In chapter 6 of his Philosophical Theology (1969), James F. Ross undertakes the very ambitious task of showing that the evil in the world does not provide even a prima facie case against divine moral perfection. Ross takes the phrase 'a prima facie case' in the legal sense: to provide a prima facie case is essentially to bring charges that need answering. So, for instance, someone who says that the evils in the world are justified by some greater good which would be impossible without them is conceding that there is a prima facie case and attempting to answer it. Ross believes that there is no such case that needs answering. After explaining his argument, I will show that, even if Ross's answer to the alleged conflict between the evils of the world and divine moral perfection succeeds, the evils of the world can still be used to make a prima facie case against divine benevolence, and Ross's strategy cannot be used to defuse this.
Ross's case hinges crucially on the dialectical situation which obtains between the theist and the atheist. We are to suppose that the atheist is aiming to establish the non-existence of God. To do this, she must use premises which are known to be true, and her procedure must not exhibit epistemic circularity - that is, she can't appeal to premises which she (allegedly) knows only because she (allegedly) knows that God does not exist, or that if God existed he would be morally responsible for the evils in the world, or some other similarly contested claim. Given that this is the case, the atheist's argument needs to appeal only to general moral principles that can be seen to hold in our everyday experience. She cannot appeal to principles that are specifically about God (unless, of course, there is some reason why the theist must accept those principles).
Now this, according to Ross, puts the atheist in a tricky position, because God's relationship to the happenings in the world is quite unique. In ordinary cases, when, say, Smith murders Jones, it is because of Smith's role in the natural causal chain leading to Jones's death that secures Smith's moral responsibility. But God, according to the broadly Thomistic metaphysics Ross endorses, is not a cause among other causes in the way Smith is. Rather, God atemporally causally sustains the world as a whole. God is not an actor in the order of natural causes.
Fleshing out this Thomistic line of thought, Ross says that God is at a higher level of reality than natural objects, and the cosmos is metaphysically dependent on God. This sort of dependence is what God's 'causal' role in the history of the world amounts to, and this is a very different sort of thing than ordinary natural causation. As a result, to formulate some general principle like 'it is prima facie wrong knowingly and intentionally to cause the death of a sentient being' and then apply this to God is to equivocate on the word 'cause.' The principle that we know is a principle about natural causation, not the sort of metaphysical conservation that God is engaged in.
So suppose the atheist tries to formulate some principle that does apply across reality levels. According to Ross, the only case of this that we have a good grasp of is the relationship of human persons to fictional and imaginary objects. This is a case where an agent freely and intentionally brings about beings at a lower level of reality. It is also similar to the creation in that the author is not one of the causes within the fiction, although the author is the cause of the fiction as a whole, and the cause of each individual event in it. (We can distinguish, perhaps, between the 'fictional' cause of each event and the 'authorial' cause of each event.)
Now, Ross is quite clear that he does not mean to argue by analogy between divine and literary creation. The claim he is making instead is that the only kind of general, known, non-question begging principles the atheist could appeal to would be principles broad enough to apply to some ordinary, concrete cases we know about, and be supported by these instances. So, if the relation of authors to fictions (and imaginers to imaginings) are the only cases of this sort that we have a good grip on, then the atheist will have to formulate some general principle known to be true in the case of fictions. The atheist, that is, will be forced by the dialectical situation to argue by means of this analogy.
However, the relevant principles, as applied to fictions, are false. Macbeth is morally responsible for the murder of Duncan, and Shakespeare is not. Of course, Shakespeare is responsible for the fact that, in the fiction, Macbeth murdered Duncan, but, although Macbeth is morally criticizeable for this, Shakespeare is not. Likewise, authors are not blameworthy for writing natural disasters and other such things into fictions. So the atheist's argument from analogy fails, and no prima facie case against divine moral perfection has been made.
The atheist may, of course, say that the cases are quite different because the fictional characters are, after all, fictional; they don't really feel pain, for instance. So the analogy breaks down. But Ross, recall, insists that he is not arguing by analogy; instead, he is claiming that the atheist is forced by the dialectical situation to argue by means of this kind of analogy. To say that the analogy fails, according to Ross, is really to say that God's relation to his creation is sui generis to such an extent that we don't have an independent grasp of the moral and other issues involved in that relation.
Now, it seems questionable to me whether the atheist can really be forced to make this kind of argument by analogy, and I also don't think that Ross has adequately answered the objection that, on his view, we can't take any events in the world as demonstrations of the goodness of God. But even leaving this out, there is another problem. The atheist is in a position to use this kind of analogy to make a prima facie case against divine benevolence.
Divine benevolence is the doctrine that God wills that each of his creatures be as well-off as possible. This is, of course, a claim about what Leibniz and others have called God's antecedent will - that is, there is a ceteris paribus condition here. The well-being of a particular creature may, in a particular case, be overridden by some other consideration. So we could state this as the thesis that God values the well-being of creatures, or that the well-being of creatures is among God's ends in his volitional activity.
Now consider this argument:
Premise 1 is extremely plausible as applied to authors of fictions. One is inclined to think that if one of Shakespeare's aims was that his characters should be as well off as possible, he wouldn't have written tragedies. And, of course, the theist must accept premise 2 and 3.
Of course, this provides only a prima facie case. One can imagine a tragedian being extremely troubled by the suffering of his characters, and truly aiming to mitigate it, but feeling that he had some strong overriding reason for writing tragedies, rather than fictions of some other sort, and needing to abide by the conventions of that literary form. But the point is, there is a charge that needs answering here. Someone who claims that Shakespeare aims that his characters should be as well off as possible has a lot of explaining to do. So Ross's argument, even if sound, does not render the project of theodicy unnecessary.
(cross-posted at The Prosblogion)
Posted by Kenny at May 5, 2012 11:41 AM| Trackbacks |
TrackBack URL for this entry: https://blog.kennypearce.net/admin/mt-tb.cgi/680
|
I don't think benevolence as defined here is a plausible divine attribute, but supposing it were, I'm not sure why you think (1) is plausible as applied to authors of fictions, because it seems to me extremely implausible. And I don't think it actually fits authorial experience; if I read her correctly, for instance, Dorothy Sayers outright denies this in *Mind of the Maker*. What kind of argument in its favor do you have in mind?
Posted by: Brandon at May 5, 2012 1:52 PMI was thinking of this as a very general, and fairly weak, principle about agents and their ends: if an agent claims that E is among her ends, but neglects opportunities to pursue E, then some kind of explanation is in order. Of course, there are plenty of cases where such explanations can be provided: the particular opportunity was too costly, interfered with some other, more important end, etc. Authors do sometimes say that they wanted the story to go a certain way, but just couldn't get it to go that way. However, there is not actually some kind of inability here. The author knows perfectly well how to write the sentences in question, saying that the outcome was such and such. What is actually going on is some kind of incompatibility between the desired outcome for the characters and the author's aesthetic aims (the author couldn't get it to go that way and still be a good story), or the integrity of the characters (the author couldn't get it to be plausible that this character would act in that way) or something. These are the sorts of explanations that would serve to answer the prima facie case, but some kind of answer is still required. If you have an opportunity to pursue E (where opportunity includes both having the relevant knowledge and the relevant power) and the opportunity costs you nothing and doesn't conflict with any of your other aims and nevertheless you don't take it, we are typically entitled to conclude that E is not very important to you (or, at least, wasn't very important to you at the moment when you made that decision).
Posted by: Kenny at May 5, 2012 2:11 PMInteresting. I don't think it is in fact true that if any agent claims E among her ends but neglects opportunities to pursue E that some kind of explanation is in order. What requires explanation are things like neglecting necessary or the only or the best opportunities, where knowledge of such things is available -- and that, in fact, you need to have some such knowledge in order to make the principle reasonable. This is one of the reason why the principle rings so false in the authorial case: authors are concerned not merely with achieving E, whatever it may be, but with selecting the best opportunities for achieving E. I think it's much too strong to characterize this as an incompatibility between desired outcome and aesthetic aim or integrity of character -- no accurate notion of desirable outcomes for a character can be formulated independently of either of these, and if it ever actually becomes an incompatibility, it's bad characterization and plotting.
Likewise, I don't think the principle in your last sentence is plausible; it has the odd consequence that any E that is not urgent -- i.e., the next opportunity definitely needs to be taken or E can't be achieved -- is prima facie not important. But urgency and importance often come apart in means-end reasoning, and I think you're trying to throw them together in a way that is often not tenable. Lots of E's don't need to be achieved at the next opportunity, regardless of their importance. And we know this, and act on it all the time: we can easily distinguish in practical life between our most urgent ends and our most important ends. Likewise, it's entirely possible for us to have ends that are neither urgent or important (all things considered), but nonetheless are stable components of our goal, so this second principle narrows (1) considerably (although given the definition of benevolence here, probably not in a way that makes much difference to the argument).
I agree that the principles could become plausible if we knew certain things -- e.g., the whole space of opportunities, or if E is of a very specific kind. But I think the exceptions are far too many to put much weight even prima facie on the principles without also postulating that we know things of this sort. (And such postulations I think are often going to fail in the authorial case.)
Posted by: Brandon at May 5, 2012 6:47 PMSo, first, you may be right that some kind of qualification is needed; I'm not sure. But I would definitely say that if the agent thought there was going to be an opportunity to achieve E in a better way in the future, that would be the sort of thing that could explain the agent's not taking the means to E in a particular case. The claim that it costs literally nothing is doing a lot of work in my original formulation, of course. If something is an important but non-urgent end, and so you neglect the means for it now so as to pursue some less important but more urgent end, that doesn't violate the principle, because in this case taking the means to E would cost something.
The sort of case I have in mind is this. People sometimes make the following criticism of Kant's ethics. According to Kant's ethics, no positive duty is a perfect duty, so no one is ever blameworthy for failing to take some course of action. In some cases this is extremely implausible. For instance, if you are walking by a drowning child, and you could reach in and pull the child out with minimal effort without even being late to wherever you are going (which isn't that important anyway), and you refrain from saving the child, you are clearly blameworthy.
To this objection, I answer: Kant's ethics requires that you adopt the happiness of others as an end. But if someone refrains from saving the child in such a case then we have extremely strong evidence that that person is not at all serious about pursing/valuing the happiness of others. So the Kantian does have grounds for moral blame against the person who refrains from saving the drowning child.
Even if there are defects in my original formulation, I think it is quite clear that some principle in the neighborhood is true and will do the work in the argument.
Posted by: Kenny at May 5, 2012 7:01 PMI've realized that I didn't actually write the 'at no cost' part, which I just said was important, in my original formulation. Oops. So the principle definitely needs to be qualified in some way.
Posted by: Kenny at May 5, 2012 7:37 PMOK, that makes a bit of sense. As I see it, the real difficulty is that (1) (or its appropriate counterpart) must meet a fairly high standard: it has to apply, at least plausibly, to authorial cases, it has to be something at least plausibly transferable to the providential analogue, and it has to make the argument go through (i.e., the way in which it is author-plausible and providence-transferable has to support the conclusion). But the authorial case is a complicated case to begin with. For instance, I think the 'at no cost' qualification, while it makes the principle plausible, is actually a very strong qualification, and I worry about how often we are in a position to think that it is plausibly met in the authorial case -- authors, even pretty straightlaced ones, select from a pretty hefty space of opportunities, and ingenious ones from a mind-bogglingly large space of opportunities. It seems to set the epistemic bar too high for using the principle in the authorial case. But at the same time you're right that "at no cost" ends up doing a fair amount of work in making the principle plausible.
Posted by: Brandon at May 5, 2012 8:04 PM