Saturday, March 26, 2011

False Senses of God's Goodness

What does God's omnibenevolence consist in? I think that we need to do a little more work on fully explaining what we mean by this. In fact I'm somewhat distraught, because it seems that if we affirm God's all-goodness we should already know what we're talking about. I think that some of the notions which are typically bound up with omnibenevolence really don't make sense.

In a post a while back I presented the Prussian Free Will Defense. I think the argument is sound. The conclusion that we end up drawing from it is the following. It will be important in determining what omnibenevolence can be and what it cannot be:

(C1) It is not the case that there is some moral principle in the nature of an omnibenevolent and omnipotent being which prevents him from creating a creature who does something immoral.

"Actualizing all possible goods":
I think it's relatively easy to understand omnipotence and omniscience. Omnipotence is just the ability to do all that is possible. Omniscience is just the knowing of all things. So does omnibenevolence consist in the actualizing of all good things?

I'm convinced that this notion is incoherent. (C1) shows that God's omnibenevolence is compatible with God's not actualizing worlds which he could have actualized in which all moral agents always freely choose to do the right thing. And that is certainly incompatible with God's omnibenevolence meaning that he actualizes all possible goods.

The idea that omnibenevolence means "actualizing every possible good" is probably in contradiction with the idea of omnipotence anyhow. One could think of two goods which are mutually exclusive but are both within the range of God's power. For example, my marrying Lucy and my marrying Edith are both goods which are mutually exclusive and possible for God to actualize. However, God can only bring about one of these. Hence, if God is omnipotent, he can't be omnibenevolent if that means that he actualizes every possible good; there are some goods which God could actualize but doesn't.

"Preventing all unnecessary evil":
It seems that the existence of an all-good being must preclude the existence of unnecessary evil. Responses to the problem of evil like Plantinga's Free Will Defense say that some evils which take place are justified by greater goods. But the atheist objects that there are some evils which are probably not justified by greater goods, and in general, the theist agrees that if this were true then an omnibenevolent God would not exist. The typical attempt to solve the problem is to claim that no evils are unnecessary.

But what if the theist were to take a different route and say that even if there were evils which could have been prevented that this doesn't rule out the existence of an all-good God? In fact, I think (C1) shows that this is true. For take again the possible world W1 in which all agents always freely choose to do what is good. (C1) shows that there is no moral principle in the nature of an omnibenevolent being which prevents him from actualizing a world W2 where an evil is committed even though he could have actualized W1 instead. But in that case, some of the evils which take place are unnecessary; they could have been prevented by God's actualizing W1.

A Problem:
So we see that, based on (C1), omnibenevolence cannot imply the actualizing of all possible goods. Neither does it imply the preventing of all unnecessary evils. Yet this seems to sap omnibenevolence of any real content. How could we call something "all-good" which allows needless suffering and withholds a great many good things from us?

To me this is a big problem. I want to find some possible ways in which we can understand omnibenevolence in light of these objections. Any thoughts?


Brandon said...

As far as I can tell people only began talking about divine omnibenevolence in the nineteenth century in the context of the problem of evil: the whole point of the term was to make a nice parallel with "omnipotence" and "omniscience" (also, simply saying that God is good and benevolent, or even the best good or good itself, which is pretty much what people have traditionally stuck to, doesn't seem to be strong enough to generate any of the contradictions for the logical problem of evil which dominated for so long). So the problem really is that it's a completely foreign term, introduced in order to create a refutation of a position that didn't historically use it prior to its becoming widespread in the attempted refutation. Prior to that, the closest people usually came to calling God omnibenevolent was to say that he was infinitely good. When they said, on rare occasions, that he was all-good, they typically meant that all good things come from His goodness.

awatkins909 said...

That's pretty interesting. That would seem to explain the perspicuous absence of any substantial logical problem of evil within the medieval tradition (at least from what I've read). Still, isn't a being A which is omnibenevolent greater than a being B which is less than omnibenevolent, and wouldn't this mean that if there is an Anselmian greatest conceivable being then it must be A?

Brandon said...

To the extent the medievals recognized a logical problem of evil, it was based on infinite good, e.g., the one Aquinas gives:

If one of two contraries were infinite, the other would be altogether destroyed. But the word "God" means that He is infinite goodness. If, therefore, God existed, there would be no evil discoverable; but there is evil in the world. Therefore God does not exist.

But it's easily answered by pointing out that you could just as easily argue the other way: since God is infinite goodness, if God is known on other grounds to exist, and evil exists, it must be the case that evil is not inconsistent with infinite goodness.

On your question, it would depend, of course, and as you worry in the post, on what is meant by omnibenevolence. I don't think it's usually coherent: it's intended to convey something that monotheists accept generally, and that creates a logical problem when combined with omnipotence and omniscience, but what notion of omnibenevolence can do both is hard to say.

Leo said...


Thanks for the information. I have long suspected that "omnibenevolence" enters into philosophy of religion discussions insofar only as it is needed to make the problem of evil problematic. I don't think I have ever seen it argued for independently. And now my suspicions are confirmed!

(What's even more unfortunate is that omnibenevolence is hardly unique in being unargued for independently: I rarely hear any discussion in contemporary philosophy of religion of what good reasons we might have for thinking God omniscient, omnipotent, etc. And no, "x is a great-making property" is not a good reason for thinking anything.)

BTW, I think that James Chastek has a recent post up on this subject. Check it out.

awatkins909 said...

I talked to James Chastek about it a little bit, and he confirmed my suspicions. I'm not entirely satisfied with what he says we can understand God's benevolence to be.

Also, why not the x is a great-making property, Leo? Isn't it essential to some ontological arguments like Godel's for instance (positive properties)? I suspect you might be right though.

awatkins909 said...

It seems maybe that I was somewhat wrong in my post. Is gratuitous evil the same as unnecessary evil? Gratuitous evil is evil which has no ultimate purpose whatsoever. Unnecessary evil is evil which could have been prevented. I think what we should conclude from the argument is that God's goodness must be compatible with unnecessary evil. This doesn't say anything about gratuitous evil one way or the other. Is that correct?

Leo said...

Re: great-making properties. My difficulties are twofold. First, recourse thereto looks suspiciously intellectually lazy: whereas an Anselm, Aquinas, or Scotus will spend pages arguing for a single divine attribute, a perfect-being theologian achieves his task with the simple syllogism "All GMPs are Divine properties; A is a GMP; therefore, etc." Now, perhaps the neo-Anselmian really has proven Aquinas' efforts, say, in arguing for Divine omniscience to be a needless waste of time, but I somehow doubt it.

Second, the notion of a great-making property is hopelessly intuitive and subjective. My list, for example, of great-making properties would include simplicity, timeless eternity, and absolute sovereignty; many a neo-Anselmian, however, would counter that these attributes conflict with smoe other "great-making properties," such as personality, omniscience, and moral goodness, and so long as we retain the language of great-making properties, we have little by way of a means to rationally dispute the matter.

It is in this that GMPs differ from Thomistic pure perfections and Gödelian positive properties, which both contain precise criteria for the admission of a given attribute into their ranks (pure attribution and lack of potency, respectively).

Leo said...
This comment has been removed by the author.
Leo said...

Possible answers to your original question:

A.) Anything A is omnibenevolent iff any x is good only if A wills that it be good.
B.) God is the whole and entire good; God wills Himself; therefore, God wills the whole and entire good, i.e. is omnibenevolent.
C.) Dispensing with the "omni" altogether, we might say that God is perfectly or supremely benevolent because He wills the best of all possible worlds (which of course requires that God actually does will said world).

awatkins909 said...


Isn't (A) just the usual divine command theory one? That doesn't seem adequate does it. Also, in my post didn't it show that God's goodness is compatible with his not actualizing the best of all possible worlds, meaning (C) must be wrong (the coherence of "best of all possible worlds" is dubitable anyway)?

(B) is probably true but what does it mean to say God is the whole and entire good? Maybe that's just basic though.

Leo said...

I don't think you understand my point in (A): I am not saying that an action-type t is a good action-type only if God will that t be a good action-type; rather, I am saying that saying that any individual creature is good only on the condition that God wills it to be so. Mary, for example, is pious and wise only because God wills that she be pious and wise. This thesis is really just an extension of universal Divine sovereignty.

I do not see how you've refuted (C): to do that, you would have to prove that the BoaPW contains no unpreventable evils, which is a tall order. Or am I misreading you here?

awatkins909 said...

Sorry to if I'm pushing the point too much but I am pretty curious about all this. That does help with option (A).

But yes, my argument is that this is not the best of all possible worlds, and that God could even actualize a world which is worse than the best of all possible worlds. I think there is a world where each moral agent always freely chooses to do what is good. God could actualize that world, but as we see in the post, God could actualize a worse world and still be omnibenevolent. Hence, omnibenevolence can't consist in actualizing the best of all possible worlds.

It also seems that there is no upper bound in the ordered set of "good worlds". For each world there is always some world that is better. This means the concept of "best of all possible worlds" remains undefined, just like "largest natural number", and hence God's omnibenevolence can't consist in actualizing the best of all possible worlds.

awatkins909 said...

It seems Aquinas does not believe in the BoAPW hypothesis either:

Leo said...

Thanks for the link. I'm not sure that I agree with the author's argument, but I'd have to chew on it before making an adequate reply.

This conversation will get us nowhere if we don't agree upon some standard of world-goodness. I was invoking a Leibnizian view, according to which a world is better insofar as it combines a greater economy of means with a greater diversity of ends, whereas you seem to hold to a standard in which a world without any immoral decision is better than one with the same. (On this, see Theodicy, I:10ff.) Similarly, we cannot debate the possibility of a best world without such a standard.

I should add that, while I find the case for a Leibnizian best of all worlds more credible than it usually gets credit for, I'm not entirely convinced by it, and only brought the matter into this discussion as a possible account of Divine omnibenevolence that avoids your criticisms.

(I have a post at my blog up on the subject of omnibenevolence, if you're interested.)