I just finished Peter Woit‘s book Not Even Wrong, about why he’s convinced String Theory is not really science. His book is more than that, it’s a wonderful romp through the history of 20th Century physics and the deep importance of mathematics to the development of quantum mechanics — not to mention his championing of the great mathematician Herman Weyl, the unsung hero of this story — with detailed explanations of the core principles that went into everything from QED to the Standard Model, etc. Very well done.
At the end he rightly critiques String Theory for not producing falsifiable hypotheses — ones that can be tested, and potentially proven wrong. Like Woit, and most hard-scientists, I believe the general principle of Falsifiability of Karl Popper’s is the defining characteristic of a scientific theory vs every other kind of speculation humans can make. While one can quibble about details (rightly) regarding exactly HOW to falsify one theory or another, the essence of being able to draw a hard line in the sand between science and non-science is of fundamental importance. Whether something is good science or not is far less important than if it is simply not science at all.
My own view regarding String theory is that it is properly placed within the realm of mathematical philosophy (note: I did NOT say “philosophy of mathematics”!), that is, it is a philosophical quest aided by mathematical tools.
Who knows, someday it may turn into either true mathematics (positing theories and proving them) or physics (positing theories that are falsifiable).
So far, that hasn’t happened.
If I was in the world of physics, I might be more concerned about this, as theoretical physics has been swamped by string theory, and so many of the brightest minds are being sucked up into it. That isn’t all bad, except that it means by definition that those minds are not being let loose on other problems that are testable in physics, and need work.
Reading between the lines, I take it that Woit is suggesting is that maybe the only sensible thing to do is to move String Theory over into the philosophy departments until it produces some testable theories, and leave open the funding for physics that is… physics.
Here’s a video with Peter Woit at Big Think talking about this topic. It includes a snippet on what mathematicians can teach physicists, among other cool things: