More on logical & ontological possibility

Necessity-page-0Allow to me pull out a few relevant passages from the Hartshorne piece (Religious Studies, Vol. 13, No. 2 [Jun., 1977]), I linked to in the preceding post regarding logical and ontological possibility/necessity (all bold emphases are mine):

“Note that the meaning postulate used to define God can be rejected as lacking self-consistent meaning. Also some hold that ‘there might have been nothing’ is consistently conceivable, granting which the notion of a nature such that there logically must be something having it is absurd. So far as I recall, Hick nowhere discusses the conceivability of ‘there might have been nothing’….”

“What is Hick conceiving when he ‘conceives’ the divine non-existence? Is it ‘the existence of bare nothing’? I take this to be a series of words with no clear, consistent, specifiable meaning….”

“I insist there is a problem here that Hick has merely ignored. It is not incidental to my reasoning but central to it. I think Hick has entirely failed to show what ‘God does not exist’ means, assuming that ‘God exists’ has consistent meaning. He is comparing two allegedly conceivable but not mutually compatible ‘states of affairs’, but has given us no help in conceiving the negative alternative.”

“Consider now the points made by Anselm and me that a logically contingent statement must be such that, were it false, its being true would remain conceivable, and that this conceivability of both truth and falsity is intelligible only with assertions of things that are not eternal, that could come into or go out of existence. A still-born child that never became an actually thinking animal might have survived to become such an animal. A couple that had no child might have had one. Even the logical contingency (conceivable falsity) of assertions about non-eternal things is dependent, Anselm and I hold, against Hick, upon the non-eternality of their subjects. As Aristotle said long ago, eternal things are necessarily necessary and temporal things necessarily contingent. Hick wants necessity here to be merely ‘ontological’, that is, to mean self-sufficient, ungenerated and indestructible. I think Aristotle neither implied nor would have accepted any such view, any more than Anselm would have accepted it. Hick identifies God’s necessity with his eternity and self-sufficiency. However, no one, not even God, can wait forever to see if something is always there; for then he would never know its eternal status. It is necessity that explains eternity, not vice versa. And Hick partly sees this. God will live forever because he could not be destroyed. And this is the logical could not! Hick says so…but adds that this logical necessity depends upon an hypothesis that there is an ontologically necessary being.”

“The issue now grows clearer. Granted that God exists at present, there is no further logical contingency about his always existing. But his non-existence, Hick tells us, still remains conceivable. Yet, how do we conceive this allegedly conceivable negative case?”

“Anselm holds that some existential assertions, all ordinary ones, are logically contingent, that is, their denials involve no intrinsic absurdity, while at least one extraordinary existential assertion is not logically contingent, that is, its denial does involve intrinsic absurdity, either a misuse of words or a contradiction. Thus I cannot at all accept the gulf between talk of logical modality and what Anselm intended. He had the idea of logical necessity, though not the distinction between those cases which involve more than the constants of formal logic as now recognized, and those which involve meanings additional to purely logical ones. Either way the necessity in question turns entirely on the meanings of the terms employed.”

It is analytic that a thing to whose existence there is a conceivable alternative is either something producible (at least indirectly) by causes capable themselves of not existing or of not acting as they acted or would act in producing the thing, or else is something that could exist without cause of its existing and could fail to exist without cause of its non-existing. Since God is conceived as eternal and without cause of his existence, only the uncaused case is relevant. Such a causeless yet contingent existence is without connection with our ordinary ways of understanding contingency. I do not believe that any extraordinary way has been or is likely to be arrived at.”

We do not seek causal explanations of non-contingent truths, as in mathematics, but we do seek them for contingent truths. The empiricists tell us in effect to forget all this when considering God. They accuse Anselm of violating rules; but they violate the elementary rule that logically contingent matters are intelligible in genetic and causal terms, or not at all.”

“As Leibniz with his marvelous clarity saw long ago, a mere definition cannot establish existence, not, however, because there is no logical connection of ‘idea’ with reality, but only because it is possible for sensible sounding phrases to lack definite and consistent meaning, so that the assumed ‘idea’ is only a set of words. I find this more illuminating than all Kant’s lengthy verbiage on the proof. Some definitions, like ‘greatest possible number’, fail to express a coherent thought. Thus the conviction of so many that existence cannot be derived from a mere definition is fully justified, but not for the usual reasons. Thought does have a necessary connection with reality, for even contingent ideas make sense only because there is a creative process able to produce or not produce various things. What lacks necessary connection with reality is only words and sentences. If they fail to capture a thought, they will certainly not capture a true thought.”

“The foregoing theory of contingency, without which I take no stock in any ontological argument, means that to exist contingently is to be, or to have been, contingently produced, that is created. It follows of course that no eternal entity can be contingent. Am I now speaking of ontological or logical contingency? Neither, as Hick explicates terms. His logical contingency of the divine existence is, to me, a meaningless business of a way we can talk about nothing, while pretending to talk about something. All existence implies God as its creative ground, according to both Hick and me. Still there might be, he thinks, neither the creative ground nor its contingent creations. This might be is merely logical, it has no ontological referent. I think it is mere words.”

Logical & ontological possibility/necessity

HartshorneI’m not a process theologian, but I like a lot about Charles Hartshorne (1897-2000, pronounced ‘hearts-horn’ by the way), an extremely brilliant philosopher. Recent conversations about the distinction between logical and ontological possibility have taken me back to Hartshorne’s view that ultimately logical and ontological possibility coincide and are one and the same. But philosophers today (exclusively of the analytic sort?) distinguish between these two modes of speaking and being. Logical possibility/necessity is strictly about the formal-logical consistency of the terms of a proposition. Do the semantics of some claim or description violate the axioms of logic (about which there is ongoing debate), which traditionally are the law of identity (A is A), the law of non-contradiction (A is not not-A), the law of excluded middle (between A and not-A there is no third option). Others (e.g., Schopenhauer) add the principle of sufficient reason (PSR) as a fourth which states that everything that exists has a reason sufficient to explain its existence. Some of these four are hotly debated today.

It is generally (as far as I can tell) held that logical possibility/necessity concerns itself with language and conceivability. It’s primarily about speaking. Something is logically possible just in case its definition doesn’t involve any violation of the rules of logic. Ontological (or metaphysical) possibility/necessity describes realities themselves. It’s about being.

The question then is, what is the relationship (if any) between logical and ontological possibility, between logical modalities and ontological modalities? One option is to conflate them, take them as coincident and say that whatever is logically possible is ontologically possible, or that whatever is ontologically impossible is logically impossible. Another option makes a hard distinction between them and holds that neither mode of possibility can be assumed from the other. What is logically possible may in fact be ontologically impossible. But to begin to expose the difficulty of positing too absolute a distinction between the two, I don’t know of any philosopher who argues that what is metaphysically possible may be logically impossible. That is, it must be logically coherent to say of what exists that it exists, or what may exist that it may exist. Additionally, and controversially, others (e.g., Hartshorne) argue that it is not coherent to conceive of what exists necessarily as not existing.

A clear example of how the two manners of speaking relate would be the existence of God. Take the proposition:

P1  “A necessary being of infinite intelligence and goodness exists.”

(Assume whatever apophatic qualifications you wish.) I don’t know of any atheist who argues this is logically impossible (i.e., the words generate one or more violation of the axioms of logic), although all atheists are committed (though very few recognize it) to the ontological-metaphysical impossibility of such a being, for the only way a proffered ‘necessary’ being fails to exist is if its existence is impossible. Contingent existence isn’t an option for the God folks debate. He either exists necessarily, or his existence is impossible.

However, many theists who hold to the truth of P1 agree that its contradiction (“No necessary being of infinite intelligence and goodness exists”) is logically possible. That is, the non-existence of God is ‘conceivable’, because it is thought that conceivability is purely about the formal-logical consistency of language. “God does not exist” does not generate any obvious logical contradiction, hence it’s logically possible (i.e., conceivable).

Charles Hartshorne took the rare view that ultimately logical and ontological possibility coincide and are one and the same. I’ve always gravitated toward this view. It doesn’t make sense to me that once committed to the necessity of God’s existence, one should concede the conceivability of the non-existence of God. Hartshorne felt the same thing. He expressed this on many occasions and the linked article here is one such instance.


No one who believes in the metaphysically necessity of God can find his non-existence conceivable, at least not if one’s understanding of God is that he is the ground and source of all being (as opposed to believing in “a” God, Zeus for example). We’re speaking about a necessary being here, so our options are either he exists necessarily or his existence is impossible. Hence, to the extent one grants the logical possibility of God, one is committed to his actual existence. When theoretical physicist Steven Weinberg, for example, grants that God’s existence is indeed logically possible but he doesn’t believe God exists, he’s being self-contradictory.

So what do we do with logical and ontological possibility? Well, since it isn’t immediately obvious what metaphysical necessities there are, we do require a consistent and predictable way to talk about candidates, and to this extent Hartshorne granted a distinction between logical and ontological possibility/necessity. The laws of identity, non-contradiction, etc., remain the immediate rules for meaningful discourse. But, as Hartshorne’s arguments revealed, once one had reasoned one’s way through to conclude the existence of a necessary God, one was justified in maintaining the inconceivability of God’s non-existence. And the rules of logic held one to this. The inconceivability in question had to be argued for, of course. Hartshorne didn’t think metaphysical (a priori) truth was always obvious, laying on the surface. And he granted that debate over the status of such truth would continue. But where the mind embraced the truth of God’s existence, it could not consistently concede the meaningfulness (conceivability) of his non-existence, even if to enter into conversation about God he admitted atheist claims for examination alongside his own rather than dismiss atheist claims as obviously incoherent, like “Married bachelors exist.” But he did maintain that there are legitimate grounds for affirming the logical impossibility of the metaphysically necessary. So long as one is up front and careful about one’s use of language, all is well. There’s no logical reason why one one’s formal-logical language game must in all cases remain unrelated to one’s metaphysical commitments. Metaphysical commitments carry logical implications.

I’m doing a poor job of describing it. Forgive me. Enjoy the Hartshorne piece.