Richard Carrier on the argument from contingency

I’ll be blunt. I’m not a big fan of Richard Carrier. Read almost any reply from Carrier responding to a critic of his work, and you’ll notice how arrogant, condescending, and antagonistic he is with his interlocutors. It can be very difficult at times, for me, to read through a polemic work of his without being extremely put-off. (Just survey his articles wherein he claims that the late Maurice Casey, Joseph Hoffman, Stephanie Fisher, and Robert Eisenman are all literally insane.) However, while I dislike him, I do find that many of his theses are well-thought out, well-argued, and worth responding to—I think this specifically about his book Sense and Goodness without God  and what I’ve seen from his arguments for Jesus-mythicism (though I haven’t yet read his On the Historicity of Jesus yet).

Now, the other day I was writing an upcoming post regarding atheist responses (and mischaracterizations) of the cosmological argument, and I realized that I had never read what Carrier’s specific replies were regarding said argument. (I have surveyed Carrier’s thoughts on God as an ultimate explanation here, but I had never seen him directly address the cosmological argument.) I decided to look up some of his responses, and one response I came across consisted of a handful of Carrier’s supposed refutations of the argument from contingency (the argument I find most persuasive, and have defended here) promulgated in a rebuttal, which was part of a debate he engaged in with Tom Wanchick in 2006 (see here). I (obviously) found Carrier’s rebuttal to be inadequate—though to be fair his opponent Wanchick formulated the argument poorly—and I will argue presently that he failed to refute the argument from contingency, when the concepts are formulated properly.  

So, let’s dig in by surveying Carrier’s first objection:

First, Wanchick has not shown that the universe does not exist necessarily. Contrary to what Wanchick falsely claims, scientists now agree that we cannot know whether the whole of existence had a beginning, even if the observable part of it did, nor do scientists agree that everything that exists (including all space-time) will end, even if the present cosmos will.

While Wanchick might have attempted to ground his argument in a universe which had a beginning, the argument from contingency need not assume such a thing. For the claim “the universe does not exist necessarily” does not require a universe that must have had a beginning. That is to say, the eternality or non-eternality of the universe is completely peripheral. For the theist argues that even an eternal universe can still be contingent.

But how can we actually establish whether the universe, or any other existing thing, is contingent or necessary? Well if it is not of the nature of X to exist, then X must derive its existence from something outside itself, and therefore it is contingent. That is, if the nature or essence of X does not provide the reason for its existence, then X is contingent. Another way to think about this is that if I can contemplate the nature of X without simultaneously contemplating its existence, then the reason for the existence of X is not contained within its nature, and it is, again, contingent. (There are many other criteria that can be applied: if something is ontologically composed of parts, or if it is a composite of act/potency etc.) Therefore, we need only apply these criteria to the universe to determine whether is satisfies said criteria. And indeed, what does happen when one applies this standard? Well, the nature of the universe is that it is all space-time, matter, and energy. But obviously, contemplating all space-time, matter and energy does not clue us in regarding why space-time, matter, and energy exist in the first place. That is to say, contemplating the nature of the universe does not provide us with the reason for why it exists at all—and it certainly doesn’t tell us why this universe exists and not another, or why any universe exists instead of none at all. Therefore, the universe’s nature does not contain the reason for its own existence, and thus it is, by definition, contingent.

So while Carrier might have been correct that his interlocutor did not demonstrate why the universe is contingent, we surely have. And none of our arguments need rest on the premise that the universe had a beginning, or will come to an end.

Carrier moves to his second objection:

Second, Wanchick has not shown that "every substance [including the universe] has an explanation." Wanchick only offers as evidence our observations regarding the effects of a universe. But he hasn't demonstrated that these observations hold for a universe itself. The only way we can logically infer that what is true of "the effects of a universe" is probably true of the universe itself is if we assume the universe is an effect, since otherwise we only have knowledge of effects, and whether the universe is an effect is precisely the matter in dispute. Wanchick hasn't demonstrated that the universe is an effect, and if the universe is not an effect, what we conclude about effects within a universe will not necessarily apply to the universe.

Carrier’s point here is spot on. However, it’s only adequate based on the fact that his interlocutor is (again) doing a poor job promulgating the argument from contingency. We can do better. For the way to demonstrate that the universe necessitates an explanation is to demonstrate 1) that the universe does not contain the reason for its own existence (which we have done); and 2) that the existence of brute facts is impossible. For if one cannot logically rule out brute facts, then it is always possible that the universe just exists with no rhyme or reason, end of story. Luckily I’ve already argued that brute facts are logically impossible (here), and therefore we’ve demonstrated that the universe requires an explanation for its existence, contrary to Carrier.

Carrier articulates his third objection:

Third, Wanchick claims "only minds" can cause any time or location to exist, but this cannot be true. It is logically impossible for a mind to think or act without a time in which to think or act, and a mind that has no location exists nowhere and what exists nowhere does not exist.[2] Therefore, he has offered no logical explanation for space-time. Likewise, Wanchick claims "there was no nature prior to the universe" as a reason to reject natural causes of the universe. But if "the universe" includes time, then there can never be a time when the universe didn't exist--even if the universe began--and therefore it is logically impossible for anything to exist at any time "before" the universe, whether a person or a thing. So if there was no nature prior to the universe, there was no person, either.

Here Carrier is implicitly begging the question. (Let us ignore the Scholastic doctrine of analogy presently since Carrier was not debating a Thomist.) It is not clear that one cannot act without time or space unless you a priori assume metaphysical naturalism. That is, to act must be to be actual, or to exist, in some way—you cannot act and not be actual—and thus Carrier is essentially saying that there cannot be existents without space-time—meaning there can be no existence outside of the natural world—which begs the question in favor of naturalism. He does the same thing with regards to time. I agree that it is nonsensical to say that anything existed “before” there was time, but this only means that temporal existence is nonsensical “before” the universe. However, we cannot assume that existing things can only be temporal unless we assume naturalism beforehand—that is, unless we assume that existents can only exist inside space-time and the natural world. Carrier is simply begging all the questions he possibly can.

Carrier moves to his last objection:

Finally, it's not logically impossible that "it is in the nature of a universe to exist," and if it's in the nature of this universe to exist, then the existence of this universe is self-explanatory. Even if we accept that "the explanation of the universe must be a metaphysically necessary, uncaused being," which "metaphysically necessary, uncaused being" would that be? Wanchick hasn't demonstrated that this "being" can't be the universe (or some part of it), nor that it could be a god.

A few responses. First, while it might not be logically impossible that it be of the nature of the universe to exist—though I would challenge this as well [1]—this tells us nothing about whether the actual universe is indeed necessary. And reasoning this way, though I’m not claiming Carrier is, is exactly the fallacy that the ontological argument commits. Second, if we accept a necessary being as an explanation of the universe I don’t understand why Carrier thinks the question “What necessary being would that be?” is efficacious. For if we arrive at a necessary being we arrive at the ultimate explanation of everything. That is, we arrive at the terminus of explanation itself in that which explains itself as well as everything else. And unless Carrier thinks there can be more than one of these necessary existents then I fail to see how his question makes any sense, or has any force against his interlocutor’s argument. Moreover, while Wanchick might not have successfully demonstrated that this necessary being cannot be the universe (or part of it), we certainly have.

So we have surveyed Carrier’s arguments against (a specific formulation of) the contingency argument. While some of Carrier’s objections pose a problem for his interlocutor’s formulation of the argument, I think this is due to the lack of said interlocutor’s ability to promulgate a more persuasive version of said argument. For we have seen that Carrier’s objections really pose no problem whatsoever when compared against a more rigorous formulation.

_______________________________________

[1] While we might be able to imagine a universe which is necessary I don’t see that we can conceive of one. For by universe I, again, mean that which is all matter-energy, and space-time. But to reiterate, contemplating all matter, energy, and space-time cannot tell you why this matter, energy, and space-time exists in the first place—for if it could then matter, energy and space-time would be self-explanatory already, and this current universe would be necessary, which we saw above it isn’t.

Richard Carrier on the God hypothesis

Richard Carrier’s book Sense and Goodness without God is one of the best defenses of metaphysical naturalism that I have read. He is very well-read, especially when it comes to philosophy and science, and even though I don’t agree with many of his conclusions, one can infer that his position, as he presents it, is warranted.

That being said, there are a few places in Carrier’s book where he surveys the justification for belief in God, and (shockingly!) finds said justification wanting. Obviously since I’m a theist I vehemently disagree with Carrier, and (also shockingly) I found his reasoning to be a bit sloppy and confused. As such, this post is my attempt to demonstrate where exactly Carrier goes wrong, and why what he says actually points towards God, instead of away from him.

First, Carrier considers the idea that God could be an explanation for the universe. He claims that this explanation isn’t a “good one,” and articulates a few points attempting to demonstrate such:

[T]he idea that there was a God around before there was a universe—in other words the idea that something existed when there was no place for it to exist, that something acted when there was no time in which it could act—does not make much sense[.] (p. 72)

A few problems here. First, remember that God is claimed to be provable via logical deduction, and thus if this is indeed true (see e.g. the argument from contingency) then the existence of God is logically necessary., and therefore quibbling about how hard it is to fathom such an existence would be irrelevant. Let me demonstrate this with an example—from math! A math professor once showed me a proof that attempted to demonstrate that the sum of all positive integers equaled one-half. Hopefully it’s obvious that this conclusion seems absolutely absurd, since common sense clearly dictates that the sum of positive integers would converge to infinity.  Yet the fact that this proof infers something that is hard to fathom does not itself refute the proof. The proof still stands, and the only way I can refute such an inference is to find a mistake in the proof itself. Similarly, even if we couldn’t explain logically how God exists outside of space, time, and physical existence, this wouldn’t call the conclusion of such proofs into question, and thus it wouldn’t call God’s existence into question. Of course, remember that God’s existence would first need to be demonstrated in order for my answer here to be valid.

Moreover, classical theists have always posited what is known as the Doctrine of Analogy, which states that attributes predicated of God can only be attributed analogously, instead of literally. So when we say that God acts or thinks, for example, we don’t mean these words in the same way—nor in a completely different way—that we predicate them of ourselves or of natural things. Thus, Carrier’s claim that predications of God make no sense is only a problem if we are thinking strictly univocally. Now, this isn’t an ad hoc attempt to avoid the implications that Carrier has postulated. That is, the Doctrine of Analogy wasn’t formulated specifically to answer objections like those of Carrier. Rather, this doctrine is seen as a direct entailment of the doctrine of divine simplicity. Because if God cannot be composed of parts—as the doctrine of divine simplicity states--then there can be no parts in God that are distinguishable from each other. Thus since our predications of God are in fact distinguishable, then this must only be because our predications are analogous and not literal. So, when seen in light of these two points, Carrier’s problem here loses efficacy.

Carrier continues his explanation for why the “God hypothesis” does not constitute a good explanation:

And even if we can come up with an intelligible theory of creation, it still isn’t the best logical inference to make. Can we infer from what we see as a completely natural universe that a sentient Creator is behind it? Not really. Given the lack of any clear evidence for a god, and the fact that everything we have seen happen, which was not caused by humans, has been caused by immutable natural elements and forces, we should sooner infer the opposite: that immutable natural elements and forces are behind it all[…]Since we can explain everything by appealing to only those things and their properties, then (all else being equal) such an explanation is the most plausible one around—leaving no need and no sound reason to go beyond them and invent all manner of unproven entities, like gods and spirits and miraculous powers. Hence, even if the ‘god hypothesis’ were plausible, it would not by any means be the most plausible. (p. 72)

Again, we run into some complications. Carrier is viewing the “God hypothesis” through the lens of abductive reasoning here. That is, he’s comparing theism with other explanations (e.g., naturalism) and attempting to determine which provides the best explanation for the evidence we observe. “So what’s the problem?” you might ask. For isn’t reasoning to the best explanation a good thing? Indeed it is, depending on what’s attempting to be inferred. If you’re attempting to determine what a fossil is doing in the Cretaceous period, then abduction is perfect. But if you attempting to infer that the square root of negative two is an irrational number, then deduction, and not abduction, is your best tool. You see, if theists used abduction to try to reason to God—think of William Paley—then I would agree with Carrier’s claims. For Occam’s razor would entail that the God hypothesis multiplies entities extraneously and makes too many unnecessary assumptions. However, classical theism makes no use of abductive reasoning—and shouldn’t--but rather makes use of deductive reasoning. Deductive reasoning doesn’t use inference to the best explanation. Rather, deduction attempts to prove that something is necessarily the case. You see, classical theists didn’t argue that God is the most plausible explanation for the universe, rather they argued that God is a necessary condition for the existence of anything at all. And abductive reasoning cannot ground such a claim, but only deductive reasoning.

So Carrier’s claim that God would not be a good hypothesis is only efficacious if we follow him in using abductive reasoning to determine the ontological foundation of reality. But this is neither necessary nor logical. For if we can utilize logical deduction to infer what the ontological foundation of reality must be, then punting to abduction is superfluous here. The conclusions of deduction, if true, are necessarily true, and thus Carrier’s talk of best explanations is simply a category mistake regarding how the existence of God is usually arrived at. All this is to say that theists are playing chess while Carrier is playing checkers. If he wants to attack theism, then he needs to attack theists on the actual grounds that they utilize to make their arguments, otherwise he’s attacking a strawman.

Carrier then moves from talk of God as an explanation, to explaining God himself:

If everything must have an explanation, then you do not really get anywhere by explaining the universe by proposing a god. For then that god needs an explanation. Why does a god exist at all? Why that particular god and not some other? And where did this god come from? […] So either there is an eternal string of endless explanations, in which case there is no “ultimate” explanation because the explanations never end (and so the universe remains ultimately unexplained), or else there is something that has no explanation, something that just “is,” what we would call a “brute fact.” There are no other possibilities. (p. 73)

On the contrary, there is another possibility--something can be self-explanatory. Remember that classical theists believe, deductively, that God is a necessary being. The very nature of a necessary being is that its essence just is existence. That is to say, a necessary existent cannot possibly fail to exist. Therefore, when we ask “Why does the necessary being exist?” we’re really asking “Why does that which must exist, exist?” And such a question is self-evident and therefore self-explanatory—it answers itself. So, there is a false dichotomy here that Carrier is presenting between infinite explanatory regress and brute facts—which are impossible. An explanatory chain can end in something which is self-evident and explains itself—in fact this is done in mathematics! Now, Carrier (sort of) anticipates this answer:

 Thus, the question for us really is: Where do we stop [the explanatory chain]? What is the one, ultimate “brute fact” that needs no explanation? Certainly, most people say this is God, that God is self-explanatory, having no origin, that God exists necessarily as the one brute fact. But that requires resting on a huge number of assumptions. Why not just stop with what we actually know—the natural world? Certainly this is just as viable. After all, if god needs no explanation, then why does nature need one? (p. 73)

Many problems here. First, Carrier is conflating a brute fact with that which is self-explanatory. A brute fact is something which admits of no explanation, at all, while something self-explanatory explains itself, and thus does in fact have an explanation--this explanation is simply contained within the existent’s own nature. So when theists say that God is self-explanatory, they are not claiming that God is therefore a brute fact.

So, Carrier is confused by claiming that extending our existential understanding to include God violates Occam’s razor due to the fact that we could just as easily stop at another brute fact that requires less assumptions, namely the universe. Since God would not be a brute fact, then Carrier’s appeal to parsimony here is misplaced.

Second, there is a perfectly logical reason why we can stop the explanatory chain at God but not at the universe, as Carrier wants. For remember that something is contingent if we can conceive of it failing to exist—that is, if it’s nature does not contain the reason for its existence. Because if Y is part of the nature of X, then you cannot have X fail to be Y. The contrapositive is that if X can fail to be Y, then Y is not part of the nature of X. Thus, if we can conceive of something failing to exist, then existence cannot be part of the nature of that thing. And it should be obvious that the universe—including a multiverse—satisfies the definition of contingency. That is to say, we can easily conceive of the universe failing to exist in its current state, or at all, and thus the universe is indeed contingent. But this entails that the universe does require an explanation, since to be contingent is to be contingent upon something else. We see then that the universe cannot be a brute fact—especially since they’re impossible—and therefore we cannot possibly stop at the universe in an explanatory chain.

Carrier still tries to avoid this conclusion and claims that “The multiverse explains everything that exists,” but we’ve seen that this is false. For the multiverse does not explain itself, nor does it explain why anything at all exists in the first place. For this we need a necessary existent, and the universe is not it, since it satisfies the definition of contingency. Carrier again anticipates this objection:

Of course one could ask why the multiverse exists at all, and why it has the exact properties it does. But as we’ve already shown something must exist without any explanation at all, so it may as well be the multiverse. (p. 82)

On the contrary, we’ve seen that an explanatory chain need not—and cannot possibly—end in an unexplained brute fact. It can, and must, end in that which explains itself and which exists necessarily. And unfortunately for Carrier this cannot be the multiverse, for it is contingent and therefore necessarily entails an explanation outside of itself.

So Carrier did not accomplish his goal in this part of his book. He did not show how God is an unnecessary and extraneous hypothesis. His thesis rested on surveying the wrong type of reasoning method when put forward by those attempting to prove God’s existence. Then his thesis rested on conflations regarding the nature of explanatory chains and how these relate to the ontological foundation of reality. So while I admire Carrier’s attempt to vindicate naturalism in a systematic nature, I believe he ultimately fails.

Why something rather than nothing? Part III

In the last part of our series regarding naturalistic answers to the question “Why there is something rather than nothing?”, we shall survey the arguments of atheist Richard Carrier. Carrier’s explicit answer to this question can be found on his blog in a post entitled Ex Nihilo Onus Merdae Fit. Carrier begins by defining what he means by nothing:

[Nothing] can only mean that nothing whatever exists except anything whose non-existence is logically impossible. That latter caveat is unavoidable for the obvious reason that if it is logically impossible for something not to exist, then there can’t have ever been a state of being where it did not exist. So if by “absolutely nothing” you mean even the non-existence of logically necessary things, then “absolutely nothing” is logically impossible, and thus there can’t ever have been “nothing” in that sense.

A few caveats here. First, when someone begins the definition of nothing with “nothing, except…” then they have already gotten off on the wrong foot. Nothing, by definition means the absence of anything, at all. Period. It does not, contrary to Carrier, mean nothing except X,Y, and Z. For then, one is indeed talking about something.

Now, obviously Carrier has reasons for this definition, which he explains in the passage following the above quote. Carrier claims that logically necessary things must, by definition, always exist, and, therefore, there cannot be a state of reality without them. Is it just me, or is this the very answer that theists give for the existential question—namely, that since God is a necessary being, absolute nothing is not a possible state of affairs.(Now, this already demonstrates how ridiculous Carrier’s definition is. For if God did exist, even by himself, then by Carrier’s definition, this would still constitute “nothing”. But, I digress.) However, Carrier obviously does not adhere to this answer—and his reasons for it are given elsewhere. So, he must be talking about some other necessary thing:

[A]ll the fundamental propositions of logic and mathematics are necessarily true[…]and therefore there can never have been a state of being in which they were false.

 Carrier is claiming, then, that the laws of logic and mathematics are these necessary things that must exist. Fair enough. While it is not clear that these laws could be predicated in a state of “nothingness”—since these laws only describe the way existence behaves—it seems harmless to allow this premise in Carrier’s argument. Carrier continues:

Now, when nothing exists (except that which is logically necessary), then anything can happen (whose happening is logically possible). Because the only way to prevent something from happening, is to have some law or force or power or object or agency, in other words some actual thing, that prevents it. If you remove all obstacles, you allow all possibilities.

This is where Carrier goes off the deep end, and subsequently where his argument collapses. He is claiming that since nothing exists (that is, nothing is actual), then there are infinite possibilities (that is, infinite potential). This is completely incoherent. Why? Because actuality is ontologically prior to potentiality—that is to say, only something already existing can have potential. For example, water has the potential to become ice, but this potential is dependent on the already existing substance, namely, that of water. This potential would not exist by itself. Yet, this is what Carrier would have us believe, namely, that a big state of “nothing” has infinite potential! (Note: this objection was brought to Carrier's attention by individuals in the comment section of Carrier's post, but his reply boiled down to exclaiming "prove it".)

Another way to think of the problem here is the following. Potentiality must be predicated of something. That is to say, for potentiality to be a predicate, you must have a subject. To say “(x) has potential” one must substitute some thing for (x). Yet, Carrier can only substitute “nothing” for (x), which renders such a proposition nonsensical. Just ask the question “What has potential?” to which the reply will be “nothing.” So, not only is Carrier’s claim here silly, it is logically incoherent.

 But, wait, Carrier’s not finished:

Therefore, in the beginning, nothing existed to prevent anything from happening or to make any one thing happening more likely than any other thing.[…] Of all the logically possible things that can happen when nothing exists to prevent them from happening, continuing to be nothing is one thing, one universe popping into existence is another thing, two universes popping into existence is yet another thing, and so on all the way to infinitely many universes popping into existence[…]Therefore, the probability of some infinite number of universes having popped into existence is infinitely close to one hundred percent.

 So, Carrier has basically argued that from “nothing”, everything comes, multiple universes and all. There are many problems with Carrier’s argument here. First, Carrier has still given no metaphysical explanation of how “nothing” can spontaneously pop into something. This is seen to be metaphysically impossible—hence the old adage “from nothing, nothing comes”. For something to go from one state (nothing) to another (something) would be for potentiality to be reduced to actuality. But, 1) we’ve already seen that potential must be a potential of some actual thing, which is not nothing, and 2) something can only be reduced from potentiality to actuality by something already actual, which, again, cannot exist in nothing; that is, potentiality cannot move itself to act because mere potential is not actual.

Thus stated, we have seen that Carrier begins his argument with a dubious definition of “nothing”. Then he tried to predicate metaphysical concepts of this “nothing”, yet, said concepts can only be coherently predicated of something, not nothing. Third, even after all this ridiculous argumentation, Carrier has given no metaphysical explanation of how this ‘nothing” can produce anything at all. Rather, it is simply flatly asserted that this “nothing” can pop into multiple universes. Taking all of this into consideration, we can see that Richard Carrier has not given a coherent and tenable answer to the existential question.