The following is an attempt to give a strong argument for a view that I myself do not hold. I think that the argument eventually fails but that it is worth pursuing, for the reasons why it fails are intriguing.
There are two versions of the problem of evil, the logical and the evidential. It has been admitted by most for some time now that the former is of little trouble to theism; various theodicies have been given to demonstrate the logical compatibility of evil in the world with the existence of an all-good God. The latter problem, however, is not so easily refuted or answered. Herein I shall argue that the strengthened challenge of the hypothesis of an evil god, combined with platonist moral realism, can withstand a strong plausible criticism directed towards it. If the argument stands, then, we are left with the conclusion that there is no evidential or empirical reason to rationally believe in a good God and, thus, no rationally compelling reason to believe in the God of theism.
2. The evil god challenge and good/evil symmetry
The evidential argument from evil – the argument that the existence of vast amounts of evil (moral and natural) is strong evidence against the existence of an all-good God – poses a serious threat to theism. The evidential argument can be bolstered by what Stephen Law calls the ‘evil god challenge’. This is to suppose the hypothesis that there is an omnipotent and omniscient creator of the cosmos, but that this being is maximally evil. The reason this hypothesis can go through is that the common arguments for the existence of God (e.g. the cosmological argument, or Aquinas’ five ways) leave out any description of the moral character of the creator. Quite possibly, God is really the evil god. Here the hypothesis runs into a parallel ‘problem of good’: there just seems to be too much good in the world for there to be an evil god. If this is the case then the evil god hypothesis can be bolstered by reverse theodicies, symmetrical to those put forward by theists for the compatibility of a good God with evil in the world – only now it is the compatibility of an evil god with the good in the world. For instance, one could say that free will makes possible moral creatures that will attain greater evil than they would without free will (with the inevitability of some good too), or that the existence of some very great second-order evils presuppose some first-order goods, or that (parallel to John Hick’s suggestion) the world is a ‘vale of soul destruction’ that requires the existence of some good.
The challenge has another part, namely the symmetry of the problem from evil and the problem from good. For every, or very nearly every, theodicy given to counter the problem from evil, on behalf of a good God, a symmetrical argument can be given to the problem of good on behalf of the evil god. Some examples were given above. The symmetry thesis states that there is a nearly isomorphic symmetry between good God and evil god theodicies, such that considerations as to reasonableness between the two are the same. And, for all we know, the amount of good and evil are on a par in regards to giving evidence towards the good God or the evil god, which means that both kinds of theodicies are equally reasonable, or symmetrical. Nevertheless, it is plain that evil god theodicies are unreasonable – and unreasonable to believe – given the great amount of good in the world. But if the two sets of theodicies are truly on par, then the good God theodicies are equally unreasonable – and unreasonable to believe – given the amount of evil in the world. As Wes Morriston has put it, “If this is right, then the really fundamental issue is not whether evil and suffering disprove theism, but whether there is any basis at all on which a reasonable person can believe God (if he exists) is wholly good.” This is the evil god challenge.
3. William Lane Craig’s argument for a good God
William Lane Craig has argued that the best defense of the goodness of God is a form of an argument from the existence of objective moral facts. That is, given the existence of objective moral facts God must exist as the source of those facts. Since those facts describe moral goodness God must be good, being the source of moral goodness. The argument runs as follows:
1. If God does not exist, objective moral values do not exist.
2. Evil exists.
3. Objective moral values exist.
4. God exists.
To this should be added a further deduction, mentioned already, that (5) such a God must be a morally good God, given that he is the source of moral goodness. An evil god would be incompatible with being the source of moral goodness and, therefore, there is an asymmetry between the evil god hypothesis and the God of classical theism. Now, Craig takes it that premise (2) is fairly straightforward and the atheist can grant it (after all, the evidential argument against a good God claims that there is too much evil). Likewise, premise (3) is granted by the atheist, either as a matter of principle or for the sake of argument. Premise (1), then, is the key to Craig’s argument (although, there are genuine concerns with premise (5) as well).
The standard atheistic response to an assertion such as that in premise (1) is to invoke the Euthphryo Dilemma: either x is moral because God commands it, or God commands x because it is moral. This is because the natural way to take the premise, and as Craig intends it, is to insist that if there were no God there would be no objective morality; thus, God must be responsible for the creation of objective morality. The dilemma, however, has two disconcerting horns: either morality is arbitrary (given the assumption that God could command willy-nilly), horn one, or the source of objective morality is independent of God, horn two. Russ Shafer-Landau takes it that the metaethics involved here is that of the Divine Command Theory and that, given the Euthyphro Dilemma, the theist should abandon horn one and cling to horn two. Horn one implies that God is imperfect and God cannot be imperfect, according to theism. But this is simply to deny the Divine Command Theory and affirm that the source of objective moral truth is independent of God after all – contrary to premise (1). The only way to preserve God’s integrity is to deny that he is the source of objective morality.
Strictly speaking, the classical theist has a way out of the Euthyphro Dilemma. Tim Mawson has suggested, following closely to the solution proposed by Richard Swinburne, that the theist could take both horns of the dilemma. That is, take horn one for contingent moral truths and horn two for necessary moral truths. God cannot change the necessary moral truths, since they are conceptual necessities – of the order of all bachelors are unmarried – yet these necessities depend, often enough, on the contingent features of God’s creation. So, given that God has created men that can marry (a contingent matter), it follows that no bachelor is married (a necessary matter). There is a similar story for all of the moral truths, and thus the objective moral values do depend upon God. Now, also strictly speaking, this need not be the only story: if the moral values or truths were platonistic abstract objects, akin to numbers, then the situation is much different. The best position for the atheist is thus to push the argument for platonism.
4. Platonist realism
Platonist realism holds that there are genuinely existing abstract objects, independent of concrete reality, that do not enter into causal interactions with any other existent entity. Realism about abstract objects holds that there really are necessarily existing entities such as propositions, properties, numbers, perhaps sets, possible worlds, relations, universals, etc, such that they could not fail to exist. Further, abstract objects, being necessary beings exist a se; they do not depend upon anything for their existence. The atheist (and perhaps even the theist, if Shafer-Landau is correct) can hold in all good order that moral values or truths are likewise platonic abstract objects, necessary in the way numbers are necessary. If this were the case then, clearly, premise (1) in the argument for the goodness of God is undermined; for it is no longer true that if there were no God there would be no objective moral values. At most God, if he were good, would be something of a ‘perfect thermometer’, accurately describing the moral landscape. Similarly, if there were an evil god the moral landscape could be absolutely identical – the only difference being that an evil god will not describe it accurately.
5. An indispensability argument for platonist realism The atheist criticism of Craig’s argument, and upholding the symmetry hypothesis of the evil god/good God theodicies, requires a defense of platonist realism. This defense can be readily given by the Quine-Putnam indispensability argument. Consider that mathematics is applicable to empirical science not just in a passing way but rather in a way that it seems impossible to do science without numbers (contra Hartry Field). A slightly weaker version of the Quine-Putnam argument, without the recourse to a question-begging naturalism, can be put as follows:
6. We ought to have ontological commitment to all the entities that are indispensable to our best scientific theories.
7. Mathematical entities are indispensable to our best scientific theories
8. We ought to have ontological commitment to mathematical entities.
If one were to change the discussion from science and mathematics to ethics and moral values, then the argument would straightforwardly vindicate the line of argument for platonist moral realism. Our best ethical theories require us to posit moral values (or truths), whereas they do not require us to posit the existence of God as the source of those moral values (or truths). We should therefore be committed to the existence of moral values even without God, but this pushes us towards platonist moral realism for the same reasons that the classic indispensability argument pushes towards platonist mathematical objects.
Now, it stands to reason that if we have a good argument for the existence of some platonic entities we have good reason to hold to the whole realm of potentially platonic entities. That is, it would be unparsimonious to hold to a platonic realm of real numbers but not to the platonic realm of propositions, say. The moral truths are merely a subset of all true propositions and so, given indispensability, we should think that they are in the platonic horde also. But, then, given the indispensability argument we have strong reason to think premise (1) above undermined.
The point of the discussion, from the point of view of atheism, is that the evil god challenge is not met by the plausible assumption that God must be good, given the need for God to ground objective morality. That argument was meant to demonstrate that there is an asymmetry between the evil god and the good God, but if platonism concerning moral truths is correct then the argument is undermined. Consequently, the evil god hypothesis is on safe symmetrical grounds with the good God various theodicies. In that case, however, the various good God theodicies against the evidential problem of evil are on a par with the evil god theodicies against the evidential problem of good – both sets of theodicies are equally reasonable or unreasonable. Both theism and atheism are compatible with a platonistic understanding of moral value; hence, appealing to objective morality is not sufficient to answer the atheist and, if William Lane Craig is correct, the best defense against the atheist’s evidential argument from evil is undermined.
Colyvan, Mark. “Indispensability Arguments in the Philosophy of Mathematics”. The Stanford Encyclopedia of Philosophy (Spring 2012 Edition), Edward N. Zalta (ed.)
Craig, William Lane. “The ‘Evil god’ Objection”.
Feser, Edward. “Laws ‘evil-god challenge’”.
Girwarnauth, Samuel. “No Platonism Given Theism: God and the Problem of Abstract Objects”. Presented at the 2013 meeting of the Society for Orthodox Philosophy in America, Kendalia, Texas.
Law, Stephen. “The evil-god challenge”. Religious Studies 46 (2010) – references to the online (2009) version, pages 1-21.
Mawson, Tim. “The Euthyphro Dilemma”. Think 20, Vol. 7 (Winter 2008): 25-33.
Morriston, Wes. “The Evidential Argument from Goodness”. The Southern Journal of Philosophy XLII (2004): 87-101.
Shafer-Landau, Russ. The Fundamentals of Ethics. Oxford: Oxford University Press, 2012.