Thought for the Evening: The Logical Argument from Evil and the Free Will Defense
Analytic philosophy of religion is not a field that tends toward consensuses, but one of the most commonly accepted views, such that it can reasonably be said to be the scholarly consensus, is that the logical argument from evil is defunct. And in fact this is right, and so obviously so that it could just be taken for granted. I say 'could', however, because there have recently been pushes -- and they are pushes, because they are not based on any actual change of argument or evidence -- to try to reject the view. Setting aside cases of mere obstinacy, I find that all of these pushes tend to involve not understanding what is meant either by 'logical argument from evil' or by saying that it is defunct or dead.
So what is it? It's an argument with premise that evil exists and a conclusion that God does not exist that has certain specific characteristics. It is (and this seems to elude some people) a specific kind of argument. It is very important to grasp that there are many kinds of argument from evil that are not classified as 'logical argument from evil'; arguments that are probabilistic and thus not directly making the inference once based on logical consistency and inconsistency are the most obvious, but also many arguments that are based on some supposed logical problem associated with holding both that evil exist and that God exist. That is, to count as a 'logical argument from evil' there has to be a logical inconsistency, but it has to be the right kind. For instance, it has to be a general inconsistency, not an inconsistency with a specific fact. In particular, however precisely it is formulated, it has to be equivalent to the following set:
(1) God is omnipotent
(2) God is omniscient
(3) God is perfectly good
(4) There is evil
And it has to take this set to be an aporetic set, that is, it has to be a claim that it is logically impossible to accept these four. What is more, it was an ad hominem argument (in Locke's sense, not the fallacy sense), so all four have to be understood in ways that would fairly typically be accepted by thoughtful theists; that is, you can't cheat by making up definitions that get you the inconsistency -- you have to use definitions that would be generally accepted -- indeed, generally have to be accepted to be a typical theist.
Such an argument had been argued by a number of people (Mackie is generally held to have given the best exposition). Plantinga is often credited with 'killing' it, i.e., with showing that it is dead or defunct. This often seems to be understood as saying that Plantinga refuted it. That is, people who say that the logical problem of evil is not 'dead' often seem to mean that it is was not refuted, in the sense that Plantinga didn't show either that one of the premises were false or that there was no logical way to get from those premises to the conclusion that there is no God. Well, yes; but in that sense, Plantinga did not intend to refute it and never claimed to have refuted it, and nobody thinks he did, except for people who don't understand the topic. What Plantinga did -- and conclusively -- was show that an argument of the relevant kind (a general aporetic tetrad (1)-(4) whose premises would typically have to be accepted) was not possible, because using only standard methods of analytic philosophy and without even assuming as definitely true any particular features of any particular form of theism, one could always construct counterexamples in which (1)-(4) on their own were consistent. Plantinga actually did a very large number of things to lock this in as conclusive, but the most famous and for many people the most persuasive component of his work is what is known as the Free Will Defense or FWD.
There are slightly different ways of formulating the basic FWD, but the general framework is to construct by a set of hypotheses a scenario in which persons with free will are importantly valuable, and the relevant kind of persons with the relevant kind of free will cannot be had without the possibility of the existence of evil. For instance, let us suppose that morality requires free will in the sense that one can refrain or not from doing something without being determined one way or another; let us suppose, too, that morality, and thus any conditions required for it, are very valuable; let us suppose, too, that the world is one that gets its value in part from being a world with morality in it. Then it seems that it is impossible for this world to be arranged in such a way that people could be determined not to do evil.
It's important to grasp a few things about this argument, and it's easy to slip up on them.
(A) The FWD does not assume that the hypotheses in question are actually true. They just have to be intelligible. If they are, then you have just constructed a counterexample to (1)-(4) being an aporetic tetrad. They may be inconsistent with each other for some other reason, but you have shown that just taking them themselves, you can make suppositions that are not ruled out by any one of them on which they would be logically consistent. This is why the FWD is not a refutation but a 'defense'; it doesn't show that (1)-(4) are actually consistent in the world as we have it. It shows that if they are inconsistent, it is not because they are themselves logically inconsistent. There are possible situations in which they would be consistent, even if those possible situations are not actual.
(B) Because of this, the actual strength of the FWD is not in the particular hypotheses or suppositions chosen. Sometimes people get confused about this. Plantinga has his own preferred hypotheses for constructing a counterexample, and there are lots of arguments that have been made that this or that feature of one of his hypothesis is a problem for this or that reason. But Plantinga's preferred hypotheses for describing the FWD are just examples. What the FWD shows is that there is a general template for constructing counterexamples to (1)-(4) being an aporetic tetrad, one that is consistent with many different hypotheses, and this general template doesn't even have to appeal to features of any particular theistic view -- notice that none of the particular suppositions I mentioned above actually said anything directly about God, being only about morality and free will and their importance. An implication of this is that a theist can reject the argument that (1)-(4) are logically inconsistent in and of themselves without even having to defend any theistic claim at all. It follows from this in turn that rejecting the logical inconsistency of (1)-(4) doesn't require an atheist to accept any directly theistic claim. And there are lots and lots of different possible variations on sets of hypotheses about free will that could get you the same general kind of counterexample. If even one such set of hypotheses is logically consistent, then (1)-(4) is not an aporetic tetrad, and this will be the case whether God exists or not. In fact, you could add to the FWD the supposition 'God does not actually exist', and still get a counterexample to (1)-(4) being logically inconsistent.
(C) And notice, again, that this is all independent of any distinctive features of any particular form of theism. This all plays out independently of any argument that any theist might have that (1)-(4) are consistent on his interpretation of (1)-(4). It can all be done purely hypothetically, with the inevitable conclusion that if (1)-(4) cause any problem for theism, it is not because of (1)-(4) and their logical implications themselves, which was the point under discussion. Of course, it's likewise true that if a theist wants to give a positive argument that (1)-(4) are consistent in themselves, this would give yet another reason for thinking that the logical argument from evil would fail; the FWD is not the only conceivable way of proving that (1)-(4) are not an aporetic tetrad -- it's just a general recipe that allows you to construct extremely large numbers of reasons for holding that it is not, and if any of those reasons are even possible, it is not an aporetic tetrad.
(D) Because the FWD provides a general template for counterexamples, it doesn't matter what evil we are considering, either. An early attempt to get around the FWD argued that it could only handle moral evil and not natural evil. And Plantinga famously pointed out that this is not true. Here's another hypothesis we can add to our set to run an FWD for natural evil: All natural evil derives from moral evil -- to give just one of many possible ways one might understand this, perhaps all natural evil is due to ultimately to the instigation of the devil. Again, it's not at all relevant to the immediate point whether this is true or not. All it has to be is intelligible, and if it is, you have a created a counterexample, for natural evil, to (1)-(4) being logically inconsistent in and of themselves. And since FWD is a general template for endless numbers of counterexamples, if you don't like one, you can pick another by either modifying the hypotheses or adding new ones. For instance, maybe the version of the hypothesis you pick is that all natural evil is due to the choices of Adam and Eve, or for that matter due to your own choices. It doesn't have to be true; it doesn't even have to be plausible; it just has to be coherent and intelligible. And here is one of the grand kickers (and one of the reasons the FWD was able to convince so many atheists): If you use a hypothesis that is provably false on evidential grounds, that would show it to be coherent and intelligible enough to prove false and show that it can be used for constructing counterexamples using the FWD. You can generate logical counterexamples from counterfactuals, so proving that something is counterfactual can prove it to be a logical counterexample. You can be an atheist, using hypotheses you think are provably wrong on the evidence, and still get a logical counterexample even for natural evil. And we can do this kind of thing with any kind of badness, so modifying which evils you are talking about doesn't change anything. The FWD does not apply only to moral evil.
Now, as previously noted, there are many arguments from evil other than the logical argument from evil, in the sense being considered; there are, for instance, arguments purporting to prove from (1)-(4), plus additional suppositions that are thought reasonable by those who propose them, that God cannot exist. In effect, since the FWD is purely based on hypothesis, and is proving something about logical structure, an atheist could likewise just as easily propose hypotheses that, when added to (1)-(4), make the set inconsistent. The FWD is not a refutation of the claim that you can get an atheistic conclusion from (1)-(4), and was never intended to be. But what people had been wondering was whether (1)-(4) on their own, understood in typical ways, prove such a conclusion, and Plantinga in the FWD, as well as other lines of argument proposed by both Plantinga and others, showed that they do not. Why does this matter given that there are so many other possible arguments from evil? Well, first, the argument that the FWD shows to be impossible was an actual argument that people had proposed; it was not a hypothetical argument for atheism, but one that people actually put forward. Second, the way the FWD was formulated did not presuppose a position about theism or atheism as such, and since it showed that the question of God's existence was not logically provable from (1)-(4) on their own, it showed that the real point of dispute was about one of two things, either (a) other hypotheses of the sort that you could add to (1)-(4) to create inconsistency, or (b) whether you could get from (1)-(4) to the atheistic conclusion by a weaker sort of provability than logical provability. And, indeed, everything that has been argued since is one of these two -- either people argue that you can still get logical provability if you add particular suppositions to (1)-(4) or that (1)-(4) make God's existence less likely. So it was undeniably a significant conclusion for the shape of later discussion.
And it still holds. There are practically endless numbers of ways in which you can show that an argument with the properties the logical argument from evil was intended to have can't be constructed. Even if you assumed (what no one has proved) that every single one of the possible counterexamples were not actually possible, the FWD would still show that actually constructing an argument with the properties required for it to be a logical argument from evil in the relevant sense would require an utterly unfeasible amount of work for anyone to get through, because constructing it would require having similar logical proofs (about free will, morality, etc., etc.) over large areas of philosophical inquiry. Thus people say that the logical argument from evil is dead or defunct.
Various Links of Interest
* Irem Kurtsal, Russell on Matter and Our Knowledge of the External World (PDF)
* Craig K. Agule, Defending Elective Forgiveness (PDF)
* An interesting news story about the legal right not to be 'fun' at work
* Gideon Lazar, Creation and Divine Freedom
* Researchers recently figured out how to decipher a coded message written by Emperor Charles V in 1547.
* Eleanor Parker, The Sermon of the Wolf, on St. Wulfstan
* Daniel van Wachter, Roman Ingarden's Theory of Causation Revised (PDF). This is interesting, although I think Ingarden's theory is in some ways better than the suggested revision.
* Billy Trout, Monkey Speak, Monkey Learn
* Christ A. Kramer, The Philosophy of Humor: What Makes Something Funny?, at "1000-Word Philosophy"
* Louise Daoust, Shepherd on Causal Necessity and Human Agency
* David Landy, Shepherd on Meaning, Reference, and Perception
* Kenneth L. Pearce, Astell and Masham on Epistemic Authority and Women's Individual Judgment in Religion (PDF)
Ayn Rand, Atlas Shrugged
Thomas Joseph White, OP, The Trinity: On the Nature and Mystery of the One God
D. G. D. Davidson, Rags and Muffin
On audiobook: James S. A. Corey, Cibola Burn