Reflections on the Carroll-Craig debate-2: Carroll’s response

In his opening remarks that I discussed yesterday, William Lane Craig made the standard opening move within the college debate framework for those speaking first and arguing the affirmative by framing the question in the most minimal way, thus making it harder to counter. He said that all he was seeking to show was that modern cosmology provided “significant evidence in support of premises in philosophical arguments for conclusions having theological significance”.

In his opening, Sean Carroll made a non-standard move for a respondent by ignoring Craig’s framing and reframing the debate as to whether naturalism or theism is the better way of understanding the universe. In a college debate contest, this would have been a dicey move with judges looking askance. But here Carroll did not have to worry about being judged by a panel using a debate-scoring rubric. Here he was trying to persuade a general audience and for them these debates, however they are worded or framed, are always about naturalism versus theism, so Carroll only risked annoying Craig with this gambit.

So Carroll began by saying that he was going to show that naturalism is superior to theism in understanding the universe. He pointed out that at cosmology conferences no one takes god seriously as a proposition to be included in understanding the universe. He said that naturalism is preferable to theism because:

  1. Naturalism works
  2. The evidence is against theism
  3. Theism is not well-defined

He immediately challenged the first Kalam premise (that if the universe began to exist, then there is a transcendent cause which brought the universe into existence) that Craig said was so obvious as to not need any argument in support. Carroll invoked Wolfgang Pauli’s famous dismissal of a paper by saying “It is not even wrong”. He said the first Kalam premise is worse than false because it is not even the kind of question that should be asked or the vocabulary that should be used, and that the kind of causal reasoning it invokes is dependent on a two-thousand year old form of metaphysical thinking that is no longer valid.

What modern science deals with are not transcendent causes but differential equations that reflect patterns, models, unbreakable rules, and laws of nature that connect events at one time to events at another. There is no need for any transcendent causes. The real question is what is the best model (i.e., formal mathematical system) of the universe that most closely matches what we observe.

He said that there are many models in which you can have universes that have a beginning but did not have a cause and this has been known for over 30 years. We also have many models of the universe that are ‘past eternal’, i.e., did not have a beginning.

Carroll pointed out that ‘theorems’ in science (like the BGV theorem repeatedly used by Craig) are not ironclad brute facts of nature but are conclusions that depend upon the assumptions that go into them. All that the BGV theorem states is that classical space-time breaks down at some time in the past. But this does not mean that the universe had a beginning because it may mean that it is eternal and that quantum mechanics becomes important.

Also when cosmological models speak of a finite lifetime of the universe, what they mean is that the models they use point towards a moment when t=0 in their equations. But while cosmologists do use locutions like “the universe began to exist” or “popped into existence”, they do not necessarily imply what Craig thinks they do, that there necessarily existed some pre-existing period before t=0 and that the universe came into being at that time.

Carroll granted that the teleological argument is the best argument that religious people have because it plays by the rules and compares two models that can be compared with the data, the theistic one and the naturalistic one. As for the argument about fine-tuning, the real question is why the universe ‘began’ (i.e., at t=0) in a low entropy state. In this way of looking at things, the thermodynamic ‘beginning’ suggested by the Second Law is not a beginning in time but the time at which the entropy of the universe was at a minimum. While we don’t know yet why this is so, we have various possible explanations.

He said that the theistic model fails for five reasons.

  1. It is not clear that there is a fine-tuning problem at all because we don’t know a priori what conditions are life-permitting. There may be many values of the fundamental constants that are life-supporting.
  2. God does not need to fine-tune anything because he can do anything what he wants. Fine-tuning may be apparent rather than real and the common arguments given by theists that the expansion rate of the universe must be tuned to one part in 1060 is simply wrong and the correct derivation shows that the probability is 1. i.e., all the early universe models have the right expansion rate for the universe to exist.
  3. The cosmological multiverse is not a theory but a prediction of a concise self-contained theory and allows for many universes being created continuously of which ours is just one. It allows for comparing with data
  4. There are plenty of viable models of the universe in which random fluctuations produce ordinary observers and not Boltzmann brains. Different multiverse models make different predictions for the ratio of ordinary observers to random observers. There are plenty of viable multiverse models in which Boltzmann brains or random fluctuations do not dominate. Furthermore a better understanding of quantum fluctuations suggest that Boltzmann brains and random fluctuations occur much less frequently than previously thought.
  5. Theism is a terrible theory of the universe. If it is done properly, then theists should be able to say what kind of universe they expect to see under theism and compare with the data. The universe we see has the kind of randomness and messiness that one would expect from naturalism than from any kind of purposeful design. Furthermore, theism would predict that life should be important in the universe while naturalism predicts it should be insignificant, that the presence of god should be obvious while naturalism predicts that god hides, that religious beliefs should be universal while naturalism says that it would reflect local conditions, that religious doctrines should be eternal and unchanging while naturalism says that it should reflect social progress, that moral teachings should be transcendent and progressive (i.e., sexism and slavery should have been always wrong) instead of inconsistent and tribal, that sacred texts should give us useful and interesting information which they do not, that biological forms should be designed rather than the end product of twists and turns of evolutionary history, that minds should be independent of bodies instead of changing with our bodies, should explain the problem of evil and not have random suffering, that the universe should be perfect instead of a mess, and so on. While theists can give explanations for all this, this is because theism is not well defined and so can come up with all manner of ex post facto explanations for what we see.

In reframing the debate in terms of the relative merits of naturalism versus theism in understanding the universe, Carroll wrong-footed Craig because the debate shifted to the merits of cosmological models and Carroll knows far more about this stuff first-hand than Craig, who has to depend on quote-mining of the scientific literature. Furthermore, Carroll is honest and does not make claims that are without merit just to score debating points. So when Carroll says that Craig has misunderstood what various cosmological models say and is wrong, it carries far more weight than Craig’s assertions to the contrary.


  1. machintelligence says

    Craig also got nailed in the Q&A by a questioner asking that, since cause must precede effect , how could this happen when time does not exist. His answer, that cause and effect were simultaneous, was weak because one could not then distinguish between the two. Did God cause the universe — or did the universe cause God?

  2. says

    What is a “transcendant cause”?
    The dictionary says “transcendant” is: beyond or above the range of normal or merely physical human experience.
    In other words, a supernatural cause. So, supernatural causes: therefore god. Sounds like WLC is arguing at about an 8th grade level, begging the question.

  3. Wylann says

    Marcus, that’s pretty much the Kalam argument in a nutshell. I remember first encountering it in a college philosophy course. I asked (aloud and in the middle of class) if theists had ever come up with something better, that actually relied on evidence. Some of the looks I got were pretty priceless. 🙂

  4. brucegee1962 says

    Here’s what I’ve come up with so far on the “why is there a universe, and what else is out there” front. As I see it, there are four broad categories of answers.

    1. Reality is coterminous with this universe we occupy. There are no universes anywhere, aside from this one. In that case, it makes perfect sense to go with the “began but not caused” model. You could also come up with some weird sorts of science fiction scenarios with this if you wanted to – say, in the future we invent time travel, then go back to the beginning of time and start up the big bang ourselves.

    2. The “infinite universes” theory – or at least the very large number of universes. The idea that every quantum fluctuation creates a new universe in which some different fluctuation took place, and all possible universes coexist equally. Aside from squashing free will and making all decisions and human existence pointless, this theory just seems wasteful. And it only involves infinite variations of our universe, with our own particular laws – what about all the other perfectly good ways that a universe might develop? And of course, this theory still doesn’t offer any explanations for where the whole mess came from.

    3. Having read Flatland at a very young age, I do tend to think that three seems like a rather arbitrary number for how many dimensions a universe might have, and ones with more are eminently feasible. One theorem that I don’t hear people talking about, but which seems perfectly valid, is that, if some kind of extra-dimensional space exists, then why must we assume that the generation of a universe from that space must be directed by some intelligence? Perhaps, in a universe with some higher number of dimensions than ours, the formation and destruction of universes is a natural process that comes with its own set of rules and laws. Maybe they grow on bushes, like berries. We’ve only got the one to run tests on – we should have the humility to admit that the mathematical conditions under which universes can come to exist are still pretty far beyond us.

    4. If there are other universes besides our own, it seems reasonable to assume they might be inhabited. And if they are inhabited, then there is no reason to assume they’d be incapable of creating a universe. Heck, we may be capable of creating one ourselves someday (if only in a Matrix sort of way). I just don’t think this hypothesis seems any more likely than any of the other hypotheses. And of course, just because an intelligence created us certainly doesn’t imply any kind of Abrahamic deity – I tend to favor the “made it for a science fair project, then left it in a box underneath his bed” type of theory. Whether or not there was any designer, there certainly doesn’t seem to be any evidence that it stuck around for very long.

    Are there any theories that don’t fit into one of these four categories?

  5. Mano Singham says

    There are the universes that are infinite in time. Also ‘bouncing’ universes. And other modes as well. I wish I could tell you more about them but not being cosmologist I am not as familiar with the literature as I might be.

  6. Wylann says

    I would recommend ‘The Elegant Universe’ as a good layman level book on the subject, although it might be a bit out of date now. brucegee1962, our universe is thought to have 11 dimensions, not 3.

  7. brucegee1962 says

    Thanks for the recommendation. Are the “rolled-up dimensions” of string theory really widely accepted by now, though? I thought they had enough problems in terms of not really predicting much, that they hadn’t caught on widely.

    I always liked the toroidal universe, but I guess dark energy blew that one right up.

Leave a Reply

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