More objections to the science of cosmic fine-tuning

August 30th, 2015 in clues. Tags: , , ,

Growing cocoa

The science of fine-tuning is summed up in this statement: of all the possible universes allowed by theoretical physics, an extremely small number would allow the evolution of intelligent life.

Very few cosmologists seem to contest this statement, and Luke Barnes names more than 20 of the best who support it. Yet it is common to hear objections to some aspects of it, or at least suggestions that fine-tuning isn’t as certain as sometimes claimed.

In a recent internet discussion, several suggested objections were made, mostly based on a debate by cosmologist Sean Carroll. I am not a cosmologist, so I am struggling a little with some of these objections, but I thought they were worth a look. (This post covers similar territory to The science of universal fine-tuning.)

In this post, I am just considering the science of fine-tuning, as defined above. I will look at the theistic argument, based on the science, in a follow-up post. It is important to distinguish between the two.

The evidence briefly stated

In the past half century, cosmologists have developed their understanding of astrophysics to the extent that they can know calculate the effects of changing cosmological laws and constants. They have found that only small changes in any of about a dozen physical constants (e.g. the gravitational constant or the mass of a neutron) would not only make life as we know it impossible, but would generally result in the universe either having only a very short life or consisting only of Hydrogen or Helium.

The science is well enough developed that equations can be evaluated and graphs drawn, showing how only a small part of the sample space allows life, a conclusion most eminent cosmologists support.

The objections

My friend and Sean Carroll between them had the following objections or questions:

  1. Do we understand the laws well enough, and can we model them accurately enough, to know the fine-tuning is real?
  2. Can we calculate probabilities with any accuracy? Could there be an infinite number of universes that make probability calculation meaningless?
  3. Do we know the conditions under which life can and can’t form, or even define life?
  4. Could there be universes with totally different structures than we have even imagined, that might support life?
  5. The multiverse provides an explanation of how fine-tuning could occur.

Sean Carroll made two other points that relate to the theistic argument rather than the science, so I will defer them until my next post.

1. Do we understand the laws well enough?

Cosmologist Luke Barnes says: “The theories on which one bases fine-tuning calculations are precisely the reigning theories of modern physics. These are not “entirely new physics” but the same equations (general relativity, the standard model of particle physics, stellar structure equations etc.) that have time and again predicted the results of observations, now applied to different scenarios.”

It is hard to see how the 20+ eminent cosmologists who are quoted to support the conclusion that the universe really is “fine-tuned” would have come to that conclusion lightly, and in some cases (e.g. Rees, Susskind, Davies, Penrose, Smolin) written books on the topic.


Sean Carroll suggests that one of the common fine-tuning claims, that “the expansion rate of the early universe is tuned to within 1 part in 10^”60” is mistaken. In fact, he says: “If you ask the same question using the correct equations you find that the probability is 1”

I have insufficient understanding to assess this claim, but I note that:

  • it appears that most other cosmologists don’t draw the conclusion Carroll does;
  • Luke Barnes makes the point that if inflation is not itself evidence of fine-tuning, it may be that it is itself even more fine-tuned by some other factors (he quotes Hollands & Wald: “although inflationary models may alleviate the “fine tuning” in the choice of initial conditions, the models themselves create new “fine tuning” issues with regard to the properties of the scalar field”); and
  • Carroll himself says: “The homogeneity of the early universe, however, does represent a substantial fine-tuning …. inflation only occurs in a negligibly small fraction of cosmological histories, less than 10^-6.6×10^7” and “… analysis shows that inflation doesn’t really change the underlying problem — sure, you can get our universe if you start in the right state, but that state is even more finely-tuned than the conventional Big Bang beginning.”

I don’t really understand Carroll here. Has he changed is mind between those quotes and his debate with Craig, or did he introduce a red herring into the debate? I don’t know enough to say.

But I conclude then that the fine-tuning of the universe is a well supported scientific conclusion. We can only go with current science – if it changes, we re-assess, but there seems little likelihood that this science will change much.

2. Can we calculate probabilities with any accuracy?

Luke Barnes shows graphs that depict the fine-tuning in mathematical terms, and offers the following fine-tuning numbers: “the cosmological constant alone gives 10^-120. The Higgs vev is fine-tuned to 10^-17. The triple alpha process plausibly puts constraints of order 10^-5 on the fine-structure constant. The “famous fine-tuning problem” of inflation is 10^-11 (Turok, 2002). The fine-tuning implied by entropy is 1 in 10^10^123 according to Penrose”. To these we can add Smolin’s calculation of the probability that stars will form as 10^-229.

These cosmologists didn’t just make these numbers up! Penrose was Professor of Mathematics at Oxford University and presumably knew what he was doing. Certainly the numbers are not precise, but Barnes makes the point that with numbers this large, order of magnitude calculations are quite sufficient.

I have seen several arguments against the calculation of probabilities.

Perhaps there is an infinite number of possibilities, so how can we calculate probabilities?

The probabilities are calculated based on what is known. If the sample space is even larger, this further reduces the probabilities. If, somehow, the sample space could be infinite, the probability approaches zero. I can’t see how this argument takes us anywhere except reinforcing the improbability of the fine-tuning.

Probability estimates assume an even probability distribution.

I guess this is true, but two points seem relevant here:

  • Our definition of fine-tuning doesn’t depend on a numerical probability, but on the observation that “of all the possible universes allowed by theoretical physics, an extremely small number would allow the evolution of intelligent life”.
  • Luke Barnes points out that “to significantly change the probability of a life-permitting universe, we would need a prior [probability] that centres close to the observed value, and has a narrow peak. But this simply exchanges one fine-tuning for two — the centre and peak of the distribution.”

So I conclude that probabilities can be calculated with sufficient reliability to draw general fine-tuning conclusions.

3. What about life of a very different form?

This is perhaps the most common objection to fine-tuning. Aren’t we being a little parochial to only consider carbon-based life which depends on oxygen? But the fine-tuning argument is stronger than that. Here I will quote Luke Barnes at length, commenting on Sean Carroll’s statements (shown in italics):

“There are changes we can make to the laws of nature that result in a universe so simple, so barren, that by any definition of life, this isn’t it. The cosmological constant is a good example: we have a 120 orders of magnitude to play with, but after even 10 or 20, the universe contains nothing but an expanding hydrogen soup. Such a universe is very easy to predict – the universe never leaves the “linear regime”. We can solve the equations of cosmological structure formation. Compared to calculating the behaviour of our universe, this one is a doddle.

“Carroll says: “The results are going to sound like they come from a science fiction novel.” I think that that statement is false. Consider a universe with too large a cosmological constant. Shortly after the beginning of the novel, the only thing that happens in the universe is two hydrogen atoms colliding every trillion years or so. The results will make for the most boring science fiction novel conceivable.

(Barnes says elsewhere: “If the strong force were weaker, the periodic table would consist of only hydrogen. We do not need a rigorous definition of life to reasonably conclude that a universe with one chemical reaction (2H → H2) would not be able to create and sustain the complexity necessary for life.”)

[Carroll again:] “We just don’t know whether life could exist if the conditions of our universe were very different because we only see the universe that we see.”

“I don’t know how a theoretical cosmologist can make a statement like that. …. If Carroll’s problem here is an in principle problem, then his objection amounts to a denial that we can do theoretical physics. The job of the theoretical physicist is to take a given law of nature (and its constants), and predict its consequences. This usually involves solving the equation. Asking whether a given set of laws and constants would produce life is the same type of question as whether they would produce atoms, rainbows, galaxies or a CMB.

“Granted, life is a more difficult task. But …. we can be conservative. Rather than identify every island that life may or may not inhabit in parameter space, we can just note the huge lifeless oceans.

“The best-understood cases of fine-tuning are too dramatic to think that nit-picking over the definition of life would make any difference. Carroll’s point is essentially appealing to an as-yet-unknown fact about life that will hopefully reveal why, against all appearances, it could form and survive in a wide range of universes. In the absence of any specific idea about what this unknown fact might be, it is just as likely that what we don’t know about life will make it rarer in possibility space, i.e. more fine-tuned than we think.”

As Paul Davies says, the universe “is fine-tuned for the building blocks and environments that life requires.”

4. Universes with totally different structures

Here we are starting to get very speculative. But I think there are good reasons why this argument isn’t convincing.

Luke Barnes has addressed this question briefly in s4.1.3. He lists 6 different laws that have been suggested (for example, reversing the electromagnetic or gravity forces, no quantum physics at small scales, or making electrons bosons rather then fermions), but says that in every case, such a universe would either be impossible, or unable to support any form of life. He concludes that “that region of possible-physics-space contributes negligibly to the total life-permitting subset”, and thus tends to confirm fine-tuning.

Most changes we can think of are based more or less on the physics of our own universe, but with some twist (as in the examples Barnes mentioned), and the effects of these changes can in principle be assessed using theoretical physics. But what about more “exotic universes?

For example, we might say “what about a universe not made of electrons and protons?”. But if we are talking science, as we are here, we need to follow the scientific method, which requires a hypothesis that can be tested and possibly falsified. Fantasy universes don’t really fulfil that requirement.

I have been pondering this question too, and I think we can see that, mathematically, alternative universes don’t help the anti-fine-tuning argument. The sample space of known physics is very large and the life-permitting region within it is very small. Let’s just consider Lee Smolin’s estimate (p 45) of the probability of stars forming randomly (=1 in 10^229) as being the probability of life, and suppose that we could accept a level of fine-tuning that was just 1 in a billion (10^9). Even if there were a billion other forms of physics, each with its own sample space of randomly possible universes, every one of those alternative physics would have to have a probability of permitting life of less than 1 in a billion for the final probability to fall within our acceptable range. In other words, in a billion other types of physics, the probability of life is around 1 in 10^9, but just in ours it is 1 in 10^229. That doesn’t sound like a reasonable explanation to me!

So it seems that, if we are talking science, totally alternative universes don’t negate fine-tuning, and perhaps tend to confirm it.

5. The multiverse

The one possibility that seems able to explain fine-tuning, and the one adopted by many cosmologists, is the multiverse – there are a multitude of universes (or separate domains within the one huge universe), each of which has randomly determined parameters, and of course intelligent observers (that’s us, in case you didn’t recognise that description!) could only appear in one of the few that could support life. Lee Smolin suggests a slightly different scenario, but it seems to my inexpert eye to amount to a similar result.)

Even proponents of the idea agree that it is still speculative and faces many difficulties before it could be accepted as “science”. Some believe it can never be science because alternative universes cannot be observed, but others argue that the multiverse idea grows out of established physics, and may perhaps one day be regarded as a logic inference. (Leonard Susskind even suggests science may have to abandon the principle of falsifiability!)

But this isn’t relevant to the argument here. The multiverse isn’t established science (yet), and even if it was, it confirms that our universe is fine-tuned, but offers a possible explanation. I’ll consider the multiverse again in my next post.


The conclusion seems inescapable. Sean Carroll’s statements notwithstanding, current science has found that “of all the possible universes allowed by theoretical physics, an extremely small number would allow the evolution of intelligent life”. Fine-tuning appears to be a conclusion of modern cosmology that most accept.

Whether that scientific conclusion offers support for a theistic argument is a much harder question to assess, and I’ll look at it next post. Please check it out in a week or so.

Picture: Andromeda galaxy (NASA).


  1. Hey Eric,
    I’m interested to know your thoughts on another possible objection. It’s kind of a reverse of the probability objection you noted.

    It basically asks whether the presumption behind fine-tuning – that the constants could be something other than what they are – is valid. Could it be that by describing the regularity of the universe in mathematical terms we have created the illusion that things might be otherwise by simply tweaking some of the numbers in those approximations, when in fact the fundamental nature of reality is immutable? Or, in even more abstract terms, maybe the laws aren’t “number-like” and casting them in terms of variable values is misleading in this respect? What is the reason to think that there is a probability space at all? Is it just that this is the default assumption for everything and so we apply it here as well?

    I’m not proposing an answer, just wondering whether it’s something that you’ve seen discussed or considered yourself.

  2. Hi Travis, obviously I don’t know enough to say, but I have read a few comments on that by cosmologists.

    Martin Rees in his book Just Six Numbers offers 4 possible explanations of fine-tuning – providence (i.e. God), coincidence (i.e. chance), necessity (outcomes of a theory of everything) and the multiverse. He rejects coincidence (i.e. the odds are so far against it) and he rejects necessity (which is what you are mentioning here) because he can see no way it can be true.

    His reasoning (from memory) is that it is easy to imagine different values in the equations, and the numbers don’t look like they are all fixed. So we would need a theory of everything that shows why the numbers and laws actually are fixed, even though they don’t look like they are, but we have nothing like a theory that could satisfy that.

    Luke Barnes also addresses the question, and says that so far nothing looks like being able to tie everything together and make our universe necessary. He says the current best hope for a theory of everything, String Theory, doesn’t do it, and looks very anthropic.

    From my reading, I think all cosmologists hope there may be a theory of everything one day, but I think few are thinking it will happen. Of course if it does, then the whole matter will have to be revised.

  3. Thanks. Reminds me of an old comedy sketch. We have in Australia a partially Government sponsored TV network, SBS (Special Broadcasting Service), that caters for minority ethnic groups – many Australians were born overseas and speak a language other than English at home.

    SBS has a reputation, not fully justified, for showing obscure films on obscure topics from obscure countries. And so a comedy sketch (on another channel) many years back had a woman announcer talking up the shows coming up that night, one of which was a film which was something like “Surprise for the Goatherd. In this film from Transylvania, Zlatko the goatherd ventures down in the valley to find a new ribbon for his favourite goat. [pause, then in thick accent ….] Can’t wait for that one.”

    So my response to Sam – can’t wait for that one! 🙂

  4. Hi Abby, I haven’t watched that video, I’m sorry. I just don’t have a lot of time to watch 71 minutes, and for that reason I rarely watch long videos.

    But I can say I have seen very similar arguments before, and I have seen other people say that numerical patterns can be found in all sorts of writings. I would need to see a mathematician compare several different texts before I could assess this sort of thing.

    I’m sorry, but that is as much as I can say.

Comments are closed.