Today, more on the work of William Lane Craig on the fine-tuning of the universe for intelligent life. The issue today is whether the laws/constants/initial conditions of the universe are, in some way, necessary. We think that the probability of a randomly chosen universe (with its laws, constants and initial conditions) being life-supporting is vanishingly small. We reach this conclusion by altering the laws/constants/initial conditions and predicting the outcome.
But perhaps when we have a deeper understanding of the laws of nature, we’ll realise that these constants couldn’t have been different. Or at least, we’ll realise that many of them are related, and thus cannot be altered independently. This would significantly reduce the probability of “getting the universe right”, as there are fewer dials to be tuned.
Let’s have a look at Craig’s response to this argument:
Physical necessity is the hypothesis that the constants and quantities had to have the values they do, so that the universe is of physical necessity life-permitting. Now on the face of it this alternative is extraordinarily implausible. It requires us to believe that a life-prohibiting universe is physically impossible. But surely it does seem possible. If the primordial matter and anti-matter had been differently proportioned, if the universe had expanded just a little more slowly, if the entropy of the universe were marginally greater, any of these adjustments and more would have prevented a life-permitting universe, yet all seem perfectly possible physically. The person who maintains that the universe must be life-permitting is taking a radical line which requires strong proof. But there isn’t any; this alternative is simply put forward as a bare possibility.
Sometimes physicists do speak of a yet to be discovered Theory of Everything (T.O.E.), but such nomenclature is … quite misleading. A T.O.E. actually has the limited goal of providing a unified theory of the four fundamental forces of nature, to reduce gravity, electromagnetism, the strong force, and the weak force to one fundamental force carried by one fundamental particle. Such a theory will, we hope, explain why these four forces take the values they do, but it will not even attempt to explain literally everything.
For example, in the most promising candidate for a T.O.E. to date, super-string theory or M-Theory, the physical universe must be 11-dimensional, but why the universe should possess just that number of dimensions is not addressed by the theory … All this has been said with respect to the constants alone; there is still nothing to explain the arbitrary quantities put in as boundary conditions.
Craig here is responding to some comments by Richard Carrier, some of which are pretty daft. Here’s Carrier: “in the 19th century there were some twenty to forty ‘physical constants,’ there are now only around six”. So Carrier’s physics credibility is non-existent.
Craig makes three main points:
- These other universes seem to be possible.
- A Theory Of Everything (TOE) won’t literally explain everything e.g the boundary conditions of the universe.
- String theory is our best candidate TOE, and even it doesn’t uniquely predict the constants of nature.
My central criticism of Craig concerns his first point – it is far too weak. Scientific discoveries are surprising precisely because they are unexpected. Consider the following three constants: the electric constant, the magnetic constant and the speed of light. Prior to 1865, these constants seemed completely unrelated. Then Maxwell showed that light is an electromagnetic wave, and in so doing showed that these three constants are related.
So the constants of nature will always seem to be changeable. And, most importantly, no physicist believes that we have heard the last word on what we now think are the fundamental constants. To dismiss necessity, we need reasons to believe that, even as physics progresses, the laws of nature will still need to be fine-tuned for life.
Craig’s third point is similarly unconvincing. String theory is certainly a work in progress – we don’t even know the exact equations. The failures of string theory as a TOE should lead us to look for a better TOE, not conclude that no TOE will predict all the fundamental constants of nature.
Craig’s second point is starting to get somewhere. He is correct in that a TOE won’t explain literally everything. In particular, the initial conditions (or, more generally, boundary conditions) of the universe are a worry. Here’s a quote from John Wheeler that should keep cosmologists awake at night:
Never has physics come up with a way to tell with what initial conditions the universe was started off. On nothing is physics clearer than what is not physics: equation of motion, yes; initial position and velocity of the object which follows that equation of motion, no. (At Home in the Universe)
The problem is more general than just initial conditions. Even if the equations that describe our universe are unique, containing no free parameters, it doesn’t follow that the solution to the equation, and thus the universe itself, is unique. For example:
The equations of [string theory] have no adjustable constants, but their solutions, describing different vacuum states, are characterised by several hundred parameters-the sizes of compact dimensions, the locations of the branes, and so on. (Alex Vilenkin, “Many Worlds in One”)
The closest thing in physics to a “law of initial conditions” is the Hartle-Hawking no-boundary proposal. And if I had any idea what Hartle and Hawking were talking about, I’d tell you.
Craig’s objections aside, the idea that the fine-tuning of the universe can be explained by physical necessity hasn’t won many supporters. It’s not my preferred scenario. Let’s look at a few case studies to see what goes wrong.
A. An unknown physical process sets the constants/initial conditions to be in the life-permitting range. A good example here is inflation. In standard big bang cosmology, the matter density at early times needs to be fine-tuned to one part in , otherwise it will be too dense (and recollapse into a big crunch) or not dense enough (and be unable to form galaxies/stars/planets/people). However, if the universe inflates (expands exponentially) in its early stages, then the density of the universe will be driven into the life-permitting range. Then we wouldn’t need special initial conditions – any old universe can be made to have a life-permitting density by inflation.
Could this type of scenario be true for all the other constants/initial conditions? I don’t think so, for the following reason. Inflation makes the universe life permitting by driving the density towards the “critical density”. The critical density can be specified independently of the requirements for life – it is the density that makes the universe spatially flat. It is then a very fortunate coincidence that life-permitting universes need to be very close to spatially flat. If this were not so – e.g. if life permitting universes needed to be 100 times less dense than critical – then inflation would make every universe life prohibiting.
So the success of inflation in making the universe life permitting relies on a coincidence between a life permitting and an independently given physical parameter. This would seem to be true for any physical mechanism that hopes to make the universe life permitting. But for every physical constant/initial condition to have an explanation of this type would require a multitude of as yet unknown physical parameters, which by some coincidence happen to lie in the life permitting range, and a physical mechanism that drives the actual constant toward this new parameter. This seems incredibly unlikely, and even if it were true, it would simply be another astonishing example of fine-tuning.
B. A deeper understanding of physical laws reveals that many/most/all the constants are related. Perhaps there is only one constant of nature, which itself sets the scale of the universe and is thus unobservable. Other physical constants turn out to be mathematical constants. Let me make three quick responses.
- Suppose string theory achieves this goal. Even if there no alternatives within string theory, there are alternatives to string theory. Paul Davies says it well: “I think [the idea that there is only one possible universe] is demonstrably wrong. There is not a shred of evidence that the universe is logically necessary. Indeed, as a theoretical physicist I find it rather easy to imagine alternative universes that are logically consistent, and therefore equal contenders for reality”.
- The problem of parameters in the solution, noted above by Vilenkin. It could turn out that what we think of as parameters of the laws of nature are simply parameters of the solution that describes our universe. They would, then, remain fine-tuned, since other solutions would presumably not be life permitting.
- It would be the mother of all coincidences that the only universe permitted by the laws of nature would happen to permit intelligent life. In their classic paper on fine-tuning in 1979, Carr and Rees comment that “even if all apparently anthropic coincidences could be explained [by deeper physical law], it would still be remarkable that the relationships dictated by physical theory happened also to be those propitious for life.”
Let’s sum up. I agree with Craig that necessity isn’t the right explanation for the fine-tuning of the universe for intelligent life. I have an issue, however, with some of his arguments is favour of this conclusion. In the end, the idea that life-permitting universes are in some sense not possible seems incredibly implausible. As Craig says, we would want some very good reason or believing that.
More of my posts on fine-tuning are here.