"The critic of this can claim that all the explanations involved can be subject to translation into full explanations. Indeed they might go further and claim that all incomplete explanations only gain their explanatory force from being debt notes to be cashed out in the form of a full expatiation (even if these further features are at present epistemic brute). What is at work here is linguistic indeterminacy and not any genuine explanatory stopping point in reality. "

Yes, he can do that, but how plausible is that view, really? Prima facie, we *do* take incomplete explanations as good enough explanations, that's why one can even call them that. Whether or not they determine the outcome, they make it intelligible. The ball is on the critic's side and he has to argue that the intelligibility in question is just an illusion or "debt notes to be cashed out in the form of a full explanation". I find this immensely implausible. When people normally think about the proposed explanations, they don't actually treat them as debt notes, or expect them to be fleshed out eventually, they just accept that they make the explanandum intelligible. And they do so regardless of whether or not they believe such explanations can be fleshed out to entailment.

Explanations and intelligibility lie in the frontier between metaphysics, phenomenology and epistemology; we can study them ontologically, but we know about them from our experience of intelligibility.

I disagree with you about the quantum example; even when considering the question "why they do it at this moment of time rather than earlier or later", it is only brute if you insist that the explanans must determine the explanandum. But this is what is contended. There is an explanation in the conditions involved, only this explanation does not necessitate the explanandum. But it makes enough sense of the event and makes it intelligible.

Even if the quantum scenarios described turn out to be deterministic, the point is that none of this implies that a genuine explanation can't be "incomplete" or entailing. This is an open contention that the Van Inwagen objection proponent has to defend. And prima facie it seems very implausible, as 1) there seems to be no requirement of entailment in the notion of explanation, something can make another intelligible without necessitating it; 2) we do in fact accept non-necessitating explanations as genuine explanations as Anscombe, Pruss and Feser do in their examples.

Besides, libertarian free will, which many people believe in, also shows a way in which people explicitly accept non-necessitating explanations as genuine explanations. And that is the case whether or not LFW is true, because at the very least the model in question shows how many people accept non-necessitating explanations as intelligible and as making the explanandum intelligible. So, again, I think it immensely implausible to insist that every explanans must necessitate its explanandum. At the very least, it is up to the other side to argue their point.

Again speaking on free will, I think both Aquinas's and Saint Edith Stein's treatment of the issue are illuminating (his acceptance of finite goods as explanations for behavior, even though by their very nature as finite goods they cannot determinately attract the will; her distinction between causality and motivation, and how the second suggests or enlightens a certain action without causing or necessitating it, etc).

Last edited by Miguel (1/09/2018 11:10 am)

UGADawg wrote:

FWIW it's not exactly difficult to prove the move from the necessary being to the BCCF must be an entailment, even if we allow that some explanations aren't entailments. To quote from one of my friends:

What I'm trying to get at is if explanation are (sometimes) not entailments, then it's possible that they fail, even when the conclusion is also true - the explanation from P to Q has defeaters. In the example I gave, even though someone being pushed off a building explains that he died, and he did die, it was defeated by the man being shot on his way down. And, if the PSR is true, there is a fact to whether or not any given explanation is defeated for each defeater, and these facts must have an explanation. 

Therefore, given the explanation E, for the Big Conjunctive Contingent Fact (BCCF), I believe we can demonstrate that it is an entailment, if we accept than an explanation that gives an account of every possible defeater of it, forms an entailment. I haven't sketched out the formal proof of that part, but it seems reasonable - if it's impossible for there to be any other ways for the explanation could fail, then we know whether it does in fact succeed or not. And a true explanation that succeeds entails the truth of what it explains. 


i) An explanation with necessary defeaters or no defeaters is an entailment (as it would either always succeed or never succeed). 
ii) Therefore, if E is not an entailment, it must have contingent defeaters.
iii) Therefore, the defeaters of E are part of the BCCF
iv) Therefore, E explains the defeaters of E
v) Therefore, E is an entailment.

 
Perhaps I don't understand the argument. I don't follow how "a true explanation that succeeds entails the truth of what it explains". I don't see why an explanation being the only one and not failing to explain the explanandum, it means that the explanandum was necessitated by the explanation. It simply means that it's the explanation for it, and we could say that in every possible world w in which the explanandum exists, there is the explanans. But this does not mean that there ar eno possible worlds in which the explanans does not exist without the explanandum.

I'd like to see the formal proof, if there is any.

Last edited by Miguel (1/09/2018 5:02 pm)

Miguel with the "a true explanation that succeeds entails the truth of what it explains" he's just talking about those explanations that account for all their possible defeaters. He's not assuming, prior to the argument, all explanations are entailments; rather he's arguing that for any situation in which an explanation accounts for all its contingent defeaters (and it must have contingent defeaters in the first place, otherwise it's already an entailment), and the explanation is true, then all the defeaters must be false, so we have an entailment. This kind of situation obviously holds for the relationship between E and the BCCF, therefore it's an entailment.

Last edited by UGADawg (1/09/2018 6:37 pm)

If P explains Q, and there are possible words with P but not Q, given PSR there must be some reason Q doesn't follow from P in those possible worlds (otherwise brute), which he's calling a defeater. In every possible world with P & no defeater, then, Q obtains as well. So if there are no contingent defeaters it's an entailment.

At least that's how I understand him. If I can get in touch with him I'll ask him if he ever formally checked it with a computer; he's literally a professional logician, though, so I tend to trust his intuitions when it comes to formal logics.

Last edited by UGADawg (1/09/2018 8:55 pm)

UGADawg wrote:

If P explains Q, and there are possible words with P but not Q, given PSR there must be some reason Q doesn't follow from P in those possible worlds (otherwise brute), which he's calling a defeater. In every possible world with P & no defeater, then, Q obtains as well. So if there are no contingent defeaters it's an entailment.

At least that's how I understand him. If I can get in touch with him I'll ask him if he ever formally checked it with a computer; he's literally a professional logician, though, so I tend to trust his intuitions when it comes to formal logics.

 
My questions are:

1) Why assume that, If P explains Q, there must be a reason why Q *doesn't* follow from P in a certain possible world in which P obtains? It seems to me that there is only an explanation required for "why Q doesn't follow" if we are already presupposing that P entails Q, which of course would beg the question. If we don't believe that explanations entail what they explain, then there is no explanation needed for why we'd have P without Q, and the whole view that some positive defeaters are needed to account for why Q does not follow from P would be wrongheaded from the start.

2) If we grant that some contingent facts (namely, some free actions or decisions, as Pruss suggests and DanielCC mentions in his article) are self-explanatory, then in a possible world w God can choose not to create the universe in a contingent and self-explanatory act. This would be a defeater and part of the BCCF, but self-explanatory. Wouldn't this also defuse the argument?

3) What if God always has necessary reasons for both creating and not creating the universe? The necessary reasons for not creating the universe can explain why God didn't create the uniferse in possible world w. Why would this be a problem if we don't assume explanations must necessitate what they explain?

Last edited by Miguel (1/09/2018 10:58 pm)