Reflections on the Carroll-Craig debate-3: The back-and-forth

To follow up my earlier posts on the debate (see here and here), after the two main talks, we had two further talks by each that allowed for some back and forth, beginning at around the 1:04 mark (with an intermission between 1:27 and 1:35) and where the speakers repeated and reinforced their arguments and tried to rebut the other. There was some repetition, especially on the part of Craig. I will not go through it in order but summarize the main points.

Craig seemed a little frustrated at the way that Carroll had broadened the debate to one between theism and naturalism. He tried his best to get the debate back within the minimalist framing with which he had started and protested that the debate was not between theism and naturalism that Carroll had made it into and that the latter had introduced extraneous non-cosmological arguments to show that naturalism was superior. While this complaint may have had an effect on judges in a formal debating contest, here it came off as somewhat whiny.

Craig said he was surprised that Carroll had challenged the first Kalam axiom that “If the universe began to exist, then there is a transcendent cause which brought the universe into existence” and said that all it means is that ‘things couldn’t just pop into existence’. He said that to deny this is to be unscientific because the whole project of contemporary cosmogony is to find a cause for the universe. He was ‘astonished’ that Carroll would think that the universe could ‘pop’ into being out of nothing. He asserted, without explaining why, that there must be a metaphysical first principle that being cannot come from nonbeing. It is inconceivable to him otherwise, because if a universe can come from nothing, then why not bicycles or Beethoven?

He reiterated his fine-tuning argument and invoked the low-entropy of the initial universe and gave a list of scientists who think that fine-tuning is real. And Boltzmann brains made repeated appearances, with Craig saying that this is the biggest problem for past eternal universes. He says all such universes end up as De Sitter spaces and these are dominated by Boltzmann brains.

He said that anything could be the transcendent cause and that he was not arguing that the cause had to be an interventionist deity. He also said that he was not arguing that the purpose of the universe was to create us, and that there may well be life scattered throughout the universe. This of course is the classic Craig gambit, to argue for the absolute minimal non-natural influence and once you have that wedge in the door, then blast through that tiny opening with an interventionist god, Jesus, miracles, the whole religious kit-and-caboodle. It is completely disingenuous.

He tried to make the case that all of the models of the universe without a beginning or are uncaused are either not successful or are incomplete, depending heavily on Vilenkin’s views, which of course only pertain to classical space-time. He also said that Carroll’s argument that a quantum universe could exist for past eternal then raises the question of why it transitioned to a classical universe 13.7 billion years ago and not long before.

In response, Carroll reiterated that this debate was about whether naturalism or theism was better and that naturalism was preferable because of the three points he had made earlier: that naturalism works, the evidence is against theism, and theism is not well-defined

He said that Craig’s argument fails because he thinks that the same kinds of causal reasoning that applies to events within the universe that exist in time also applies to the universe as a whole. What applies to things inside the universe is different from what happens to the universe. Thinking they are the same results in asking the wrong questions, like looking at an iPhone and asking where the film goes. He says that saying things like ‘popping into existence’ for the universe presupposes something happening in time. It is more accurate to say that there was a first moment in time for finite models of the universe and there are models that do have such a first moment.

He said that we have no right to expect more than a complete model of the universe and to do so ends up asking the wrong questions. Inside the universe there are unbreakable laws and an arrow of time going from low-entropy to a higher entropy state and we can find cause and effects but these are not applicable to the universe or multiverse and there is no reason to expect them to do so.

He says that Craig’s criticisms of the cosmological models that are past infinite are simply wrong. There are, for example, bouncing cosmology models that are perfectly well-defined. In this bouncing cosmology, time is infinite and it is a classical description everywhere. There is no possible sense in which this universe ‘comes into existence’ at some moment in time and he challenged Craig to explain why this model is not valid. Of course Craig could not and ignored this question.

He also showed Alan Guth (one of the authors of the BGV theorem that Craig heavily relies upon) saying that he thinks the universe is likely eternal but that nobody knows for sure yet and that his theorem is only about classical descriptions of the universe and not about the universe itself.

Carroll also responded to Craig’s poll of scientists who said that fine-tuning is a real problem by saying that if we are using polls of scientists to make our arguments then he could just as easily do a poll of cosmologists about whether god had anything to do with the universe and he would win by a landslide. He said that there is fine-tuning but there is no evidence that this level of fine-tuning is required for life to exist.

Regarding the suddenly-famous Boltzmann brains, he said that the multiverse does not predict that everything happens but it predicts that certain things happen with certain frequencies so that you can look for multiverse models in which universes are not dominated by such brains and such models exist. In some models, it is actually easier to make a universe than it is to make a brain.

Carroll pointed out that Craig seems to use incredulity a lot as an argument, repeatedly saying that that he is ‘astonished’ by some of Carroll’s statements. But as Carroll said (quoting someone else), incredulity is not an argument and that “I do not know how to refute an incredulous stare”.

I thought that Carroll’s final statement from 1:46-1:52 was particularly good. Just move the cursor to that point.

Carroll said that 500 years ago it would have been reasonable to be a theist because that was the best that people could do with the knowledge they had but that got undermined by modern science beginning with by Newton, and after Darwin it was a complete rout. He said that a recent survey of philosophers, a notoriously fractious bunch, found a surprising level of agreement (73%) on three questions: (1) An external reality exists; (2) Science tells us something about that external reality; and (3) God does not exist

He said that people don’t become religious because of cosmological arguments but for other reasons. He threw Craig and theists in general a lifebelt, saying that they risk being completely marginalized by science unless they rethink the role that religion can play in society. They should stop arguing for the existence of god and instead focus on how to live life better, something that everyone should do. I don’t think they will, though they should.


  1. Marshall says

    Mano, I don’t completely understand the Boltzmann Brain argument. I mean, I *get* it, but the logic behind it just seems misguided to me, and I’m hoping you can enlighten me a bit.

    The claim is that a steady-state universe of high entropy undergoes random fluctuations, given enough time (and we’re talking unfathomable scales). These random fluctuations into low-entropy states are accompanied (and defined, I suppose) by configurations of matter, one of which is our universe, which includes billions of galaxies and human brains. While some may argue that this is how our universe came to be the way it is, the “entropy dip” required to create are universe is FAR greater than the entropy dip required to simply produce a single, sentient brain, a “Boltzmann Brain,” and therefore it is unfathomably more likely that we would be living in a “Boltzmann Brain” universe as opposed ours.

    Here is my problem with this argument: we can assume that the “rules” of the universe remain constant despite the magnitude of the entropy dip. In that case, doesn’t the “dip” simply provide some initial conditions from which the universe naturally evolves by chance back into a higher-entropy state? It is hard to imagine a universe beginning with higher entropy than ours that is capable of producing brains, period. We need to end up with an Earth-like planet that’s temperate enough for life to develop over some four billion years. In order for that to occur, we require a large mass to revolve around a star at a reasonable distance. In order for those scenarios to occur, we need galaxies, etc.

    My point I’m trying to make is–the universe is so homogenous that it has produced billions upon billions of galaxies that, from an entropic perspective, are “indistinguishable,” in the same way that we view a of gas of constant temperature as a microstate indistinguishable from others of the similar temperature. In that respect, isn’t it “easier” to make a billion random universes, and have life spring from one of them, than it for a random entropy dip to produce ONE universe that happens to have a brain?

  2. Marshall says

    P.S. I wish I could edit silly grammatical errors, such as “are universe” -> “our universe.”

Leave a Reply

Your email address will not be published. Required fields are marked *