On reductios and the argument from evil
January 4, 2008 — 10:59

Author: Alexander Pruss  Category: Uncategorized  Comments: 15

Bill Vallicella and I have been worrying about variants of the problem how an atheist who does not believe in objective good and evil can formulate an argument from evil against the existence of God (see this post on Vallicella’s blog and this post on Prosblogion). The problem is that the argument from evil requires objective evil. (If we were talking of subjective evil, theodicy becomes trivial–we can say that the evils of this world are evils from our point of view, but not from God’s.)
Here is one way of formulating the problem. On the view that there is no such thing as objective evil, the term “objective evil” becomes nonsense–the view isn’t merely that there happens not to be objective evil, but that the concept makes no more sense than the concept of a perception with no perceiver.
Here I want to raise the problem of auxiliary premises. I will formulate the following in the case of the deductive argument from evil for simplicity, but I think the issues generalize to the inductive case.
The typical theist believes that God exists and that objective evil exists. She is very unlikely to grant that if evil exists, then God exists. Thus, the deductive argument from evil requires some additional premises besides the claim that evil exists. At least one of these premises will involve both “God” and “(objective) evil”, and our atheist takes the second of these terms to be nonsense–and the first, too, most likely.
Perhaps, this will be a premise like: “If God exists, he eliminates all evils he can.” Now this premise will not be a mere tautology. Even if we grant that tautologies using nonsense terms make sense, a controversial question (see this this post and discussion on my blog), so that all borogoves are borogoves, a non-tautologous statement involving nonsense terms, such as that some borogoves are mimsy, seems to be simply nonsense.
So our atheist’s argument from evil involves premises that she believes to be nonsense. This creates, first of all, a formal problem. What does it mean for an argument involving nonsensical sentences to be valid? Perhaps a purely formal solution to this can be given–if it is has the form of a valid argument, it is a valid argument (If all borogoves are mortal, and George is a borogove, then George is mortal). I think we shouldn’t be too sure. The distinction between narrowly logical necessity, on the one hand, and conceptual necessity on the other hand, is not as clear as one might wish.
But I want to raise a second worry. These auxiliary premises cannot be something the atheist coherently believes, since the atheist takes these premises to be nonsense. The argument thus is not just a reductio, but an ad hominem one.
But now we have a problem that I touched on in my last prosblogion post on this topic, though perhaps not as clearly. Suppose the theist simply rejects one of these auxiliary premises. The atheist cannot say that the premise is true. The most she can say is that by the theist’s lights it is true. But to say that the premise is true by the theist’s lights when the theist rejects it seems to involve at least a little chutzpah. Moreover, how can the atheist criticize the theist for denying a premise that by the atheist’s own lights is nonsense? Perhaps the atheist can give a formally valid auxiliary argument from other premises that the theist believes. But similar problems are going to come up, since the conclusion of that auxiliary argument is nonsense, and hence at least one of its premises is nonsense or else its premises are inconsistent.
The basic point here is this: The theist can always get out of the atheist’s argument by rejecting something that the atheist thinks is nonsense or contradictory–and how can the atheist complain about that?
Let’s go back to the concrete issue. Our atheist does not think that God, if he exists, eliminates all (objective) evils he can–that conditional is nonsense by our atheist’s lights. The theist doesn’t believe that either. So where is the conflict? Perhaps the atheist can argue that if x is omnibenevolent, then x eliminates all (objective) evils. But our theist presumably rejects that, just as our atheist does, though for different reasons (the theist because she thinks it’s false, and our atheist because she thinks it’s nonsense). Perhaps the atheist can give a compelling analysis of omnibenevolence. But there is chutzpah in giving an analysis of omnibenevolence that you think is nonsense, and it is hard for the atheist to object to an alternate analysis that yields a different conclusion, since our atheist thinks both analyses are nonsense.
Of course some atheists do believe in objective good and evil, and these issues do not come up.

Comments:
  • Alex,
    We may need to think harder about what we mean by ‘objective evil.’ Our atheist’s denial of objective evil might amount to the claim that all propositions that ascribe good and evil are false. That is, our atheist might be an ‘error theorist’ along lines explained by J. L. Mackie in Ethics: Inventing Eight and Wrong. He might maintain that the very sense of e.g. ‘The Nazi hypothermia experiments were evil’ purports to describe an objective fact, but fails in this purport inasmuch as there are no moral facts. The idea is that moral propositions are truth-valued, but they all have the value false. If our atheist took this line, then there would be no objective evils, but ‘objective evil’ would not be nonsense.

    January 4, 2008 — 20:44
  • Peter Lupu

    Alex,
    I would like to offer a few critical thoughts about your note. Here is what I take to be the gist of your argument in this note:
    1) The atheist is committed to the view that objective evil does not exist.
    2) The claim
    (a) objective evil does not exist.
    entails or is equivalent to the claim:
    (b) the phrase ‘objective evil’ is nonsense; i.e., it is meaningless.
    3) The proponent of the logical argument from evil (LAFE) must introduced into his argument certain auxiliary premises in which the phrase ‘objective evil’ is used (not merely mentioned).
    4) But from an atheist point of view, any sentence in which the phrase ‘objective evil’ is used, not just mentioned, is not a well-formed sentence because it contains essentially a nonsensical phrase.
    Therefore,
    5) From the atheist point of view, any such sentence cannot be a premise, auxiliary or otherwise, of an argument.
    Therefore,
    6) An atheist cannot be a proponent of LAFE.
    7) Since a theist will certainly not be a willing proponent of LAFE, it follows that no one can be a proponent of LAFE.
    Here is my brief reaction to this argument:
    (I)Premise (1) is highly controversial; still it is something I am willing to accept, for the sake of the argument.
    (II) Premise (2)is simply false and you do not offer a convincing case to support it. I reject it tout court.
    (III) I accept premise (3).
    (IV) I reject step (4), (5), and 6) because they all rely on premise (2).
    (IV) Therefore, i do not think that (7) is true either.
    So our disagreement focuses upon premise (2). Here are some very strong considerations against accepting premise (2):
    1) you claim that “On the view that there is no such thing as objective evil, the term “objective evil” becomes nonsense”.
    I do not see why one should accept this and premise (2) which is derived from it. Why?
    Well, for the following reasons.
    First, i suppose that the term ‘objective evil’ is a general term, like a predicate. Predicates can be quite meaningful even if their extension is empty. So (a) can be simply taken to mean that the predicate ‘x…is a case of objective evil’ has no instances in the world. Why should one conclude from this that the term is meaningless or non-sense (=lacks a sense in any language?)?
    Second, suppose we discover that there are no human beings that are over ten feet tall. Does it follow that the term “is a human being over ten feet tall” is non-sense? is meaningless?
    I don’t think so. The same holds for ‘objective evil’, provided it is taken to be a general term.
    Third, perhaps your point is rather this:
    An opponent of objective evil does not merely deny that objective evil exists; such a philosopher maintains that objective evil could not exist. i.e., that there could not be instances of the term ‘objective evil’.
    But even under this much stronger claim, it does not follow that the term ‘objective evil’ is nonsense merely from the fact that there could not possibly be instances of objective evil. Why?
    Consider the following term: ‘is a positive whole even number smaller than 2’. Now, there is nothing that satisfies this predicate and, moreover, if mathematical truths are necessary, then there could not be any positive natural number satisfying this predicate. Does it follow that the phrase ‘is a positive whole even number smaller than 2’ is nonsense? I do not think so.
    Incidentally, i do not see the analogy between this case and the case of perception without a perceiver. But that is another matter.
    peter

    January 4, 2008 — 21:17
  • Alex,
    I’m having a very difficult time understanding what you mean by ‘objective evil’ and (I’m assuming) by ‘objective value’. It’s just not clear what conditions have to be met for the property “x is evil” to be objective. Here are some possibilities:
    x is evil is objective,
    1. if “E is evil” takes a truth-value.
    2. only if “x is evil” is a nonnatural property
    3. if “x is evil” supervenes on some set of natural properties.
    4. only if “x is evil” is not reducible to any natural property or set of natural properties.
    5. only if “x is evil” supervenes on a set of properties that includes being commanded (wished, willed, etc.) by God.
    Are any of these in the neighborhood of what you mean?

    January 4, 2008 — 21:21
  • Peter Lupu

    Bill,
    I have been trying to comment on your note in your own site “Maverick Philosopher”, but was unable to do so. Perhaps, I need to wait for approval first.
    In any case, the problem with your and Alex’s view is much deeper than the question what is meant by ‘objective evil’. The problem is general regarding the following claim you both make. Let ‘x is an F’ be any arbitrary predicative term. Here is what you guys say:
    (a) There are no instances of F.
    entails or is equivalent to
    (b) ‘x is an F’ is meaningless or nonsense.
    No matter what one means by ‘x is an F’, i do not see that you offered any supporting argument for the claim that (a) entails (b) or is equivalent to it in any way. Moreover, there are plenty of considerations against the entailment/equivalence claim, some of which I have listed in one of my posts on this site.
    Unless you make a convincing case for this claim, I do not see how your argument holds.
    peter

    January 5, 2008 — 9:34
  • On the view that there is no such thing as objective evil, the term “objective evil” becomes nonsense–the view isn’t merely that there happens not to be objective evil, but that the concept makes no more sense than the concept of a perception with no perceiver.
    Peter,
    I took this passage initially to be proposing the inference you mention. But I’m guessing Alex cannot have in mind (at least not as an inference) that “x is evil” is uninstantiated, therefore the property is incoherent. I’m guessing that it is rather the converse that he has in mind: the property is incoherent, and therefore has no genuine instances. Otherwise he is burdened with such easy counterexamples as the fact that “x is a twenty-fingered human” is uninstantiated (let’s say… make it 35 if necessary), therefore the concept of a twenty-fingered human is incoherent. But, obviously, that is not incoherent. Worse still, take the property of being the first person born in 2009. Again, no doubt it is uninstantiated and no doubt it is coherent.

    January 5, 2008 — 10:13
  • Thanks for all the comments.
    I am not claiming that all atheists are committed to the claim that there is no such thing as objective evil. But some do have that commitment–particularly ones with scientistic tendencies.
    By “there is no such thing as objective evil”, I do not simply mean the view that there happen not to be any objective evils. I mean the view that the “concept” of objective evil is a pseudo-concept and does not express a property.
    What do I mean by objective evil? Well, I don’t have a theory of objectivity. But any view on which the truth value of “E is an evil” always depends on the interests, views or attitudes of the speaker or her community will not qualify.
    Consider three views:
    1. There in fact are no objective evils.
    2. There can be no objective evils.
    3. The phrase “objective evil” has no meaning.
    I think views 1 and 2 are rationally unstable for atheists, and push one to view 3 or else to the view that there are objective evils. First, take view 1. Given the apparently excellent evidence for the existence of objective evils–massacres, tortures of the innocent, rapes, etc.–I think the three main ways to have some justification in believing (1) is (a) to have some conceptual argument against the possibility of evils, or (b) to hold that evil is to be in some way analyzed in terms of the divine will, e.g., as events contrary to God’s antecedent will, or (c) to have some kind of a supernaturalistic view on which the apparent evils of the world are merely apparent.
    I don’t have an argument that (a), (b) and (c) are the only options, but I think they are the main ones. I simply reject (b) for the same reasons that I reject divine command theories. There may be some atheists who believe (b) and thus doubt the existence of objective evil, but it is not a very philosophically attractive view, I think. For one, most divine command theorists only analyze wrong, not evil, in terms of the divine will. Moreover, the Western atheist (I do not here include adherents of atheistic religions like some varieties of Buddhism) is not going hold to (c). That leaves (a), but that not only is going to show that there in fact are no objective evils, but that there can’t be any.
    Can view 2 be stable and avoid falling into view 3? I think not. If one rejects the possibility of “objective evil”, it is just not clear how that phrase can have meaning. We do not have a definition of evil, except perhaps as the opposite of good, but the same issues apply to “good”, and I assume that the atheist who thinks there can’t be objective evils also thinks there can’t be objective goods. We learn what the words “good” and “evil” mean by pointing to paradigm cases. But by (2), none of the paradigm cases are cases of objective goods or evils. So I don’t see how by pointing to the paradigm cases we could have ever acquired a meaning for the words “good” and “evil”. Suppose you learned what “mimsy” was by having paradigm cases pointed out to you: the Battle of Waterloo, the Empire State Building, the Coronation of Elizabeth II, etc. You might after a while find yourself able to use the word “mimsy”, and feel that the things it applies to have a certain specific kind of grandeur. However, supposing you then learned that none of these cases were in fact cases of mimsiness, and that in fact there can be no mimsy things or events, then I think it would become clear that “mimsy” lacks meaning.
    There is, however, one way of staying out of this. One might think “evil” has a meaning and then “objective evil” gets a meaning by compositionality, albeit it is a self-contradictory meaning, like “square circle”.
    However, I don’t think this will help the atheist. When people talk of “objective evil”, I don’t think this is something one gets by compositionality from “objective” and a non-objective “evil”. They mean something else. The same goes for “objective beauty.”
    Besides, if “objective evil” is a simple self-contradiction, then claims like “God prevents all evils he can” are weird. Does God prevent all square circles he can?

    January 5, 2008 — 10:50
  • JA

    Why cannot the atheist’s argument simply be an ‘if-then’ argument without the atheist having to assert that any of the premises are in fact true? If the atheist could show that there is an inconsistency between the premises then it is the case that at least one of the premises is in fact false. I take it that is what Mackie does in his paper and what Plantinga counters in his discussion of Mackie. It seems to me that questions of consistency have little to do with the actual truth-value of what is being claimed.
    Furthermore, are you suggesting that an atheist must believe, on the basis of a moral theory, that evil is not objective or that there are in fact no atheists who think that evil is objective? If this is so, then I suggest that the latter is an empirical claim and needs some verification. Are there no atheists who in fact believe that evil is objective and who then use this belief to argue that there is no God (theistically defined)? If you are suggesting the former then I think some argument is needed to demonstrate that atheism is theoretically committed to a non-objective thesis regarding good and evil. Maybe you have done this and I missed
    it.

    January 5, 2008 — 11:23
  • Re rejecting something that your opponent finds nonsensical, your opponent has reasons (that she thinks you should accept, if you are rational) why you should not reject it, where that “it” refers to the hypothetical referent of the term that your opponent finds nonsensical. So your opponent may simply suggest that the logical analysis of hypothetico-deductive reasoning (e.g. questions about the ‘nature’ of impossible worlds) is currently in its infancy, is in no position to judge its application, whether in reductios or science generally, or indeed ordinary reasoning; that having an intuitive grasp of how it works (and how it can work in this case) is simply part of being rational.

    January 5, 2008 — 11:53
  • Enigman

    Hi Alex, I just thought I’d add that self-contradictions are quite different to nonsense; e.g. if the Liar sentences are nonsensical then they do not contradict themselves, whence the paradox is resolved.

    January 5, 2008 — 11:58
  • Peter Lupu

    Bill,
    I have been trying to comment on your note in your own site “Maverick Philosopher”, but was unable to do so. Perhaps, I need to wait for approval first.
    In any case, the problem with your and Alex’s view is much deeper than the question what is meant by ‘objective evil’. The problem is general regarding the following claim you both make. Let ‘x is an F’ be any arbitrary predicative term. Here is what you guys say:
    (a) There are no instances of F.
    entails or is equivalent to
    (b) ‘x is an F’ is meaningless or nonsense.
    No matter what one means by ‘x is an F’, i do not see that you offered any supporting argument for the claim that (a) entails (b) or is equivalent to it in any way. Moreover, there are plenty of considerations against the entailment/equivalence claim, some of which I have listed in one of my posts on this site.
    Unless you make a convincing case for this claim, I do not see how your argument holds.
    peter

    January 5, 2008 — 14:00
  • Enigman

    Sorry if that seemed confused; maybe an example will clarify my point.
    Suppose X says “deserts might be slithy,” and Y (who believes that “slithy” is nonsense) says that that’s clearly false and points to all the places where X has explained the meaning of “slithy” in terms of it being such a property as otters, for X’s main example, might have (with the clear implication being, in Y’s opinion, that the property is akin to lithe and shiny, and very possibly sleek and slimy, et cetera).
    It would seem dialogically incorrect for X to say that Y is not entitled to such an opinion, of her statement’s falsity, because X believes that “slithy” is nonsense. It would seem to be incorrect for X to demand an account of how reductios work, rather than Y’s pointing to X’s previous statements. But having said that, I do suppose that most atheists who say that “objective evil” is nonsense actually mean that it is no more possible than an imperceptible sense-datum.

    January 5, 2008 — 14:04
  • Enigman

    (woops, a bit of a mechanical stutter there)

    January 5, 2008 — 14:08
  • What do I mean by objective evil? Well, I don’t have a theory of objectivity. But any view on which the truth value of “E is an evil” always depends on the interests, views or attitudes of the speaker or her community will not qualify.
    I’m not sure why you say this. It looks like you’re saying that the existence of objective evil must be independent of contingent facts about human beings such as (some of) their interests, desires or preferences. Is it the contingency that worries you? But then wouldn’t you want to say that some objective evils are dependent on the existence of conscious beings and conscious states? Certainly, on any view of objective evil, Smith’s excruciating pain has to count as objectively bad. If Smith were not conscious, there would be no such objective evil. But the existence of such conscious states is itself a contingent matter of fact. So it cannot be the contingency.
    Is your worry rather that objective evil cannot be dependent on subjective features of human beings (and other sentient beings)? But here again subjective experiences of extreme pain have to count as objectively evil. So it cannot be the subjective features that worry you.
    But then the worry seems to be neither the dependence of objective evil on contingent facts or subjective states. But then, what?

    January 5, 2008 — 14:32
  • Peter Lupu

    Mike,Alex, Bill,
    I gather that Alex’s last post confirms Mike’s last comments. But, now, here are some thoughts about this last version of Alex’s thesis.
    Alex,
    You begin the relevant portion of your last post as follows:
    “By “there is no such thing as objective evil”, I do not simply mean the view that there happen not to be any objective evils. I mean the view that the “concept” of objective evil is a pseudo-concept and does not express a property.”
    While I sense here a shift at least in the formulation of the original thesis, let us examine this formulation.
    1)Are the following two phrases synonymous, mean the same thing:
    (i)there is no such thing as objective evil;
    (ii)the “concept” of objective evil is a pseudo-concept and does not express a property.
    I suppose that (ii) entails (i) [I will grant that for now]. But does (i) entail (ii)? I do not think so for the reasons I and Mike in his last post have indicated.
    Therefore, (i) and (ii) do not normally mean the same thing. Hence, if an atheist asserts (i), it would not be very charitable to interpret him/her as meaning (ii).
    2) Perhaps you simply meant to say that for the sake of stating your thesis you stipulate that (i) shall mean (ii). Thus, your argument then covers two types of atheists:
    (1) Those who assert (ii) directly;
    (2) Those who assert (i) but really mean (ii).
    So let us simply focus on (ii) and forget about (i), for the later is not a view you are really discussing (or at least should not be).
    3) Here is how I interpret (ii):
    The form of words ‘objective evil’ does not express a real *concept* or a real *property*; while it is used as if it does express a property or concept, but it fails to do so. So such uses are really pseudo-uses. Therefore, all contexts in which such a phrase occurs essentially only appear to have a truth value, but really do not.
    Of course, if someone holds this view, then ipso facto they also hold that the phrase ‘objective evil’ cannot have an extension,[i.e., they hold (i)], where the ‘cannot’ here is some sort of conceptual necessity.
    But, now, if this is your original claim, then there is no need to go through the argument you have just given in your last post. Here is what you say:
    “Consider three views:
    1. There in fact are no objective evils.
    2. There can be no objective evils.
    3. The phrase “objective evil” has no meaning.
    I think views 1 and 2 are rationally unstable for atheists, and push one to view 3 or else to the view that there are objective evils.”
    But if you really mean by here what i have stated above, then the views listed as 1 and 2 here are not merely “rationally unstable for atheists” and “push one to view 3”; 1 and 2 are equivalent to 3, for according to you 1 and 2 must be interpreted as meaning that the form of words ‘objective evil’ do not express a concept or a property; but since they are the sort of words that if they express anything at all, they express one or the other, it follows that they are really meaningless. Hence, if you really intended your position to apply to atheists who hold (a) in the sense you now give it; namely, that the form of words ‘objective evil’ does not express a concept or property, then there is no need for you to prove that this position pushes one into the view that this form of words is meaningless. This consequence follows immediately simply by virtue of the claim that these words fail to express a genuine concept or property.
    4) I think that you really are of two minds how to state your position. On the one hand, you wish the argument to be substantive, so you sometimes state it in the form that (a) implies (b). And this is indeed a non-trivial thesis, except that it is vulnerable to several pretty critical counterexamples. On the other hand, you sometimes resort to the later formulation which is not vulnerable to the counterexamples in question; but then the position is not as substantive as it appears initially.
    peter

    January 5, 2008 — 15:14
  • Peter Lupu

    All,
    i am sorry about the *echo* in my last posting; it appears not to post it because of an error message; then i resubmit it only to find out that it posted it the first time i submitted it.
    sorry
    peter

    January 5, 2008 — 15:23