I have seen distinct definitions of gauge covariant derivative (in Yang-Mills theory)

$$ D_\mu \phi = (\partial_\mu + igA_\mu) \phi $$


$$ D_\mu \phi = \partial_\mu \phi + ig[A_\mu,\phi] .$$

I guess the first is the common definition, which is the same as the covariant derivative in QED. What is about the second?

  • 2
    $\begingroup$ The notation $A_\mu \phi$ is a placeholder for "whatever you get when you let $A_\mu$ act on $\phi$, given that they both live in representations of the gauge group". In the usual case it's the commutator (though it doesn't have to be) so the second notation is more explicit. $\endgroup$
    – knzhou
    May 24, 2017 at 15:03
  • $\begingroup$ The first one is incorrect in case of Yang Mills (see answer below). It is strictly for QED. $\endgroup$
    – DanielC
    May 24, 2017 at 15:09
  • $\begingroup$ Related: physics.stackexchange.com/q/514399/2451 $\endgroup$
    – Qmechanic
    Apr 13, 2020 at 6:04

1 Answer 1


As AccidentalFourierTransform points out, the second expression is the non-abelian generalization of the former. The first one is only valid for the abelian case (QED), while in general Yang Mills the fields are matrices transforming in some representation of the gauge group and the correct form is the latter.


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.