perturbation theory

References

Summary

I'd thought that all perturbation theory was about getting particles to high energies in order to reduce their interactions — that's actually not true — the theories for which this works, like QCD, are called asymptotically free.

QED, on the other hand, actually increases interactions with increasing energy; this is really surprising to me; but the first reason for my misconception is that it increases really slowly — so even at the highest energies humans will ever reasonably be able to make, it's still perturbative. I knew that QED was perturbative at high energies, so I must have just assumed that it was also asymptotically free.

The other reason for the misconception is that I'd thought you could just increase the kinetic energy of a particle, without increasing its interaction energies. So for instance, if we take a simple Lagrangian like $$ {L}(\phi) = \partial^{\mu} \phi \partial_{\mu} \phi - m^{2} \phi^{2} - \dfrac{\lambda}{4!} \phi^{4} $$

I'd figured you could just speed up the particle, increasing the first two terms, and making the λ term small in comparison, and then use it as a perturbation.

As it turns out this stops working when you quantize — in the classical theory, λ is just a fixed number. But once you quantize, it becomes something dependent on energy. This feature is called a quantum anomaly and is related to things like Landau poles and renormalization.