It’s always a nervous moment when, as a scientist, you discover that a documentary has been made on one of your favourite topics. Science journalism is rather hit and miss. So it was when the Australian Broadcasting Corporation (ABC), our public TV network, aired a documentary about the fine-tuning of the universe for intelligent life as part of their Catalyst science series. (I’ve mentioned my fine-tuning review paper enough, haven’t I?).
The program can be watched on ABC iView. (International readers – does this work for you?). It was hosted by Dr Graham Phillips, who has a PhD in Astrophysics. The preview I saw last week was promising. All the right people’s heads were appearing – Sean Carroll, Brian Greene, Paul Davies, Leonard Susskind, Lawrence Krauss, Charley Lineweaver. John Wheeler even got a mention.
Overall – surprisingly OK. They got the basic science of fine-tuning correct. Phillips summarises fine-tuning as:
When scientists look far into the heavens or deeply down into the forces of nature, they see something deeply mysterious. If some of the laws that govern our cosmos were only slightly different, intelligent life simply couldn’t exist. It appears that the universe has been fine-tuned so that intelligent beings like you and me could be here.
Not bad, though I’m not sure why it needed to be accompanied by such ominous music. There is a possibility for misunderstanding, however. Fine-tuning is a technical term in physics that roughly means extreme sensitivity of some “output” to the “input”. For example, if some theory requires an unexplained coincidence between two free parameters, then the “fine-tuning” of the theory required to explain the data counts against that theory. “Fine-tuned” does not mean “chosen by an intelligent being” or “designed”. It’s a metaphor.
Ten minutes in, the only actual case of fine-tuning that had been mentioned was the existence of inhomogeneities in the early universe. Sean Carroll:
If the big bang had been completely smooth, it would just stay completely smooth and the history of the universe would be very, very boring. It would just get more and more dilute but you would never make stars, you would never make galaxies or clusters of galaxies. So the potential for interesting complex creatures like you and me would be there, but it would never actually come to pass. So we’re very glad that there was at least some fluctuation in the early universe.
Paul Davies then discussed the fact that there not only need to be such fluctuations, but they need to be not-too-big and not-too-small. Here’s the scientific paper, if you’re interested.
The documentary also has a cogent discussion of the cosmological constant problem – the “mother of all fine-tunings” – and the fine-tuning of the Higgs field, which is related to the hierarchy problem. Unfortunately, Phillips calls it “The God Particle” because “it gives substance to all nature’s other particles”. Groan.
Once we move beyond the science of fine-tuning, however, things get a bit more sketchy.
Leonard Susskind opens the section on the multiverse by stating that the multiverse is, in his opinion, the only explanation available for the fine-tuning of the universe for intelligent life. At this point, both the defence and the prosecution could have done more.
Possibilities are cheap. Sean Carroll appears on screen to say “Aliens could have created our universe” and then is cut off. We are told that if we just suppose there is a multiverse, the problems of fine-tuning are solved. This isn’t the full story on two counts – the multiverse isn’t a mere possibility, and it doesn’t automatically solve the fine-tuning problem.
There are cosmological theories, supported by some evidence, that point towards a multiverse. Cosmological inflation is a theory about the very early universe which makes some successful predictions – the density of the universe, homogeneity and isotropy, the existence (though not the magnitude) and type of density fluctuations. Inflation is an effect, not a cause, and so cries out for a physical theory to explain how it happens. Some of the most plausible theories of inflation predict that, beyond the observable universe, conditions may be different, and there may be enough variation to explain the fine-tuning of the universe.
The multiverse is thus more than a mere possibility. There is a connection, though not a particularly direct one, with observationally tested theories.
There is also more to say in critique of the multiverse. Suppose the RandomAutos car factory makes cars by randomly arranging metal and plastic, and then testing to see whether the arrangement could be driven 1 km within 2 minutes. If not, scrap it and try again. Suppose you bought one of their cars. It would be no surprise if you could drive the car around. It would, however, be shocking if you got a working sat nav. Selecting for drivability makes extra features no more likely. This is true even though, if enough cars were made in that way, a car with a sat nav would happen sooner or later.
The multiverse can only explain what the selection effect selects for. The selection effect is the anthropic principle: observers must observe a universe compatible with observers. Therein the problem lies, because observers don’t require a fine-tuned universe. Freak observers can bypass the requirements for biological life – stable stars, planets, atoms, chemistry etc – and simply fluctuate into existence. Most of the multiverse will be inhospitable to biological life, but perfectly fine for freak observers. If the multiverse contains many more freak observers than biological lifeforms, then this galaxy-star-planet-biochemistry setup we’ve got going in this universe is like the sat nav in the RandomAutos car. This universe is fine-tuned for biological life, but the multiverse can only explain observers, and so plausibly fails to explain the fine-tuning of the universe. This is true even though biological life would appear somewhere in the multiverse.
This is where the field is at. For more discussion of these ideas, see this post by Sean Carroll and his talk here, as well as this paper by Robin Collins. (I’m not really that surprised that the documentary didn’t cover these ideas. The multiverse is confusing enough as it is.)
The documentary starts off by arguing quite cogently against the idea that fine-tuning could simply be a coincidence. Brian Greene has us consider a child discovered with cookie in hand, who claims that the cookie simply fell into his hand. Such an explanation, he says, would be rejected. It requires too much fine-tuning, in that at just the right moment, the jar must tip in just the right way to make the cookie fall on just the right trajectory to land in a hand that happened to be in just the right place. Also, an alternative hypothesis is available, namely that the child took the cookie. The fine-tuning of the universe, we are told, is too improbable to be coincidence and so demands an explanation. Hence, Leonard Susskind can say that the multiverse is the only option.
Towards the end, however, the host returns to coincidence as a live option.
Fine-tuning could be nothing more than a coincidence. The way our human minds naturally work is when we see a coincidence, we straight away think something deep is going on. So instance, if four of my Australian friends suddenly turned up here, on this Californian beach, I wouldn’t think “that’s a fluke”. I think “wait a minute, there’s something going on here. Maybe they’re planning a surprise party for me”. But that very instinctive way of thinking we all have can be very misleading.
I’m getting a bit sick of these pseudo-psychological explanations. Whenever you’ve run out of arguments, just make up some just-so story about how there is a natural tendency for humans to believe such things for some made up reason and we’re all biased of course so (condescending head pat) don’t believe that, sonny. C.S. Lewis called it Bulverism, Flew called it the subject/motive shift: distract attention from the truth of a claim by explaining how some people have come to believe something so silly.
The thing to do with a perceived bias is to correct for it, not to overcorrect by surrendering to the opposite conclusion. One should think more, not less.
A coincidence is a hypothesis. Specifically, it is a hypothesis wherein a number of events which happen to coincide do so without any deeper connection. For example, the five Australians on a beach in California is a coincidence if the individual reasons why they are each there don’t involve the fact that the others will be there. Contrast this with the surprise party hypothesis, in which the four people are there because of the fifth. We evaluate these hypotheses in the same way we evaluate any hypothesis – how likely is the hypothesis before we discover the five people on the beach, and how likely it is that these five people would be on the beach assuming that the hypothesis is true? Considering and weighing alternative hypotheses is exactly what we should do, and it’s what we instinctively do. The Bayesian account doesn’t contradict our instincts. It agrees with them, and makes them more precise.
I’m not convinced that there is a tendency for the human mind to think that every coincidence points to something deeper. We notice coincidences all the time without grabbing our Sherlock Holmes pipe and launching an investigation, let alone believing a conspiracy theory. It depends how unlikely the coincidence is, and whether an alternative, plausible explanation is at hand. (If you want a mind-blowing coincidence, read this. Unlikely in the extreme, but any deeper explanation would be even more implausible.)
Fine-tuning and The Fine-Tuner
Says Phillips, against some foreboding music:
Some take fine-tuning as evidence that God created the universe. You can imagine physicists horror at the thought.
I watched that clip today in a room full of astrophysicists. We all burst out laughing when he said that. Not all atheists are atheists because God horrifies them. Not all scientists are atheists. One can be a scientist and a theist. Most of the makers of the scientific revolution were.
The scientists interviewed get a bit sloppy at this point. Lawrence Krauss says:
It’s certainly fine-tuned so that we can exist. It’s also incredibly inhospitable. If you were designing a universe for life, I suspect you might design it differently. There is no evidence of design or purpose in our universe.
That’s a logical fallacy.
- Relative to some background information B, a piece of evidence E increases the probability that a hypothesis H is true if and only if E is more likely to be true if H is true than if H is not true.
- Suppose that E is very unlikely to be true if H is true.
- Thus, E has made H less probable
- Thus, E has made H very improbable.
1 follows from Bayes theorem (homework). However, neither 3 nor 4 follows from 1 and 2. In the absence of any information on the probability of E given not-H, we cannot use 1. 2, on its own, is not sufficient. 4 would also require knowledge of the prior, , so it doesn’t follow either.
Krauss tries to argue that the probability of observing a universe like this one given that God exists is somewhat small; call this the likelihood of the universe given God. His argument is basically that he wouldn’t have done it this way, which isn’t particularly strong. It depends on how much one’s god-complex mirrors reality.
Even granted that premise, it isn’t true that the probability of God is small, or even that this universe is not evidence for God, if the likelihood of the universe given God is small. What matters in not the size of the likelihood but the comparison between the likelihood of the universe given God and the likelihood of the universe given not-God. We must also ask: how likely is it that a fine-tuned universe would exist given that God doesn’t exist? If that is anything like the probability of hitting a life-permitting universe “at random” (indifferently, to be more precise), then it is extremely small. A minor reduction in the likelihood of God because it’s not how Lawrence Krauss would have made a universe hardly matters.
For example, suppose our host sees 250 friends and family on the beach, with a giant sign reading “Surprise! Happy Birthday, Graham!”. He thinks, “well, if I were throwing me a surprise party, I might not throw it here”. While that does somewhat decrease the probability that this is a surprise party for him, that decrease is negligible compared to the improbability of all those people and signs being there by accident. Thus, the “surprise party” hypothesis is still the best hypothesis.
Sean Carroll hits the nail on the head: “Is God the best explanation?”, he asks. Unfortunately, having asked the right question, he isn’t given time to answer it. His answer can be found here.
The fact that much of the universe is inhospitable isn’t a particularly strong argument against God, either. John Leslie: “As if the only acceptable sign of a universe’s being God-created would be that it was crammed with living beings from end to end and from start to finish!”. I have heard the cosmology of the Middle Ages criticised for being too “pokey”, too small, compared to the glory of the universe shown by modern astronomy. If a huge universe is more glorious, more beautiful, more awe-inspiring, then there is all the more reason for a God to make such a universe. Also, there is a connection between the size of the universe and life. We require stars to make the elements out of which we are made. We also need them to be long-lived to give energy. Stars need time. So the universe will need to be old, and old universes must be big and mostly empty, otherwise gravity will make them recollapse. A closed universe with a mass equal to our galaxy recollapses within a month. Space isn’t a waste of space. This isn’t to say that God must make a big universe like ours. But it does have its attractions.
A few more quibbles
Phillips says that the Earth has the advantage of a protective giant planet Jupiter to prevent rogue asteroids from hitting us. Debatable.
The documentary considers John Webb et al’s evidence that the constants of nature (specifically the fine-structure constant) actually vary from place to place. This is taken as evidence that the laws of physics are not set in stone. Webb’s case has not convinced everyone, to put it mildly. It doesn’t show that the laws of nature vary, only that what we had previously believed to be a constant of nature was actually a field which varied from place to place.
More importantly, if the fine-structure constant (“alpha”) is actually a field, then this creates a far greater fine-tuning problem than it solves. The strongest life-permitting limit one can place on alpha is one part in , related to the triple-alpha resonance (see my paper). If alpha is actually a field, then the mass scale associated with that field is about times smaller than the Planck scale. This would be another hierarchy problem, and much worse than the problem for the Higgs field. Out of the frying pan …
Not bad. A reasonable introduction, some science well-explained, an understandable wander into the weirder implications of the field, and the right people being interviewed. Most of my criticisms above are about understandable omissions or the views of the interviewees. Well-worth half an hour.