The Other Problem with Nothing (and Kalam Shazam!) • Richard Carrier Blogs (original) (raw)

In May philosopher Alex Malpass will debate William Lane Craig on the Kalam Cosmological Argument for God, and in preparation he is running a weekly series at Thoughtology on every argument there is for or against every premise of the KCA. It’s looking like a great series so far (you can start with Kalam Debate Prep 1). Those who like to spend way more time diving the analytical semantics of the KCA than is usually healthy will love this. Highly recommended. It adds to my own brief Conclusively Ending the Kalam Cosmological Argument. And to all that you can add Phil Harper’s videos I’ve recommended before (Physicists & Philosophers Debunk the Kalam Cosmological Argument and Physicists and Philosophers Strike Back).

Malpass might not agree with all my takes. But so far he is covering things I have pointed out before, and with useful reference to the academic literature at every turn, as well as a patient, methodical discussion that really clarifies things. Which includes a point that inspired me to write this article (in circa 5,000 words), because one of my arguments did come up, from someone who asked him about it. Malpass didn’t have time to describe or analyze my argument, he only says that he thinks he has a better way to tackle the same point—and in the context of a debate, he’s right. I’ve used exactly his preferred tack in debates myself.

But digressing on the point Malpass didn’t go into is warranted here, because I think people could benefit from a short, to-the-point explanation of what it is, and why I think it’s important and useful outside clocked debates. Though I used it as an effective point of debate in my destruction of Michael Jones’s cosmological argument last year, I organized my entire debate around developing that, and for the productive reasons I will explain here. And that wouldn’t be apt for what Malpass is aiming for in his debate. I think his own tack will work fine and is worth developing there precisely to see how Craig handles it (or fails to).

Say What Now?

Okay. So, what are we talking about here? In his opening video Malpass examines the Argument from Intuition for the first premise of the KCA, that “everything that begins to exist has a cause.” And his last point about that is the Reification Fallacy, which in this context comes down to the claim that theists are illicitly “reifying” nothingness as a state of affairs—as (ironically) a “thing,” from which (then) nothing could arise.

Malpass prefers to come at this by kicking the legs out from under the fallacy itself: that it simply is illegitimate to insist that “nothing” is even a “thing” at all so as to produce their intuition in the first place. He does not mean that there was no beginning “before which there was nothing” (that’s premise two, and he will give some apt defeaters for that premise in a future video), but that even if there was, the theist’s intuition still fails. Because the idea that a nothing could never be followed by an uncaused something reifies “nothing” into a “something” with predictive powers or properties. And that makes it too much of a something to properly count as “nothing” in the sense they require.

In other words, if nothing is not a “thing” (if it is not even “a state of affairs”), then you cannot say things like “from nothing only nothing comes,” as that entails “nothing” predicts a sequence of events (that “the nothing” will not change, and will thus remain forever nothing). But if there is not even nothing, then there is nothing that can entail or predict any sequence of events. For example, a Kalamer might say “a state of nothing, qua ‘nothing’, contains no potentials, but an uncaused something popping into existence would only be possible if the state of nothing had the potential to do that, to suddenly change into something else—but it doesn’t; so it can’t.”

Malpass responds to this by saying that “nothing” is not a thing or state of affairs that can have any such power or consequence. It is literally nothing, and therefore can’t even “lack” the potential to change—because there is no “nothing” when there is no thing. Therefore (literally and figuratively) nothing is changing. All that exists is simply the first thing, which just exists when it does. There is nothing before that—not even a “state of nothing,” just as there is no north of the North Pole, “from which” one might fire a bullet at you, say, or “that which” can have properties like “lacking any potential to allow or produce something at the North Pole.” There being nothing north of the North Pole can have nothing at all to do with what can or can’t be at the North Pole. The theist’s intuition is therefore incoherent and thus unsuccessful in grounding their point. It just circularly presumes the premise that things can’t just “happen.” But what if they can? That’s what we are supposed to be debating.

The rebuttal then might be that if things can just “happen,” then bunnies and big bangs could just “happen,” and we would see that. My preferred approach answers that objection differently. But Malpass could answer it (as I also have) with permutative probability, as in quantum mechanics: why we don’t see that is the same reason you don’t ever quantum tunnel through your chair when you’re having lunch, even though QM does entail that’s possible. Because you actually see this all the time: spontaneous virtual particles are constantly just “hapenning,” and this actually causes things like the electromagnetic force to work the way we observe, but these are vastly more probable events, and at such small scales you can’t “see” them (other than through their effects). Whole rabbits popping out of hats are going to be vastly less common, and that’s why we don’t see that. I already explained this to Andrew Loke, albeit to no avail (because he’s not the brightest bulb). But the point is correct. If there is a continuous “things can just happen” property of reality, it will eventually spontaneously create whole worlds like ours, over and over. We just won’t see that because the cosmic die-roll needed sets its frequency at once every gazillion years. Same with rabbits. (And no, this does not allow theists to claim we’d all then be Boltzmann Brains.)

So the problem Malpass is pointing out remains. The theist is illicitly assuming that “uncaused origination ex nihilo” entails there is, first, “nothing” and then “there is something,” as if these are two points on a timeline, one transforming into or being replaced by the other. But that attributes too much substance to nothing. If there really is nothing, there is nothing there. So there isn’t a point on any timeline where there is “nothing.” No thing (not even a nothing) is transforming into or is replaced by another. There is just “a thing begins.” And that’s that. There is nothing before that that can have any properties capable of preventing that happening, or that could dictate what should happen—such as (for example) mandating that the first thing have a cause. There isn’t anything around to make or stop that being the case. Just as there isn’t anything around that “has” or “lacks” the potential to change. There isn’t anything around to change. And therefore the Kalamer’s intuition is void. They cannot say that there is some “thing” they call “nothing” that “lacks” the potential to change and thereby it “prevents” something from coming into existence uncaused.

Therefore, Premise 1, that everything that begins has a cause, cannot be established by this intuition. Malpass will address other arguments for Premise 1 in future videos. But this one is pretty much dusted.

This is the Other Problem of Nothing, in contrast to my original Problem with Nothing, which is to trace out what happens when we grant the theist’s reification of nothing (it doesn’t go well for them either). Both approaches are adequate to refute Premise 1 of the Kalam. But this one is easier. If Malpass gets hit with any similar argument in his debate I expect he will take his preferred tack and argue that you can’t establish that “if” there was a first thing (and thus a first moment of time) that anything even existed before that to dictate whether that thing could happen or not. Malpass uses Russell’s Denotation Argument to get there, and I think it will definitely stymie Craig.

But What If We Went the Other Way?

That’s all great. And Malpass explains it all well in his first video. But what I want to do is choose the other adventure, and explain why that has its uses too. The reason I became interested in the Argument from Nothing (or AFN, which presumes “nothing” is enough of a thing to have any consequences at all) is that it has surprising logical consequences that are of use for developing a proper naturalist metaphysics. Yes, it also “happens” to refute the KCA. But that’s not my actual reason for developing it.

It is still true that there are clever ways to use the AFN to defeat the KCA, which expose the Stalking Horse fallacy in Premise 1. A “stalking horse” refers to a hunter approaching their quarry by hiding behind their horse, as animals will be less spooked by a horse creeping up on them than by a predator doing so. It’s like getting the grass you’re hiding in to walk up with you. In the context of the KCA, the analogy is that theists are assuming way more unproven ad hoc things to get Premise 1 than are required to deny it, and are thus kind of “hiding behind their horse,” as it were, and in a way that lets naturalists to do the same thing right back at them.

After all, if a theist can sneak up to Premise 1 behind a bunch of unstated assumptions they haven’t actually defended much less proved, then a naturalist gets to sneak up behind their own horse of comparable assumptions. For example, if they get to just “assume” there is an unexplained disembodied superhero before there is even a universe for it to exist in, just to get to the KCA, then the naturalist gets to just “assume” there is a simple unexplained quantum potential before there is even a universe for it to exist in. And, boom! The KCA collapses. It then gets hijacked into an argument for a simple quantum potential, with no juice left to get a god in there instead. After all, who needs that complex mess, when you can replace it with something vastly simpler and backed by vastly more precedent?

But my interest in the Argument from Nothing is bigger than that. One thing I am rather tired of is theists hijacking every conversation, and compelling all philosophy to be about them and their tinfoil hat. They’re sucking up all the oxygen in the room. Naturalists should be doing philosophy, including their own metaphysics, wholly without any reference to or care about theism. I don’t actually give a shit about their weird flat-earthy cosmological theories. I want to explore naturalist cosmology. I want to come up with current “best guess” answers for why there is something rather than nothing, and why this something rather than something else, and where it all came from (or whether it’s always been around), and what grounds it all, making it what it is and keeping it going. I don’t care what theists go on about. God is just a bunch of woo balderdash and thus a theory well past time to bin. I want to think about what then is most likely the case. I want to do naturalist metaphysics. Because none other is worth the bother of doing.

So I have developed and burn-tested my Argument from Nothing because it might be a useful component of naturalist cosmology. It would have that use even if theism died out so hard no one even remembered what the idea of a god was much less how it ever functioned in philosophy.

Describing the Possibility Space

At a key point in his first video Malpass presents a tabulated representation of why his position (past eternal existence, which he labels for convenience “infinitism”) is more probable than the alternatives. I have doctored his slide here to squeeze in my own extra line for a fourth option, which I call causatum ex nihilo. Which is the conclusion my Argument from Nothing develops.

Efficient vs. Material Causes. This table has down the left four possible explanations of the universe (from three in the original), against two merits an explanation can have across the top, the two Aristotelian explanations of efficient and material cause. The explanation of an uncaused origination of everything ex nihilo gets two red exes (it fails to provide an efficient or material cause). And so does creation ex nihilo (getting another two red X's). Then Malpass put infinitism in and it gets two green checks, because that's both an efficient and a material cause. I squeezed in Causatum ex nihilo, which also gets two green checks, as explained in the text. Also on this screencap are a photo of William Lane Craig and a video shot of Malpass looking surprised (an amusing accident of when the screencap was grabbed).

Malpass shows that theists already agree that “uncaused origination ex nihilo” (the most common naturalist alternative apart from infinitism) fails to provide an efficient or a material cause “explanation” of existence. These are Aristotelian categories of explanation, which mean, in brief, the material out of which something comes to be (of which ex nihilo means there is none), and the thing, the impetus or change-agent, that brings it about. A hypothesis of “uncaused origination ex nihilo” fails to provide that, too, because it is suggesting that existence just started for no reason. Nothing brought it about or explains why it came about.

But Malpass then argues the theist doesn’t do any better with their theory, here labeled “creation ex nihilo,” because they admit there is then no material cause (God is not what the universe is made of, and nothing was already around for him to make it from—contrary to the literal text of Genesis, which actually says there was, but theists stopped actually believing their own Bible a thousand years ago), and while they usually try to insist God is at least an efficient cause, Malpass points out that, actually, he’s not. Because theists never explain how it is that a disembodied mind can do things—or at least things like “conjure stuff from nothing.” So they are actually handwaving, not explaining. They haven’t provided an efficient cause here. It’s like Donald Trump’s “concept of a health care plan,” an excuse for not admitting there isn’t one. Theists have the “concept” of an idea for a possibility of maybe God somehow being an efficient cause. But that doesn’t count as actually having a cause. It’s just begging the question. Indeed, worse, because the whole idea is not even all that plausible (see Prep 1 at 40:20). This is one of several problems theists have that were explored long ago by Evan Fales in Divine Intervention: Metaphysical and Epistemological Puzzles (IMO a treatise still unrefuted to this day). And I developed a similar point in The God Impossible.

Malpass then points out that his favored view, infinitism, ticks both columns and thus outperforms both of those other theories. Because then there is an efficient cause (it’s causes all the way back; there is no beginning and thus never a causeless thing) and a material cause (there’s always stuff around to be caused and do the causing). A theist might winge dubiously about infinities—or complain about how endless stuff can exist in the first place, what keeps everything together that whole time, but the same existential inertia that keeps God from dissolving could keep the world from dissolving too. So there is no way for the theist to win at that game (see Joseph Schmid and Daniel Linford, Existential Inertia and Classical Theistic Proofs).

So Malpass has a strong point, and it’s an entirely credible position to defend against the KCA. You might say that takes out Premise 2, and here we’re talking about Premise 1. But Malpass’s point is that any Argument from Intuition to Premise 1 actually turns into a defeater for Premise 2: if we get to intuitively say positing an actual efficient and material cause is better, then they are inadvertently arguing for infinitism, and thus the KCA is sunk by their own argument.

I would say there is a fourth possibility. And Malpass mentions others besides, but I think we agree they are harder to coherently render. So I’ll just squeeze in my fourth way here: causatum ex nihilo. It also provides a material and efficient cause by making “a state of nothing” itself the material that becomes matter-energy in space-time (by transforming an existing “flat” field into a “curvy” one, and a geometric point into an expanding space and time: see Nothing as a Field-State and Argument from Non-Locality, respectively). And it does that through its own inalienable properties, thus it is also a proper efficient cause, in precisely the way the theist fails to actually establish God to be. While they just handwave an inexplicable and counter-intuitive power for God to make things with a mere thought, I can actually prove the causal powers of a “nothing state” are logically necessary and fully explain all present observations. So that’s then now a contender. It is, like infinitism, a multiply better explanation of existence than the KCA strives to generate. Which kills the KCA. Indeed, it kills every cosmological argument there is—even “how so” arguments against infinitism. But I’ll get to that last.

I call this the Argument from Nothing and I’ve developed and tested it in a whole series of articles and debates, starting with two primary articles, the first developing the idea in detail, and the second connecting it to a related scientific theory in the literature:

I then debated this argument several times, showing how all the arguments against it fail:

And again in my debate with Michael Jones, but there’s hardly anything new in that to add.

The Argument from Nothing

Whenever theists argue that a “nothing” would prevent anything from happening, such as by “nothing” lacking the potential for anything to happen, they are granting that there was a “state of nothing” whose properties can determine what then happens. So if we grant their reification, we get the hypothesis of causatum ex nihilo: a state of nothing is the cause of something then happening, for exactly the opposite reason than the theist avers: nothing then exists to stop that being the case. I am not the first person to develop an argument like this (see bibliography below).

The gist of my own Argument from Nothing is this:

The idea here is that in extremis you can only have one or the other: the absence of all actuals, or the absence of all potentials. And a state of nothing that lacks all actuals is less (and thus more “nothing”) than a state of nothing that lacks all potentials. Therefore, the most empty state of nothing you can have is one that lacks all actual things, not one that lacks all potential things. The most empty state of nothing therefore will logically necessarily have every potential (not no potential as theists incorrectly conclude), because it has no actual thing constraining or guiding or compelling it to do only one thing, or only some things and not others, or even causing it to do some things more probably than others. Hence:

Again, if someone asks how “a state of nothing” can have the causal power to convert a potential outcome into an actual outcome, since it is supposed to contain “nothing” and thus should have no causal powers, the answer is that it is logically impossible to have anything more “nothing” than that and it not have that causal power. That simply is a logically necessary property of any such state of nothing, because there is no way to get rid of this power without adding something to prevent or control it—and a nothing lacks any such thing. So any “nothing state” logically necessarily has this power.

Put another way, a “state of nothing” that both is the most empty nothing there can be and lacks this causal power is logically impossible, and therefore there can never have been that state of nothing, Which means the state of nothing that theists want there to have been can’t ever have been. So they can’t appeal to it. It won’t ever have been the first state of anything. They have to appeal to what could be the first state of anything. And that simply is a nothing-state that has infinite causal potential.

And this follows by the logical necessity of probability. To say that nothing has no potential to become anything is to say that nothing will remain nothing. But to say nothing will remain nothing is to say that the probability is 100% that nothing will remain nothing and the probability is 0% that nothing will become something. But there is no reason why that should be. When nothing exists, there cannot be any thing that would skew the probabilities that way. When theists say nothing lacks all potentials, they are actually adding a thing to nothing (some power or rule or force that compels it to stay nothing and thus select one out of infinitely many alternative outcomes), and thus contradicting themselves. A proper definition of nothing lacks that thing. And lacking that thing logically entails the causal power to become anything.

One of the things it could become is “nothing,” i.e. one possible outcome is that it just stays nothing. But no law or power exists to decide what it will become, and thus every possible thing it could become must be as likely as every other (hence no outcome can be any more likely than any other). And from that it follows that “nothing” remaining “nothing” has effectively 0% probability, rather than the 100% probability theists incorrectly predicted. It also follows that of all possible things that it could become, infinitely many more of them will be quasi-infinite multiverses than lone universes, and therefore the probability is effectively 100% that if there was ever nothing there would immediately be a near infinite (or even actually infinite) multiverse, which explains all observations vastly better than any other theory (especially theism).

To see how that follows you’d have to deep dive my other articles. But the short of it is that if every permutation is equally likely (which is entailed by there being nothing to make any one possibility more likely than any other), then there are a whole cardinality more permutations that are vast multiverses than are single universes (see my discussion of cardinality in How All Math Is Real). So there are genuinely infinitely more multiverse permutations than universe permutations. It follows that the odds of a single-universe outcome are infinity to one against, even while a single-universe outcome is, in turn, infinitely more likely than “nothing” staying nothing—which is a single permutation of exactly zero universes. And a zeroth universe contains zero permutations, because it contains nothing to arrange differently—there is only one single arrangement of nothing. But there are infinitely more arrangements of a single universe than that—and infinitely more arrangements of multiverses _than that_—and not only that, but infinitely more arrangements of multiverses containing more randomly arranged universes than any number you care to select. Pick any number, and there are infinitely more larger numbers, and finitely many smaller numbers, producing a ratio of infinity to one in favor of the number of universes randomly selected to exist in a multiverse being higher than the number you picked. And this is true for all numbers you could pick.

One cannot object by saying it is counterintuitive that a state of nothing would have any (much less this) causal power, because intuition is useless in metaphysics (as Malpass himself explains in his first video). You have no pertinent experience to call upon here. What dictates the matter is logic, not intuition. And that gets us this:

Therefore “nothing” (the state of there being truly nothing) logically necessarily has this causal power. To make that not the case requires adding something to nothing to constrain it. And there is no such thing when there is nothing. Therefore, by its inherent structure, “Nothing” must obey an inevitable and unstoppable quantum mechanical law fixing the probability of outcomes, and the inevitably resulting probability distribution does not favor “remaining nothing” but “starting a vast multiverse.” And that matches observations better than any competing hypothesis.

This is also why it makes no sense to complain that we don’t see this causally efficacious nothing-state now. Because that state does not exist anywhere now for us to observe. “Nothing” only has this causal power when there is nothing. Now there is something. And everything that constrains and limits what can happen around us exists here and now (unlike there and then) and thus constrains and limits everything exactly as we observe (which is why, for example, All the Laws of Thermodynamics Are Inevitable and All Godless Universes Are Mathematical). Only when you remove everything—reduce space and time to zero, and remove all energy and matter, and all properties and structure producing any other physical law—will you cause this peculiarly generative state to exist.

Conclusion

This has been demonstrated on a scientific basis in an actual scientific study: Maya Lincoln and Avi Wasser, “Spontaneous Creation of the Universe Ex Nihilo,” Physics of the Dark Universe 2013 (see my discussion in What If We Reimagine ‘Nothing’ as a Field-State?). And there is correlating support in similar arguments by previous philosophers, who have all made one or another argument to the same conclusion, deriving the result that a nothing-state is maximally unlikely not to cause something else to exist, and thus the theist’s intuition about the causal powers of nothing is simply wrong. See:

The main value of this is not that it defeats the KCA (that’s just a byproduct) but that it provides us with a naturalistic cosmological theory of very potent predictive efficacy. Our world looks exactly like a randomly selected physical world, which is more likely in a vast multiverse of randomly arranged worlds. And lo, that’s exactly what we get when we posit a hypothetical starting point of a nothing-state ungoverned by any controlling physics (as being “nothing,” it wouldn’t be). As I just showed. And as Lincoln and Wasser show using a scientific framework. Both of which also rule out supernatural outcomes. It seems a strange coincidence that starting with the simplest possible hypothesis (a nothing-state that logically entails an infinitely variable causal outcome) generates exactly what we observe, and the very multiverse conditions that would make it likely (and which cosmological science is more and more finding evidence of).

It is only the more potent that this causatum ex nihilo model is compatible with all the most popular theories in the actual science of cosmology. For example, eternal inflation predicts from a single near-nothing state an infinite multiverse in which the budding universes are randomly scattered across time—and in block or “B” theory, the dominant theory of time in physics today, that scattering across all future time would arise in a single selection event. Of all possible permutations, few will have the universes bunched up early in time; therefore random selection predicts they’d be randomly scattered across time, which is exactly what eternal inflation predicts. It’s entirely possible that the hypothesized “inflaton field” that sparks spontaneous inflation ex nihilo is simply another way of physically describing the initial transition effect of a causally efficacious nothing-state (as, for example, Lincoln and Wasser propose).

But here’s the other utility function. The Argument from Nothing operates both temporally (if there ever was at some time past a state of nothing) as well as existentially—that is, if there has always been something. So it works even on infinitism. Because even then, if nothing exists to decide what that would or would not be, what turns out to exist could indeed be literally any possible thing, producing all the same logical entailments I find for the “initial nothing-state” condition. Likewise if we don’t reify nothing into a state: the existential probability of what first thing will then exist uncaused follows the same logical entailments.

So my argument applies no matter what condition you posit. If nothing makes any possible thing more likely, then “nothing” existing (in any sense at all) is literally the least likely state of affairs; and quasi-infinite multiverses, the most. And that is true on infinitism and uncaused first things (“origination from an unreified nothing”). So the Argument from Nothing provides a naturalist metaphysics for why anything exists at all, and why this thing and not some other. It’s a random selection effect—whether that happened at some specific time in the past, or is simply the atemporal “dice-roll” whereby what would exist was selected to be. “Nothing” was always one of the things that could be selected. But when nothing exists to decide what will be selected, the probability it will be “nothing” is effectively zero, while the probability it will be something like we observe is effectively 100%. Naturalism thus better explains everything than theism.

§