I find that the gauge symmetries of the lagrangian are a topic that gets obfuscated quite a bit. I'm trying to understand the big picture of this in QED. My understanding is that:

  1. Gauge derives its name ultimately from that thing that I call the
    classical gauge transformation of classical electromagnetism: $\mathbf{A} \rightarrow \mathbf{A}' + \nabla f$.
  2. In QED, the spacetime equivalent of this is $A_\mu \rightarrow A_\mu + \frac{1}{g}(\partial_\mu \Lambda)$.
  3. Basic spacetime symmetry requires us to have global symmetry, for which we have $\phi \rightarrow e^{i \theta} \phi$.
  4. Other spacetime symmetries require us to have local symmetry, for which we have $\phi \rightarrow e^{i \theta (x)} \phi$.
  5. You have to have a gauge covariant derivative, thus we have $\partial_\mu \rightarrow \partial_\mu - igA_\mu$.

Now my question is this: are any of the above gauge symmetries & derivatives (that is, 2 through 5) redundant? For example, does rule 5 above actually the same as rule 2?


1 Answer 1


The 2) is Lorentz invariant way to write the 1).

The 5) is a way in which one has to modify derivative in order to make the theory gauge invariant. For the simple case of complex scalar field $\phi$ with $U(1)$ gauge symmetry, the lagrangian looks as follows: $$ \mathcal{L} = \partial_\mu \phi^{*} \partial^{\mu} \phi + \ldots $$ Applying an infinitesimal position-dependent transformation $\phi \rightarrow e^{i \theta}\phi$, one gets in the lowest order: $$ \delta \mathcal{L} = i \partial_\mu \theta (\phi^{*} \partial^{\mu} \phi - \phi \partial^{\mu} \phi^{*}) $$ So to cancel this change in the Lagrangian one needs to promote $\partial_\mu$ to $\partial_\mu + i A_\mu$, with the aforementioned transformation properties : $A_\mu \rightarrow A_\mu + \partial_\mu \theta$. So that the simultaneous transformation of both fields would leave the Lagranigian invariant.

About 3) and 4) - one should distinguish between global and gauge symmetry, because they have different meaning. The global symmetry - is a set of transformations, that leaves our theory invariant, and the gauge symmetry, as is often said, is not actually a symmetry, but a redudancy in description of the theory.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.