In defense of Timothy O'Connor against William Lane Craig on Necessary Existence (Setting the Stage)

In his book "Theism and Ultimate Explanation: The Necessary Shape of Contingency" Timothy O'Connor provided some original insights on the constitution of an absolutely necessary being. Even though he himself didn't drew the implications out, what resulted was a highly interesting analysis of what accounts for the necessity of said existence, as well as a new argument for divine simplicity (DS), which didn't rely on specific metaphysics but came from an analysis of necessary existence itself. I will quote the key passage at length and then will take on the criticisms made by William Lane Craig in the review of that work. I will take Craig to be representative of most analytic philosophers here and show how simple misunderstandings and the lack of openness to the topic of existence resulted in egregious mistakes and misrepresentations. What we will be left with is a new way to look at the topic and progress in the debate surrounding necessary beings, so that the lingering problem of arbitrariness can be resolved. 

"Is the property of necessary existence something that results from the rest of its essential nature, or from some part of its nature? Apparently, neither of these can be the case, because then there being, ‘in the first place’, so to speak, a thing having the ‘base’ set of properties giving rise to necessary existence would itself be a contingent fact, which contradicts the assumption that NB is truly necessary. That is, there would be a problematic explanatory/ontological priority of these base properties relative to the property of necessary existence – the problem being that the existence of the putatively necessary being would be only conditionally necessary on the instantiation of some more basic features. So, the logic of the concept pushes us to conclude that necessary existence is not a derivative or emergent property of NB, but a basic one.

Might it be that, while necessary existence (N), and certain other properties (call their conjunction ‘E’) are alike essential to NB, nonetheless N could have been conjoined to some other nature and not to E, as it actually is? (That is, might N be only contingently connected to E?) It seems not. Were this so, there would be no explanation for the fact that N is actually connected to E. N is by hypothesis a final, ultimate locus of explanation. There is no getting behind it to explain any purely contingent connections it has with other features.

So, it seems, there must be an internal, necessary connection between N and the correlated nature, E. Which way or ways might the entailment go? 

Might E entail N, but not vice versa, leaving open the possibility that there are two or more natures, E1, E2, . . . E20, each of which entails N while N entails none of them? Again, were this possible, E1 and E2 would be explanatorily prior to N. If we asked, Why do the natures E1 and E2 necessarily exist, and not some others (E20, say)? The answer would be "Because E1 and E2 – apart from N – are the sorts of nature that simply must be (whereas E20 is not)." But this (contrary to intention) can only allow for the entities’ existing by a kind of conditional necessity: given that there is an E1, it exists ‘of necessity’. And this is inconsistent with the claim that the necessity is absolute. So, we conclude, N entails the nature E." (pp. 87-88)

Conscious or unconscious of it, O'Connor's way of arguing is ingenious on multiple levels. First of all he's not treating "existence" as something innocent, to say that a particular thing exists is to designate something real and important. Secondly, nowhere in his book does O'Connor mention the debate surrounding the thin and thick theory of existence. This is important for he let's the argument and the necessity of the necessary beings' existence speak by itself, instead of introducing an additional topic in which analytic presuppositions could abort the debate from the start. Of course, since a thin theorist ultimately will reject necessary existence as unintelligible, due to the idea of a self-instantiating concept being an unintelligible one. But O'Connor, without mentioning it, shows the interesting opposition in which the PSR and the thin theory are. However, that's a topic for a future post. 

It might not be immediately clear what O'Connor is getting at and I would go so far as to claim that without a real engagement with the nature of existence, this is not possible. So before we can turn to Craig, we need to be clear just what O'Connor wants to say here. 

At first the question "What is it that makes a necessary being necessary?"is treated. The author points at something that is mostly lost to current debates about the nature of necessity. In order to be necessary, the being requires a certain relation to necessity, the being has to essentially exist. That means the existence must be an inseparable part of the beings' nature, something that is obviously not given in contingent beings, which, at least theoretically, could go out of existence. But just declaring existence a part of a beings nature doesn't end the debate. Because while it's true that essential existence tells us what makes a being necessary, it doesn't answer the logical follow-up question as to why a particular being is necessary. Neither does it tell us whether the ascription of essential existence to a particular being in the strong sense, entailing existence a se, is warranted. After all, as O'Connor indicates later, there's still the possibility of mere conditional necessity, given that a particular entity E exists, its existence is essential. The most we could get out of that is the future impossibility of its going out of existence, however that's insufficient to satisfy the PSR, given that E was still dependent on the fact that it came about, a factor external to it. This, of course, won't be the final answer on the quest we're at. 

The properties merely being instantiated makes the being no different from other contingent items. Thus there must be a deeper principle. O'Connor identifies three ways the existence could be related to the other essential properties: 1) as prior, 2) as on par/entailed by a part of its nature or 3) as posterior to it. 2) and 3) violate the PSR though, as well as the supposition that the being in question is truly necessary. Why is that? To answer that question we have to remember what existence exactly is, namely that what makes beings be as opposed to nothing. Existence is that what makes the difference between an empty world and those with beings in it. A being can't have a determinate nature/instantiated properties without existing (or it can't have an actual determined nature without existing, to accommodate the Meinongians). What this shows is that it can't be the nature or abstracted essence, that is fundamental. It must be the existence. An arbitrary nature can exist. But without existence there's no nature. 

The reason why I'm repeating that so often is because it's exactly what constitutes a breakthrough in the debate surrounding the nature of necessity. And it's exactly the misunderstanding of this what causes Craigs confusion, as we will see later. Because now we can see why O'Connor judges it problematic that the rest of the nature or at least a part of the NB could be prior or to its existence. The issue is that there's no account of how it's possible that the properties are instantiated without being within an already existing entity. The idea that the nature could entail existence, already presupposes its existence, for without it, it couldn't entail anything. Either we're left with a circularity whose resolving comes at a cost of accepting self-causation. However it's still doubtful that this would be consistent with a PSR, given that this account is still incapable of explaining the properties' own fundamentality. Which is exactly what O'Connor accuses the account of being, namely a proposal of a brute contingent fact. For while it can be conceded for the sake of argument that it's possible that the nature entails existence, this is dependent on the actual instantiation of that nature, as already argued above. This would result in conditional necessity; given that N exists, it's existence is essential to it, which doesn't give us a being that exists a se, but rather a being that couldn't go out of existence, once existing. That at most tells us why N continues to exist (though this can also be argued), but not why N exists at all. For an ultimate being, this is thus a failure. So necessary existence can't be a derivative or entailed feature, it must be basic to NB. 

The next question O'Connor asks might be thought of as having been answered through the passage above already, but it's worth it to go through this again. The topic at hand gets analyzed from different angles. He ask whether it's possible for necessary existence to only be contingently related to a particular nature, meaning that it could have been possible for necessary existence to be related to a different set of properties? The answer is negative. Why though? Remember that we're interested in the ultimate foundation of reality, a necessary existent being. And, as O'Connor formulates it, "N is by hypothesis a final, ultimate locus of explanation", meaning that the ultimate contingency must be explained by it. However in the idea at issue, necessary existence is involved in the contingency, making it incapable of providing that explanation. If the relation between necessary existence and specific nature were contingent, making it possible for the nature to not be related to it, then we have created two problems for the hypothesis: 

1) The being at hand ceases to be necessary. Necessary beings per definition exist in every possible world. If necessary existence is only contingently related to the nature of the supposedly necessary being, then it's possible for it not to exist. Hence, it wouldn't be necessary. 

Now suppose that this conclusion would be denied. Suppose for example that the being exists in every actual world, like Howard Sobel's "Dragoon" (a dragon that somehow tracks every actual world and thus exists in every of those). Once again we can give two answers. For one, even the possibility of its non-existence makes the entity in question contingent. Secondly, necessity isn't the mere quantity of possible worlds in which a being exists, it's about its mode of being (Vallicella, "Existence: Two Dogmas of Analysis", pp. 66-67, 2014). A divine idea exists in every possible world. But it does so dependently on the intellect. Therefore we can conclude that it exists ab alio, not a se. Existing in every possible world is a necessary, but not a sufficient condition for necessary existence. 

2) The necessary being is introduced into one's ontology as a requirement of the PSR, to remove every brute fact. The idea that the relation is contingent violates that exact principle, as it introduces contingency into the most fundamental aspect of reality, where there is no further explanation for it. We can therefore conclude that the relation between necessary existence and the nature must also be necessary. 

The third question O'Connor asks is, given that necessary existence is a basic property and that it is necessarily related to the rest of the nature, how does the entailment go? Is it the nature that entails necessary existence? Or is it necessary existence that entails the nature? 

I won't rehash the argument from mere conditional necessity the author already made in regards to necessary existence being basic. The same argument applies here, the nature being prior either presupposes its instantiation, requiring it to exist before entailing existence, thus resulting in an account of self-causation as well as an explanatory circle of embarrassingly short diameter or violating the initial assumption that it is necessary, by giving it mere conditional necessity. I can make another argument though based on modal continuity. 

The problem of the idea that nature might entail necessary existence is a lack of contrastive reasons, as well as, once again, the priority of what-ness over that-ness. While it would allow for multiple necessary existent natures, the question as to why such natures as opposed to others necessarily exist would have to be answered by pointing at said natures and proclaim that  these are the natures that simply have to exist as opposed to others. This is dissatisfying though. While it not only entails that we'd have no explanation for this necessary truth, we'd have an arbitrary stopping point. A substance isn't made necessary just by being the most fundamental. Neither is something necessary by being the first. The belief in an arbitrarily limited substance/object as the necessary being can't be justified unless alternatives can be reasonably eliminated. I already cautioned in the prior post that prioritizing the actual world over the merely possible commits a modal fallacy, if the probabilities or priorities are investigated. What I mean is that we can't just ascribe necessity to a singularity of the basis that it exists, making differently constituted singularities conditionally impossible. It's ridiculous to give ontological priority for one quantity over another. The same would go for the theist who too quickly would like to give God the necessity based on the infinity of the traditionally ascribed attributes. An omniscient or omnipotent being just on the basis of these attributes can't be regarded as a candidate for necessity as opposed to a hypothetical being that is just almost omnipotent and omniscient. Mere quantity, calculable or not, doesn't cause differences in the mode of being. 

Also a mere ascription of necessity ignores the consequences this has for the philosopher's ontology. Necessary existence is no small feat, it's arguably the most astonishing thing something can be. The ascription must be made discriminately, not arbitrarily, meaning that we need very good reason to attribute it. More caution has to be paid to investigate what that entails, as opposed to attributing it to a being just judged fitting. Based on O'Connors argument I will focus on that after defending him against Craig. 

I will focus on another reason for rejecting the idea that necessary truths lack explanations in a future post, it's of secondary importance here. For now I have explained what O'Connors argument amounts to and have given further arguments against the positions he rejects. Next time I will consider Craigs objections.