〈 Berger | Dillon 〉

Unfortunately making non unique predictions is perhaps more of a hurdle then making unique wrong ones

Yes, string theory is drowning in models, which people are now cranking out at an industrial scale. I don't remember anything quite like this before in the history of physics. It seems like something out of a story by Stanislaw Lem.

Exercise: Compute in plain QFT the number of consistent models with exactly the same field content as the standard model. Do you get a finite number? If not, do you get at least a finite-*dimensional* space of choices? If not, reflect on what's going on.

Urs wrote: "Compute in plain QFT the number of consistent models with exactly the same field content as the standard model." In the old days physicists chose one theory, made predictions, and considered it a success if they were right. They did not choose all possible theories.

By one theory, do you mean a continuum space of theories plus an organizing principle that goes by the name of renormalization group, without which successful predictions would be accidental?

If you want to call QED a continuum of theories, be my guest. It's true that I don't know every digit of the fine structure constant or the electron mass (as measured at some low energy scale in the usual way). I still call it one theory.