I’m told there has been some discussion of my recent pantheism post on Michael Sudduth’s Facebook page. Since I ditched my Facebook account a couple of years ago, and Michael’s page isn’t publicly accessible, I can’t interact directly with that discussion. However, a mutual friend was thoughtful enough to send me a copy of his own critical comments, which I reproduce here:
I think James’ note is too quick. For what sort of ontological containment is at play here? Clearly many sorts of containment, such that A contains B, don’t support the inference that P(A) if P(B) for just any property P. Consider mereological containment, where A contains B just if B is a part (or perhaps proper part) of A. A very large clock tower — Big Ben, say — has many proper parts less than 1′ tall. But it doesn’t begin to follow that the same goes for Big Ben; it doesn’t follow that Big Ben, too, is less than 1′ tall.
Much the same goes for spatial containment, which James’ himself seems to dismiss as a relevent sort of ontological containment. My carton of non-fat milk and the refrigerator in which it’s contained have, among other things, very different dimensions and construction. Further, the milk can have soured and yet it still be false that the same goes for the refrigerator.
Perhaps, then, the relevant notion of ontological containment is that displayed by sets and their members. But this, too, won’t do, for of course while 7 is prime, then same can’t properly be said of (e.g.) the set of natural numbers of which 7 is a member.
Of course there’s much more to be said here. No doubt there are other notions of ontological containment which will support the general inference above, as well as (otherwise) faithfully capturing what the panentheist means to assert. Or perhaps we need to look more closely into relevant types of property; perhaps there are properties of some type, such that any property of that type does apply to the container if they apply to the contained item.
These are useful comments that raise some important issues. Here are some thoughts in response:
1. I agree, of course, that “if A contains B, and P(B), then P(A)” isn’t true for just any kind of containment and any property P. But consider the specifics of the argument: if God contains the world, and the world is a mixture of good and evil, then God is a mixture of good and evil (and thus, as I went on to argue, God cannot serve as the ultimate standard of goodness). This specific inference strikes me as a quite reasonable, and I doubt I’m alone. In fact, one might view it as an argument from the transitivity of containment. If A contains B, and B contains C, then A contains C. I cannot think of any kind of containment that violates this principle. So unless I’m missing something obvious, the inference appears cogent.
2. This raises the important question: What kind of containment is in view here? And the short answer is: Whatever kind of containment is typically had in mind by panentheists when they claim that the universe is ‘in’ or ‘within’ God. In my original post I characterized it as ontological containment. Here’s an attempt to define the notion a little more precisely: X ontologically contains Y if the being of Y can be considered (to some extent, in some non-trivial sense) the being of X. This is most obviously the case when Y is a part or a constituent of X.
So, for example, Harry ontologically contains his body insofar as his body is a part of him; his body’s being is also his being. Thus, to strike Harry’s body is no less than to strike him. Similarly, if Harry’s body is infected with the ebola virus then he is infected. These examples are particularly apt in the present context, because panentheists frequently analogize the relationship between God and the world to the relationship between humans and their bodies.
To take another example, consider Bertrand Russell’s “bundle theory”: objects are nothing more than ‘bundles’ of compresent properties. On this view, the red apple ontologically contains the property of redness (sharing it with myriad other red things, but that’s beside the point here). It may not be formally correct to say that the property is a part of the apple, but it is at least a constituent of it. The being of the apple includes the being of the property of redness. (Other than the being of its properties, what being does the apple have, according to this theory?)
3. Given this understanding of ontological containment, it seems hard to avoid the conclusion that if God ontologically contains the world, and the world ontologically contains evil, then God ontologically contains evil. God would be a mixture of good and evil, ontologically speaking, rather than absolutely good. And that gives rise to the normativity problem I detailed in my earlier post.
4. Why did I take this notion of ontological containment to apply to panentheism? Several reasons. First, it seems to follow from the etymology and standard definition of panentheism (“all-in-God-ism”) given that panentheism is a claim about the ontological relationship between God and the universe. Second, I take it to be a fair generalization from the various forms of panentheism I’ve encountered in my reading over the years. Third, it’s suggested by the very words Michael Sudduth used to describe his own panentheism:
Consequently, I now accept a panentheistic metaphysics in which the universe and human souls are, to put it roughly, in the being of God.
5. A panentheist, as one commentator suggested, might be tempted to appeal to the privation theory of evil to explain how God need not be polluted by the evil of the world. I don’t think this move will work, for the simple reason that the privation theory must apply to God in the same way that it applies to the world. If a privation of good in the world entails that the world is (partly) evil, by the same token a privation of good in God (by virtue of his containing the world) entails that God is (partly) evil. And presumably the same goes for any other theory of evil. It’s hard to conceive of a containment relation that would serve the panentheist’s purposes but isn’t transitive with respect to evil. (And it’s his burden, not mine, to identify that relation.)
6. So where do we go from here? As I see it, the panentheist has basically two options. Option A would be to accept my characterization of panentheism in terms of ontological containment, but argue that my argument against panentheism (so construed) is fallacious. Option B would be to reject my characterization of panentheism. In that case, however, I think the panentheist owes us an alternative characterization. What does he mean when he says that everything is “in God” or “in the being of God”?
In sum, what we’re looking for here is some substantive notion of containment or participation that (1) is intelligible enough for us to evaluate claims that appeal to it, (2) usefully differentiates panentheism from theism, and (3) offers the sort of philosophical or theological benefits that panentheists think can’t be delivered by a theistic metaphysics.
Consider again the principle I gave:
(*) If A contains B, and P(B), then P(A)
and let P be the property of containing C (for specific C). Then of course, so specified, this just is a statement of the transitivity of containment (assuming the same notion of containment is used throughout). More generally (as you write):
(TC) If A contains B and B contains C, then A contains C
Is this better? Does this make the original case against panentheism more clear?
Not obviously. Consider set membership, where A, B and C are sets. Does it follow that C is a member of A provided B is a member of A and C is a member of B? Pretty clearly not; for let:
C = {0}
B = {{0}} = {C}
A = {{{0}}} = {B} = {{C}}
Then B is a member of A, and C is a member of B, but C is not a member of A. Now no doubt, you might add, C is a member of the transitive closure of A; but that is a different set; and the fact remains that C is not itself — not directly, you might say — a member of A. So (TC) is false when containment is understood as set membership.
Much the same goes, I think, for mereological containment, i.e., for the the parthood relation. (Shane himself is part of the soccer team; but his two feet, while directly part of Shane are not thereby directly parts of that team. See, e.g., http://plato.stanford.edu/entries/mereology/ and the references there for discussion.) But the real issue lies in a different direction, for I think that the underlying intuition here is that if the universe or world were somehow contained in the being of God, then God himself would be infected by any of the evil displayed in that world; that evil would attach to God directly in virtue of the containment, however exactly that containment is understood. Again, there are properties of the world or universe or cosmos — most notably, the evil displayed therein — that transfer directly to God given that containment.
This is why I put the relevant inferential principle — (*) — in terms of property transference under containment, instead of (say) in terms of the transitivity of containment. But as we have seen (I say), neither principle holds in general, as it fails for such species of ontological containment as set membership and mereological parthood.
In any event what we have here is a statement of a panentheistic problem of evil. I don’t say that there’s no relevant problem here; or even that panentheism doesn’t founder on the rocks of inconsistency. With respect to classical theism, there are issues here to which Christian theists have attended for centuries.
Paul,
Thanks for the comments. I apologize for taking so long to reply.
I already dismissed the idea that set membership is the relevant containment relation. But leave that aside.
You’re right, of course, that neither set membership nor mereological containment (on most accounts) are transitive. Still, if C is a member of B, and B is a member of A, there’s a perfectly meaningful sense in which A contains C, even if C isn’t directly a member of A. Likewise, Shane’s feet may not directly be part of the soccer (football?) team, but one could hardly say that Shane’s feet sit out the matches.
In any event, it’s far from clear how your two counterexamples offer any consolation to the panentheist. The real issue, in my view, is summarized in the final paragraph of this post. Perhaps you can lend a hand to the panentheist here!
You’re right that I’m highlighting a panentheistic problem of evil. But as my original post made reasonably clear, it’s one distinctive to panentheism; it doesn’t afflict classical theism. So the standard Christian theist moves don’t seem to be relevant. However, I may be missing your point here.
No worries, James. I wrote a bit more about your post in a reply to (another) Paul below. At bottom, it’s less than clear to me that you’ve constrained a substantive objection to panentheism. And in the last paragraph (to which you refer to immediately above) you appear to agree. ;)
At all events, there is plenty here, I think, to think about!
What if one fills it out more:
If A contains B, and A is neither a set nor a mere collection, and B contains C, then A contains C.
On panentheism, I assume God is neither a mere collection nor a set.
So for mere collections, say, a soccer team. If the team contains me, and I contain my heart, the team doesn’t contain my heart.
But, for things that are not sets, like, say, a carpet, then if the section or part of the carpet we’ll call (5, 6) contains a wrinkle, the carpet contains a wrinkle.
Or, no?
I think the first thing to see is that James’ argument against panentheism — at any event, the reason he gave for not being a panentheist — depends on (*), or (TC), as a premise. But these principles are false with respect to at least some notions of containment — mereological parthood, for example, or set membership. Perhaps these aren’t the relevant sort of containment; but some notion of containment will need to be specified, on which (*) (or (TC)) is true, if that argument is to be sound.
Second, and obviously enough, not just any notion of containment in terms of which
(*) If A contains B, and P(B), then P(A)
or
(TC) If A contains B and B contains C, then A contains C
is true will do. If it is to relevantly engage a panentheist, or panentheism more generally, it should be a notion of containment that fairly explicates the relationship between God and the world to which that, if not the, panentheist cleaves.
Finally, I chose the sorts of containment I did — spatial, mereological, set-theoretic — because these strike me generally as fairly clear and reasonably well understood. (Your mileage will vary, of course. Mereologists will disagree on the matter of whether or not the is a part of relation is transitive.) And at least the latter two can plausibly lay claim to being species of ontological containment. Sets, for example, are ontologically dependent on the elements in their transitive closure; in which case, since sets also have their memberships essentially, it follows that sets that contain any contingent objects are themselves contingent.
Parthood or set-theoretic membership at all events seem fairly clear — clearer, I should think than the relationship between Harry and his body, or that, on a bundle theory, between a substance and the more basic entities of which it consists. Now, James does offer a sufficient condition for what he takes to be the relevant notion of containment:
(OC) If the being of B can be considered in some non-trivial sense the being of A, then A contains B
This is less than fully clear. Is it sufficient, on (OC), that the being of B simply be considered (or non-trivially considered) the being of A, or is the idea rather that the being of B just is the being of A? And what is the modality here?
However (OC) is spelled out, I think the trouble here is that this notion of containment doesn’t support James’ argument. For evil is no more the being of the world than is that infection the being of Harry’s body or redness the being of the bundle of compresent properties that is that apple. (That there is evil in the world is beyond doubt. We learn in Romans, indeed, that all of creation has been infected by sin and groans under its weight, waiting for liberation from this bondage. But this evil is not the being, the nature, of creation as such. Similarly with the infection or redness. There might be containment of a sort here, but it is not containment in the sense of (OC).) But then the latter do not contain the former and the application of the transitivity of containment doesn’t arise.
I find it interesting that Calvin, in his commentary on Acts 17:28, says that we are “after a sort contained in God“:
“For in him. I grant that the apostles, according to the Hebrew phrase, do oftentimes take this preposition in for per, or by or through; but because this speech, that we live in God, hath greater force, and doth express more, I thought I would not change it; for I do not doubt but that Paul’s meaning is, that we be after a sort contained in God, because he dwelleth in us by his power. And, therefore, God himself doth separate himself from all creatures by this word Jehovah, that we may know that in speaking properly he is alone, and that we have our being in him, inasmuch as by his Spirit he keepeth us in life, and upholdeth us. For the power of the Spirit is spread abroad throughout all parts of the world, that it may preserve them in their state; that he may minister unto the heaven and earth that force and vigor which we see, and motion to all living creatures. Not as brain-sick men do trifle, that all things are full of gods, yea, that stones are gods; but because God doth, by the wonderful power and inspiration of his Spirit, preserve those things which he hath created of nothing. But mention is made in this place properly of men, because Paul said, that they needed not to seek God far, whom they have within them.”
If we are “after a sort contained in God,” and “have our being in him,” then how is that much different from panentheism?
Nevertheless, I highly doubt that Calvin meant that we are “contained in God” in the same sense as panentheists do. For if panentheism requires us to be “ontologically contained” in God, or a “constituent” of God’s essence or being (which appears likely), then surely it must be rejected from a Biblical standpoint. Scripture plainly teaches that God is eternal and immutable in nature, whereas we have been created by God at a point in time. Therefore we cannot possibly be a “constituent” of God’s essence or being. No passage of Scripture can be legitimately construed to teach that.