An interesting and well-written book. Smolin started out wanting to write about the sociology of research funding in the US. He is extremely worried about the fact that it has become difficult for young researchers to get money to pursue novel ideas, with most funding concentrated on a small number of mainstream projects which are regarded as "safe". In many fields, this has already been taken to the logical extreme, with nearly everything focussed on one single direction. As a researcher (albeit in a completely different discipline), I have had good reason to make similar observations. It is indeed worrying.
Smolin's publisher, however, was concerned that a general book on this theme would be too dry, and encouraged Smolin to organise it around a specific topic that he knew well. Smolin agreed, and used string theory. So now you essentially get two books for the price of one: a very interesting, up-to-the-minute, critique of string theory, written by an insider who knows where all the bodies are buried, and a general treatise which takes this as an example of what's going wrong with the way research is organised. Both are very successful. If you are at all interested in these topics, it's a must-read.
The part I liked best was the historical section, where Smolin examines the claim string theorists like to make that "it's so beautiful that it simply has to be true". He comes up with some great examples of beautiful theories that actually just happened to be plain wrong.
_________________________________________
I just finished rereading this excellent book. There's so much interesting material that it's easy to miss many of the best bits first time around! So here are some of the things that most impressed me on my second visit:
The Kaluza-Klein theory
It's not well-known (or at least, I had never heard it before), but the idea of using higher dimensions to unify gravity and electromagnetism is far from new. Kaluza, Klein and some other smart people came up with a scheme of this kind about the same time as Einstein invented General Relativity. The theory was very elegant and beautiful. But, unfortunately, the experimental predictions it made turned out to be incorrect, so it was junked.
Lack of experimental predictions from string theory
The problem is that "string theory" isn't actually a single theory. It's more a research program based on a some extremely complicated mathematics that only a few experts understand. There are many different versions of it, and all of them have multiple adjustable parameters, so it predicts very little. Hence it's extremely hard to disprove it.
What happened to string theory when dark energy was discovered
One of the very few predictions made by string theory was that the cosmological constant had to be zero or negative. Then, in 1998, astronomers discovered dark energy, which appears to point to a positive cosmological constant (the dark energy force is stretching the Universe apart, rather than pulling it together). Under normal circumstances, one would just have said that this showed string theory wasn't correct.
But the string theorists found an even more complicated way to rejig the theory and get a positive constant. The new theory comes in about 10^500 different forms (1 with 500 zeros), so now it really is almost impossible to test it experimentally. The string theorists counter by saying that there is a "Cosmic Landscape", by which they mean there are 10^500 different universes, one for each version of the theory. Then they invoke the "anthropic principle" to explain the appearance of the world we see by arguing that we're observing it, hence we must be in one of the few universes that can support life. As Smolin says, this is not a normal way to do science.
Some weird shit that astronomers have been finding
String theorists like to say that they're the only game in town. Smolin reminds us that we don't actually have to spend all our time playing in this sterile mathematical wonderland. There is some amazing stuff turning up in real-world observations and crying out to be explained.
First, he encourages us to think more about dark energy. As far as we know, it's not like it's in any particular place. It seems to be a property of the whole Universe. Physicists like to take about what happens at different "scales" - the scale of subatomic particles, the scale of molecules, the scale of planets, etc. Different forces are at different scales: for example, quantum effects are really important at the subatomic scale but not important at all at the planetary scale, while gravity is the other way around.
So he says, well, let's call R the size of the Universe, about 10 billion light years. Dark energy appears to be operating at the scale of R. Is there anything else? Let's try deriving some other physical quantitities on that scale. In particular, let's look at R/c^2, R divided by the square of the speed of light. That's an acceleration, which works out to about 10^-8 cm/s^2. It's a very small acceleration indeed. Is anything interesting happening with accelerations at this level?
And, it turns out, there is! Stars near the outskirts of a galaxy should be accelerating at just about R/c^2. But, in fact, they're moving faster than they're supposed to, while stars near the centre of the galaxy move at the predicted speed. This has been well known for decades, and is generally ascribed to the existence of a "dark matter halo" - invisible mass which only interacts through gravity, and is spread out around the outskirts of the galaxy. Suppose, though, that dark matter and dark energy were different aspects of the same thing?
Not only that, but there is some odd data coming from the Pioneer 10 and 11 probes. These are now leaving the Solar System but can still be tracked. It turns out that they appear to be slowing down more than they should, with the difference between the observed accelation and the predicted one being around 6 times R/c^2. People have been checking the figures carefully, trying to find normal explanations, and so far nothing has emerged.
He also has some very cool stuff about cosmic rays, and how you can in effect use the whole Universe as a particle accelerator to do experiments. But let me move on to my last topic.
Seers and craftspeople
The last third of the book is about sociology. Why has physics got itself into this mess?
Smolin's answer is quite long and complicated, but let me give you one part I particularly liked. He distinguishes between two kinds of periods in science, "revolutions" and "normal science". Revolutions are when things abruptly change; the last big one was at the beginning of the twentieth century. Normal science is what happens in between, where people work out all the consequences of the last revolution.
He also distinguishes between two kinds of scientists, whom he calls "seers" and "craftspeople". Craftspeople are very smart, very focussed, very good at math. They were top of their science class all through school. They learn existing techniques quickly and are good at using them. Seers are more unfocussed, mystic, dreamer types. They're often not that great at math, but they can't stop thinking about the fundamental questions. As he says, it's easy to be dismissive of seers until you remember that Einstein was like that.
Smolin thinks we've forgotten why we need seers. He says we're trying to organise a scientific revolution using craftspeople, and it's not working. He's an inspiring guy.