Sacrificing a theistic argument to the Problem of Evil
October 8, 2012 — 10:40

Author: Alexander Pruss  Category: Concept of God Existence of God Problem of Evil  Tags: , ,   Comments: 20

It’s hard to come up with reasonable priors for such theses as Naturalism and Theism and with reasonable conditional probabilities for such evidence as Evils We Can’t Theodicize on Theism. But we can sometimes come up with reasonable comparisons of the strength of evidence. And this might lead to some helpful non-numerical probabilistic reasoning.

For instance, we might have the judgment that the evidential strength of the Problem of Evil (POE) as an argument against theism is no greater than the evidential strength of the Finetuning Argument (FTA) as an argument for theism. Two thoughts in support of this: (1) the low-entropy initial state of the our universe has been estimated by Penrose to be utterly incredibly unlikely (my paraphrase of his 10^(-10^123)) and some of the other anthropic coincidences come with what are intuitively extremely narrow ranges; the theist has proposed various theodicies–they may not be convincing, but it seems reasonable to say that the probability that together they answer the POE is no less, indeed quite a bit greater, than the incredibly tiny probabilities that FTA claims; (2) just as thinking about naturalistic multiverse hypotheses significantly decreases the force of FTA, thinking about theistic multiverse hypotheses significantly decreases the force of POE (cf. Turner and Kraay’s work); (3) just as in the case of FTA we might worry that there is some nomic explanation of the coincidences that we haven’t found, so too in the case of POE we have sceptical theism.

This means that the theist can simply sacrifice FTA to POE: the FTA either balances POE or outbalances POE (I think the latter, because of point (1) above).

Then the theist has a nice supply of other strong and serious theistic arguments, such as the cosmological, non-FTA design arguments (e.g., Swinburne’s laws of nature argument), ontological, religious experience, moral epistemology (theism has a much better explanation than naturalism of how we can know objective moral truths), etc. The atheist has a few other arguments, too, but I think they are not very impressive (the Stone and other issues for the Chisholming of divine attributes, Grim-style worries about omniscience and infinity, worries about the interaction between the physical and nonphysical). At least once POE is completely out of the picture, even if FTA is lost, the theist can make a very strong case.

  • Patrick

    The theodicy outlined below called “Theodicy from divine justice” may show that it is not even necessary to sacrifice the fine-tuning argument to the argument from evil:
    (1) God’s perfect justice prevents Him from relieving people with unforgiven sins from their sufferings (see Isaiah 59,1-2).
    (2) Unlike God Christians are not perfectly just. Therefore, unlike God, they are in a position to help people with unforgiven sins. By doing this they may make those among them who haven’t yet accepted God’s salvation receptive of it (Matthew 5,16, 1 Peter 2,11-12, and 3,1-2), which in turn frees these persons from suffering in the afterlife.
    (3) The greater God’s beneficial power due to His love, the greater God’s destructive power due to His justice (see Matthew 13,27-29). Striving to prevent as much suffering as possible God can only interfere to such a degree that the beneficial effect of the interference is not neutralized by the destructive effect of it.
    (4) Someone who dies before he or she reaches the age of accountability, i.e. before he or she can distinguish between good and evil (see Genesis 2,16-17, Deuteronomy 1,39, and Isaiah 7,16) faces no punishment in the afterlife, as he or she would not have been able to commit sins. So, God may not be inclined to prevent such a person’s death.
    (5) A person’s suffering in this life may have a redeeming effect (Luke 16,25) and consequently contribute to a decrease of the respective person’s suffering in the afterlife; the amount of suffering in this life is so to speak subtracted from the amount of suffering in the afterlife. So, God may not be inclined to relieve this person’s suffering.
    (6) A person’s suffering in this life may make the person receptive of God’s salvation (Luke 15,11-21), which in turn frees this person from suffering in the afterlife.
    (7) There are degrees of punishment in the afterlife depending on one’s moral behaviour (Matthew 16,27, 2 Corinthians 5,10), one’s knowledge of God’s will (Matthew 11,20-24, Luke 12,47-48, John 15,22-25), and, as mentioned before, one’s amount of suffering in this life (Luke 16,25).
    (8) Those people who suffer more in this life than they deserve due to their way of life are compensated for it by receiving rewards in Heaven.
    (9) As for animal suffering, animals will be compensated for it on the “new earth” mentioned in Isaiah 65,17-25, 2 Peter 3,13 and Revelation 21,1.

    October 8, 2012 — 13:45
  • Mike

    Does it matter that the FTA is not an argument directly in favor of the existence of a typical theistic God, whereas the POE is directly against the existence of a typical theistic God?
    Let’s say they do indeed balance each other out in terms of the conclusion of each argument establishing itself with high probability within the context of just those two arguments and their rebuttals. It’s tempting to say that the POE outcome is telling us it’s highly unlikely that there is a loving (etc.) God. But the FTA is telling us it’s unlikely that the initial conditions of our universe arrived by chance.
    Hopefully you see my worry there about the different thrust of each, the result of which seems to still leave the POE in a better position.

    October 8, 2012 — 13:58
  • Mike:
    1. FTA claims that it is likely that the initial conditions of our universe were generated by an intelligent being who wanted there to exist intelligent life. But unless this being is very much like the God of theism (a necessary and perfect being), this explanation doesn’t really get us very far ahead, since we’d wonder why this being exists and wants there to be intelligent life.
    2. I do think that non-theistic deities are not very plausible options. I take it most Western philosophers agree, since very few of them believe in them, though there may be cultural explanations of this.
    3. A different approach would be to weaken what is the content the theist argues for, at least in terms of divine goodness. If that’s done, POE is weakened significantly, and FTA is weakened a fair amount as well (but maybe not quite so much as POE). Arguing for an extremely powerful and intelligent person who created the universe is still going to be controversial (most Western philosophers who don’t believe in God also don’t believe in such a being). Once both sides accepted the existence of such a being, then we could have a further discussion of what this being is likely to be like.

    October 8, 2012 — 14:24
  • Mike

    What about other explanatory options that do not leave us in the same position as positing another deity or a super-intelligent contingent being? For example, it seems plausible to me that the past might be infinite. If this is the case, then there may be a physically necessary causal chain leading to the initial conditions. It seems to me like the main thrust of the FTA is *against* the option of random chance rather than in favor of any particular conception, even if you can phrase the FTA in such a way that it does argue for a traditional God. So, what I’ve proposed fits well with that and doesn’t give up any great cost on the side of naturalism whereas conceding something to the POE does seem to have a pretty big cost.
    Having said that, I think you have some arguments against the possibility of an acutally infinite past (Grimm Reaper paradox, if memory serves). If I’m remembering correctly, you’ll probably disagree that the infinite past is a plausible option. I haven’t read that argument, so can’t really comment. We’re also left with whether that option leaves us with a Big Why question, but I think that takes us outside teh context of responding to the main thrust of the FTA.

    October 8, 2012 — 15:10
  • Zia

    Hi Mike,
    “For example, it seems plausible to me that the past might be infinite. If this is the case, then there may be a physically necessary causal chain leading to the initial conditions.”
    I’m not sure how this is supposed to solve the Fine Tuning Problem. Usually, people present the Infinite Multiverse Objection instead. How do we know that this necessary causal chain must give rise to life-permitting initial conditions? You would have to show that such a chain is logically and metaphysically necessary, and that it necessarily gives rise to life permitting cosmic conditions. Otherwise, it just seems to move the problem from the initial conditions to the causal chain so either way the deck seems “stacked.”

    October 9, 2012 — 10:46
  • Dianelos Georgoudis

    I agree that should we start with some set of generally accepted epistemic principles and then directly compare theism to naturalism by the lights of them, we shall find that theism is the significantly more rational view. Naturalism suffers from just too many serious conceptual problems, many of which we have found about thanks to modern science. Incidentally I don’t consider the FTA the best argument of this kind against naturalism. The argument from the deep mathematical nature of the physical universe strikes to me as stronger, since here the multiverse defense doesn’t even go off the ground.
    Still I think naturalists have a universal defeater for any philosophical argument against naturalism, which goes like this: Since what makes an epistemology right is the nature of reality, and since naturalistic reality is not created by some intelligent being, naturalistic reality may well not be rational in the sense of making right the epistemology used in rational investigation. In short, if naturalism is true then it may not be ultimately amenable to rational investigation, and indeed is probably not. But then any argument which appears to be successful against naturalism may very well be begging the question, because it must be using some epistemic principle which on naturalism is not warranted.
    Here are some examples: Suppose an argument against naturalism uses the principle of parsimony. On theism the principle of parsimony is warranted and thus should be used because God’s perfection entails that all other things being equal God will use economy. But on naturalism this does not hold, and reality may well be such that it violates that principle by being far more complicated than it need be. So when one uses the principle of parsimony one may be begging the question, because one is assuming that reality is such that the principle of parsimony is a generally applicable principle – which if naturalism is true may not be the case. Or suppose some argument justifies a premise by pointing out that its negation is extremely implausible. Again, on theism one should assume that God has created our cognitive faculties in such a way that our sense of implausibility is basically reliable. But if naturalism is true then our intuitions about plausibility may be way off the mark. So, again, if one assumes true beliefs about reality would strike us as plausible then one may be begging the question, because naturalistic reality may be such that true beliefs about it will strike us as extremely implausible. Some naturalistic physicists insist that what counts are physical facts, and if physical facts appear to go against, say, classical logic then so much the worse for classical logic (I am thinking here of physicist Lawrence Krauss in his debate with William Craig).
    In general the naturalist may turn Plantinga’s EEAN argument on its head, and argue that our cognitive faculties are unreliable especially when they are used for producing metaphysical beliefs which by their very nature are not amenable to physical verification. Of course the naturalist will recognize that then naturalistic belief is rendered unwarranted also, but this is a much lesser price to pay than recognizing that naturalistic belief is rendered false. Incidentally, my point is not that the naturalist is cleverly embracing irrationality in order to avoid trouble, but rather that the naturalist is being truthful and consistent in pointing out the limits of rationality in a naturalistic reality.

    October 12, 2012 — 3:10
  • Dianelos,
    Although I’m out of my league here, I don’t see how your argument that even if a naturalist concedes to unreliability of his or her faculties yet says the theist has the same problem is persuasive. It hardly seems like the Pyrric Victory you describe it as because the justification for the naturalist to make such a charge was shown faulty. What legs does he or she have to contend the theist’s beliefs metaphysical beliefs aren’t epistemically justified. It seems to me akin to an someone using logic in an attempt to falsify logic. Rather, the naturalist IS being inconsistent and incoherent under your account.
    An example of a naturalist being consistent within his or her epistemic frame is Hume. His empiricism was so restrictive, in order to be consistent, he had to radically deny causation, induction and the self. That’s why he was known as a skeptic. It was later demonstrated his epistemology was so strict, it couldn’t support itself. Hence, the naturalist who subscribes to Hume, (and there are many who do) has no basis to make any attack against theism. So not only are naturalism epistemic tenants a double-edge sword in a Humean sense, resulting in an undesirable skepticism, they are incoherent, rendering the position untenable.
    So, unless I misunderstood your tone and argument as legitimate instead of jest, I don’t think your conclusion holds.

    October 17, 2012 — 22:43
  • John Alexander

    The issue that no one seems to address is that starting with a definition does not get us very far. Definitions do not prove or provide reason to think that what is defined exists. An Evil Demon could very well satisfy many attempts to defend the rationality of believing that God as defined by the theist exists. Just because arguments are consistent with God being omniscient, omnipotent, and completely good, does not mean that they are not also consistent with a being that is omniscient, omnipotent, and competely evil. After all, would not an evil being want to deceive us into thinking it is completely good? It seems that the POE should include the issue of why we should think that what created this world, if there is a creator, is good?

    October 19, 2012 — 10:26
  • 1. I discuss the evil deity hypothesis in section 5.3 here.
    2. The hypothesis of a perfect being is muh simpler than the hypothesis of a completely evil, omniscient and omnipotent being, just as the hypothesis of a perfectly healthy human is simpler than the hypothesis of a human who has perfectly healthy lungs and heart but a perfectly rotten liver.
    3. I doubt that a completely evil omniscient rational agent is possible, given that one always chooses under the guise of an apparent good. Since in the case of an omniscient agent, apparent goods are real goods, an omniscient agent only pursues real goods. Granted, the agent might pursue the goods by evil means. But the fact that it is genuine goods that are being pursued makes it pretty clear that the being isn’t completely evil. (Compare: If a being pursued genuine evils by perfectly good means, we wouldn’t say that the being is completely good.)
    Suppose the guise of the good theory is false. It’s still not clear that a completely evil omniscient rational agent is possible. Presumably, the agent would be one all of whose noninstrumental desires were for evils. But the agent would also know that these desires are desires for ends that are not worth achieving, since the agent would be omniscient. Thus, these desires would not pass the muster of rational reflection. We thus would be faced with an agent whose desires are in no way modified by rational reflection.
    There seem to me to be two possibilities. Either the reason the desires are unmodified by rational reflection because they can’t be modified by rational reflection or they are unmodified by rational reflection because the agent chooses, in each case, not to modify them though it could instead choose to modify them.
    In the first case, if one is so much in the throes of desire that one’s desires are immune to modification by rational reflection, one isn’t a morally responsible agent. But then one isn’t completely evil, because one isn’t an evil morally responsible agent.
    In the second case, we have to suppose the improbable hypothesis that the agent starts off with all of its desires not matching what it knows to be worth pursuing, and add the improbable hypothesis that in no case does the agent choose to modify the desires. So our completely evil deity hypothesis becomes highly improbably.
    Besides, if we are denying a guise of the good theory, we’re likely going to be affirming a slave of the desires theory. But an agent all of whose desires are evil desires never to change any desires in accordance with rational reflection on what is worthwhile. But if the agent is a slave of desires, then it follows that such an agent is incapable of change from the evil desires and so we have the case where there is no moral responsibility.
    Now, these arguments aren’t enough to rule out an omniscient, omnipotent and quite evil being. But such a hypothesis is even less simple.

    October 19, 2012 — 11:24
  • John Alexander

    1. I look forward to reading your paper.
    2. I think that using ‘perfect’ like Anselm and others do begs the question. Why cannot there be a perfect being who is perfectly evil? That being would lack nothing that could make it more evil. This is not self-contradictory. The point that I am making is that if p (the universe exists) is consistent both with q (created by a omniscent. omnipotent, and completely good being) and -q (the universe was created by an omniscent, omnipotent and completely evil being) then the reasonable position to take regarding p is one of agnosticism. A ED could fine tune the universe because finetuning only takes the power and knowledge to do so and says nothing about the moral character of that being just like a watchmaker can make a watch if she has the knowledge and power to do so but says nothing about the moral character of the watchmaker.
    3. The real good for the ED is to deceive. Imagine that the ED sets up the human experience so that people believe that if they love one another and ‘God’, cares for one another, etc. they will be rewarded and go to Heaven. But, when they die and have done the best they can do they go to Hell. If this scenario is possible, and I can find no reason why it is not, then what reason (evidence) does one have to believe that it is not true, even if in fact it is not true? The epistemic problem has always been to get from having a belief that p is true to knowing that what p asserts is the way things actually are. No one would admit to being an ED (or they would fail) so it seems that people set up thought experiments to show that the world is fundamentally decent or good or that we have some way to find the courage to create meaning.

    October 19, 2012 — 16:21
  • Ad 2: Evil is an imperfection (that’s an understatement).
    Ad 3: “The real good for the ED is to deceive.” Vice is always non-instrumentally bad for an agent.

    October 21, 2012 — 8:27
  • Dianelos Georgoudis

    Modus Pownens:
    It’s not so much that the naturalist uses rational arguments to argue for the irrationality of our (and her) cognitive faculties. Rather the naturalist simply observes that on naturalism the reach of our cognitive faculties is limited, and that they can easily mislead us when we apply them beyond their natural field.
    Here, I take it, is how the naturalist thinks. First, consider the following propositions:
    1. The events of the physical universe are causally closed (in the sense that no supernatural causes are observed or required).
    2. There is a perfect correlation between physical events in our brain and our experience of life.
    I posit that the truth of the above propositions is established beyond reasonable doubt by the physical sciences. But if these propositions are true then there exists a (in the logical sense) possible world Wn which is naturalistic and in which all physical and mental facts are identical to our world’s. In particular there exist in Wn intelligent beings possessing exactly the same mental faculties we have, and experiencing life in exactly the same way we do, and who therefore are privy to exactly the same set of data (objective, subjective, intuitive, whatever) we are. (The way Chalmers’ zombie world shares the same physical facts of our world but not the mental facts, Wn shares the same mental facts of our world but not the metaphysical facts.)
    Now consider any argument against naturalism (or for theism) in Wn, such as, for example, the FTA Alex mentions in the OP. If people in Wn find that argument convincing, so will people in our world. But in Wn that argument’s conclusion is false, and the naturalist will first check to see where that argument goes wrong in Wn, and will then argue that for exactly the same reasons it goes wrong in our world. Thus, thanks to the two above propositions warranted by the physical sciences, the naturalist possesses a universal defeater for any argument against naturalism.
    One way an argument can go wrong in Wn is by using an epistemological principle which strikes people as intuitively plausible and which is found to be successful in many contexts, but which is in fact unwarranted as a general principle in Wn. By rejecting that principle the naturalist is not embracing irrationality, but is rather explaining on rational grounds why that principle is unwarranted, what it is in the nature of reality that inclines people to embrace that false principle, and how embracing that false principle while discussing naturalism is to commit the fallacy of begging the question. For, by assuming that this principle is warranted one is in fact assuming that naturalism is false.
    Now if it turns out that by following this path the naturalist will reject many of the epistemological principles we associate with rationality, then the following state of affairs issues: On the one hand the naturalist will not think nor accept that she is rejecting rationality and embracing irrationality. Rather the naturalist will say that she is only accepting the minimal set of epistemological principles that she should, while explaining on naturalistic grounds why the rest she rejects should strike people as being plausible and indeed necessary for rationality. Here the naturalist is simply sticking and being consistent with a possibly true account of reality. – On the other hand though the agnostic may judge that the naturalist’s position is irrational, indeed close to raving mad. So, for example, the many worlds interpretation of quantum mechanics entails that there are parallel universes where each one of us will become the roman Pope, and that in some others each one of us will never die. (That people should believe really strange things just because their metaphysical assumptions appear to require them is a fact of the human nature.)
    So what happens if we turn the table and consider theism under the same light? Being theism a supernaturalistic metaphysics in which our cognitive faculties are basically reliable but also limited it is clear that theism is a possibly true worldview too. Nevertheless the theist cannot as easily, or as justifiably, reject epistemological principles, first because on theism God has created us with basically reliable cognitive faculties including those we use when thinking about God, and secondly because theism arguably implies many of these principles.
    From the theistic point of view I find it significant that God would create the human condition in such a way that naturalism is an intellectually viable understanding of reality. Since I am personally not prepared to abandon epistemic principles such as the principle of simplicity, or of the reliability of my moral sense or of my sense of freedom of will or of my sense of the ridiculously implausible, I find that theism is far more rationally grounded than naturalism. On the other hand I recognize that me not being willing to abandon these rational principles is the result not of some epistemic process but of a willful choice. Thus, I conclude, God has built us in such a way that what we believe about the nature of reality is ultimately an expression of how we apply our will, and not an epistemic contingency. Philosophers perhaps think that their metaphysical views are exclusively informed by a thinking process, but I think that as a matter fact about the human condition this is not so. There is I say such a thing as the will to truth, one’s active choice for truth, truth understood not just as a property of propositions but as a personal, living and fruitful property of reality.

    October 22, 2012 — 3:19
  • John Alexander

    I do not know what happened to my response but…
    Stating that evil is an imperfection and vice is always a non-instrumental bad are not arguments. I problem is that I do not think one can argue for these positions, relative to the possiblity of there being an ED, w/o begging the question.
    Consider the truth-teller and the liar paradox. Regarding some state of affairs that I do not have any independant information on, how can I tell if one is a truth-teller or a liar regarding the information given to me realtive to that state of affairs? Assume that John was absent on Friday and comes to class on Monday and says that he missed class because he was sick. Is he lying or telling the truth? They would both say the same thing. This is Descartes’ point in the 1st Meditation isn’t it – that even if we know that -(p and -p) is true we may not be able to determine if p, or -p, is the case. FTA is either true or false. Assume that it is true. Does this tell us that the designer is good, or evil? We know that if there is a designer that it is either all good or all evil (leaving aside the possibility that it is a mixture of both), but we cannot assert that it is one and not the other. At least that is what I see the problem as. Having introduced the ED, Descartes’ argument that God is not a deceiver because deception is an imperfection is very weak because deception is precisely what the ED is perfect at. It is only an imperfection for those that do not want to be deceived. This is where the question is begged, or so it seems to me. Aquinas’s argment from gradation presumes that evil is the lowest and not the highest. In “Bizzaro” world would not Satan be the ultimate being and God the rebel?

    October 27, 2012 — 11:33
  • It seems to me that the FTA is just as good an argument AGAINST theism as for it. See
    The only way to save the FTA is to include some assumption about God wanting a universe that operates entirely according to fixed physical laws (which means abandoning any argument from miracles).
    Am I missing something here?

    November 10, 2012 — 13:29
  • Robert:
    Relevant to this point is also this very interesting article.
    Back to the point: The FTA needs an assumption that there is a significant value to a universe that with very few exceptions, if any, follows orderly and elegant mathematical laws of nature. Such an assumption is compatible with miracles.
    The details will differ by parameter to be fine tuned, but at least for some of the parameters, the miracles required would probably be more than occasional exceptions.

    November 10, 2012 — 13:53
  • Thanks for your response, Alexander. Unless you exclude miracles entirely, the number of universes compatible with life will still be vastly larger under theism than under atheism, even if we require that exceptions are rare.
    For example, suppose the Earth’s orbit was so close to the Sun that the incident solar energy was too large for life to occur. But life exists nonetheless, because at some distance above the Earth, the excess energy just disappears. Thus, in this universe, energy is conserved everywhere except at a very restricted set of locations around the Earth. (And if we found such a miracle, wouldn’t it be a convincing illustration of God’s power and care for us?)
    Anyway, the requirement that the universe follow orderly laws most but not all of the time seems arbitrary and smacks of special pleading.

    November 17, 2012 — 6:28
  • Leibniz argued against Newton/Clarke that it would be inappropriate for God to rely on on-going miracles for the ordinary operation of the universe.
    One way to defend this is to say that God has reason to avoid miracles. This reason can be overridden, of course. But nonetheless there will be a preference for universes that are life-permitting without the need for miracles, if there are any such. If there weren’t any such, then God would be very likely to go for the miracle-based ones.
    Of course what this ultimately depends on is a judgment about what kind of a world is more valuable. Still, I think there can be objectivity in such judgments.
    “Anyway, the requirement that the universe follow orderly laws most but not all of the time seems arbitrary and smacks of special pleading.”
    Perhaps the value in question is something like that of independence. Suppose I have a dog. Imagine that I have the technology to do everything for the dog: I can hook up a painless IV for nutrition and hydration that will be as good for the dog’s health as actually eating (maybe it’ll be even better for its teeth); I have a special device that I can implant that will put oxygen directly in the dog’s blood, so it doesn’t even have to breathe; I can electrically exercise its muscles while it’s asleep; and I can stimulate its brain so none of this causes the dog any less pleasure than an ordinary dog’s life. But surely this dog is missing out on important aspects of canine life when I do all this for it. There is a value to its doing as much of the ordinary canine stuff as it can. Of course, I should occasionally intervene in its ordinary canine life, taking the dog to the vet when it gets sick, but I shouldn’t do this more than necessary. And I will routinely meet the dog’s relational needs.

    November 17, 2012 — 7:14
  • The miracles in question need not be on-going. For example, it could be the case that the chemical laws do not allow for the chemical evolution of life. God need intervene only once, to get life started. Since there are an infinite number of ways God could do this (from creating a single original cell, to creating a plethora of life forms), there are still vastly more such universes than the naturalistic life-containing ones.

    November 18, 2012 — 7:39
  • 1. Dougherty and Poston in the paper I linked to argue that one can’t consistently run both the FTA and ID-type biological design arguments. Your point is similar, though not the same.
    2. I think both your and their point is somewhat weakened, but not destroyed, on the supposition that God would have good reason to minimize the number of miracles, subject to other desiderata.
    3. But in any case, I think many of the proponents of the FTA would say that a number of the constants going into the FTA are such that not only need the values be “just right” (tough making sense of that rigorously, but there is an intuition there that many theists and non-theists find compelling) for life to begin, but for life to persist.

    November 18, 2012 — 8:37
  • Better link to Dougherty and Poston: here.

    November 19, 2012 — 12:38
  • Leave a Reply to Patrick Cancel reply

    Your email address will not be published. Required fields are marked *