Why does anything at all exist? Someone could utter this interrogative form of words merely to express astonishment that anything should exist at all. But it is more natural to take the question as a request for an explanation: Why, for what reason or cause, does anything at all exist? What explains the sheer existence of things? Suppose we call this the ultimate explanation-seeking why-question.
Before attempting to answer this question, one ought to examine it carefully. One ought to question the question. If we do so, we soon realize that the question why anything at all exists can be formulated in two ways. One formulation is contrastive, the other non-contrastive:
Q1. Why does anything at all exist, rather than nothing?
Q2. Why does anything at all exist?
What this post argues is that Q1 suffers from a defect that makes it unanswerable, but that Q2 does not suffer from this defect. Failure to distinguish Q1 and Q2 may lead one to reject both questions as unanswerable. It appears that Paul Edwards makes this mistake in his entry "Why?" in the Encyclopedia of Philosophy. Anthony Flood may be repeating it here.
That these are distinct questions becomes apparent when we note that the questions rest on different presuppositions. Both questions presuppose that something exists. If that were not the case, there would be nothing to explain. But Q1 also presupposes that it is possible that nothing at all exist. Call this further presupposition P. P is no part of Q2, as I will explain in a moment.
Let us first think about P and what it entails. P may be expressed in several logically equivalent ways:
There might have been nothing at all
It is possible that nothing exist
Possibly, nothing exists
There is a possible world in which nothing exists
where ‘possible’ and cognates pick out broadly logical possibility.
No matter how P is formulated, it entails that everything that exists is contingent, equivalently, that nothing that exists is a necessary being. For if there might have been nothing at all, then any thing X that exists is such that it might not have existed. That is just to say that X is a contingent being. So given that Q1 presupposes P, and that P entails that there are no necessary beings, it follows that Q1 presupposes that there are no necessary beings. But this seems to imply that the question Q1 cannot be answered.
For if Q1 – or the asking of Q1 – presupposes that no being is a necessary being, then the asking of Q1 presupposes that there is nothing in terms of which an ultimate explanation could be couched. This is because an ultimate explanation of why anything at all exists cannot be in terms of a contingent entity. A contingent explainer would need explanation just as much as any other entity. An ultimate explanation, if one is to be had, must invoke a noncontingent, but possible, entity: one that either explains itself or at least is not in need of an explanation by another. (I am assuming that there cannot be an actually infinite regress of contingent explainers. This assumption is quite easy to defend, but I won’t address that task here.)
The upshot is that Q1 entails its own unanswerability. This is not because we are unable to know the answer, but because the question itself by its very structure rules out an answer. In other words, Q1 is self-defeating in that it rests on a presupposition that rules out an answer. The proper procedure with respect to Q1, then, is to reject it, not try to answer it.
But the situation is different with Q2. Q2 does not presuppose that every being is contingent. It does not presuppose the opposite (some being is noncontingent) either. Q2 is neutral on the question whether every being is contingent. This is why Q2 is not just a truncated form of Q1. It is not as if ‘rather than nothing’ is implied but not stated in Q2. Q2, resting as it does on different presuppositions than Q1, is a different question. Q2 does not presuppose the possibility of there being nothing at all, hence, does not presuppose that only what is contingent can exist.
Thus Q2 allows the possibility of a necessary being. Nothing about Q2 entails its own unanswerability. Q2 allows the following answer: things exist because one of the things that exist is a necessary being whose existence is self-explanatory, while everything else is explained in terms of this necessary being.
Whether this answer is correct is a further question. The present point is merely that Q2, unlike Q1, is answerable.
Dr. Vallicella:
Why do you think Q1 is committed to the presupposition that it's possible that nothing exists? It seems to me Q1 can be sensibly presented (or considered) in the absence of this presupposition, which is to say, nothing appears to be sacrificed if we assume that Q1 remains non-committal with respect to whether it's possible that nothing exists.
By way of comparison, consider the following questions:
(Q3) Why does God issue commands to perform good actions, rather than bad actions?
(Q4) Why does God exist in all possible worlds, rather than in only some?
(Q5) Why does every shape also have a size, rather than some shapes having no sizes?
(Q6) Why does everything which begins to exist have a cause, rather than no cause?
It doesn't appear that these questions, or the asking of these questions, commits one to the presupposition that the latter option is possible.
Posted by: Marc Belcastro | Friday, April 27, 2012 at 10:31 AM
Mr Belcastro,
That is an excellent comment, the kind I like to get, but rarely get.
In the first three of your examples, the 'rather than' clause refers to a metaphysical impossibility: God cannot command bad acts; God cannot exist in only some worlds; shaped objects cannot be without size. The fourth example is not so clear since it is not obvious that it is metaphysically necessary that what begins to exist has a cause of its beginning to exist.
Be that as it may, your first three examples do cause me some trouble.
Your main point is that I haven't argued that (Q1) presupposes (and thus entails) that it is possible that nothing exist but simply stated this as self-evident.
This is a tricky topic, but it seems to me if you ask why something rather than nothing then you are assuming that it is in some sense possible that there be nothing at all. To spell out the question. There is something. There might have been nothing. Why then is there something? If it is necessary that there be something, then the question could not get off the ground as a genuine question. If I am convinced that there must be something, then I cannot genuinely ask why there is something. When I ask why something, I presuppose that it is possible that there be nothing.
What's more, it seems I am presupposing that it is really (not merely epistemically)possible that there be nothing.
With respect to your examples I would make the same move I made with respect to the something/nothing question: I would distinguish two forms of the question and argue that only the second is answerable.
Posted by: Bill Vallicella | Friday, April 27, 2012 at 11:26 AM
(p) there is actually something.
(~p) there is not actually anything.
1. If something can’t come from nothing, then something can’t be reduced to nothing.
2. Something can’t come from nothing.
3. Therefore, something can’t be reduced to nothing.
4. There is actually something.
5. Therefore, there can’t actually be nothing.
6. Therefore, ~p is impossible.
7. “a state of affairs is necessary iff its negation is impossible.”
8. ~p is impossible.
9. Therefore, p is necessary.
I’m not sure how well or if the above will hold, but if it does:
“Q1. Why does anything exist at all rather than nothing?”
A1. Because [if 1 & 2 are true] a state of affairs in which there is not actually anything at all is impossible, ergo, a state of affairs in which there is actually something is necessary.
Posted by: duffe | Friday, April 27, 2012 at 11:33 AM
(1) needs to be reformulated. If God can create out of nothing, then he can annihilate what he has created. In this sense, something can be reduced to nothing, contrary to what you say.
Posted by: Bill Vallicella | Friday, April 27, 2012 at 05:01 PM
"(1) needs to be reformulated. If God can create out of nothing, then he can annihilate what he has created. In this sense, something can be reduced to nothing, contrary to what you say."
If a thing, a being, is essence and existence (esse); I’d agree God could annihilate all essences. However, if God’s essence is isness, I don’t think he could annihilate existence itself without annihilating himself in the process. In other words, when God annihilates a being, I think he’s just annihilating that being’s essence not that being’s existence, because it’s not really that being’s existence to begin with; it’s God’s existence.
So if God were to annihilate every essence, all essences would be reduced to God’s essence, esse, which I have a hard time identifying with nothing at all. I mean, if God did annihilate everything, we could surely say there is no thing in existence, just as there was no thing in existence before God created the world. But to say there is no essence in existence is not to say there is no essence at all. For if God’s essence is esse, just as with creation whereby all things in existence can be deduced from God’s essence, with annihilation all things can be reduced to his essence as well.
Posted by: duffe | Saturday, April 28, 2012 at 11:15 AM
Dr. Vallicella:
Thanks for your thoughts, and kind words.
I agree that this is a tricky topic. Your comments helped elucidate your intuitions about the matter.
I might be misunderstanding you—and this might not be the best way to express my observation—but you seem to be approaching or presenting (Q1) and potential answers at the same time (as it were), where consideration about one of the answers to (Q1) influences how the nature of question is regarded. I’m suggesting that we (could) approach or present (Q1), however, without yet considering the candidates for answers. Thus, the question itself, or the one who asks the question, remains entirely non-committal with respect to judgments about the answer candidates. One particular answer to (Q1)—namely, that there’s at least one necessary entity—can explain that the “rather than” clause doesn’t represent a genuine metaphysical possibility, and this appears to serve as an adequate explanation of why there’s something rather than nothing.
Maybe this will help better illustrate what I have in mind. You write: “To spell out the question. There is something. There might have been nothing. Why then is there something?” My suggestion is that we needn’t (pre)suppose that there might have been nothing when we ask the question. Perhaps we could spell out the question like this instead. It is the case that there is something. It’s not the case that there is nothing. Why, then, is it the case that there is something, rather than nothing?
I guess as long as the answer candidates aren’t obviously impossible in the strictly logical sense or otherwise incoherent, maybe we can get the question off the ground without the propulsion of P: when we ask why something rather than nothing, we needn’t presuppose that, possibly, there is nothing, but only that there’s an answer which explains why it’s the case that there’s something rather than nothing.
Posted by: Marc Belcastro | Saturday, April 28, 2012 at 02:29 PM
Marc,
Are you the Dayton, Ohio M.D. I corresponded with a few years ago?
Does 'rather than nothing' add anything or is it redundant? If it is redundant, then the question simplifies to (Q1).
Are you saying that the two questions are really one: that (Q1) boils down to (Q2)? If yes, then I think we are in agreement.
My later post on contrastive explanation is relevant.
Posted by: Bill Vallicella | Sunday, April 29, 2012 at 05:34 PM
Yes, I’m the one you corresponded with a few years ago. Good memory. But no, I’m not a physician. That’s my father, whose name I share.
Good questions. Regarding the first, I don’t think that “rather than nothing” adds anything to the question, given that “something” and “nothing” are exhaustive and mutually exclusive. By providing a non-contrastive explanation of why there’s something, it seems to me that a contrastive explanation is also provided, for there’s only one other option—whether that option is possible or not—which is thereby eliminated. This answers your second question: (Q1) does boil down to (Q2), at least in the sense that answering one of the questions answers the other as well.
Thanks, by the way, for the pointer to your other post on contrastive explanation.
Posted by: Marc Belcastro | Monday, April 30, 2012 at 09:48 AM