https://www.onfaith.co/commentary/an-argument-from-evil-against-naturalism.
[2] Richard Otte, "Evidential Arguments from Evil," International Journal for Philosophy of Religion, 48: 1–10, 2000.