If we require the reality condition $\phi=\phi^*$ on the Lagrangian for a complex scalar field is $$\mathcal{L}=(\partial^\mu\phi^*)(\partial_\mu\phi)-m^2(\phi^*\phi),$$ two degrees of freedom $\phi$ and $\phi^*$ is reduced to one. For consistency, I would expect it should give $$\mathcal{L}=\frac{1}{2}(\partial^\mu\phi)(\partial_\mu\phi)-\frac{1}{2}m^2\phi^2.$$ But this procedure misses the $1/2$ factors. Why? Did I mess up some normalization?


1 Answer 1


The standard convention is to divide each term in the Lagrangian with its symmetry factor. Therefore the kinetic term for a real (complex) scalar field is with (without) a symmetry factor $\frac{1}{2}$, respectively. A complex scalar field $\phi= \frac{\phi^1+i\phi^2}{\sqrt{2}}$ can be viewed as 2 real scalar fields.

  • $\begingroup$ Not sure I completely follow. I know that the Lagrangian of the sum of two real scalar fields $\phi_1$ and $\phi_2$ is equal to the Lagrangian of a complex scalar field defined as $\phi=(\phi_1+i\phi_2)/\sqrt{2}$. But my question is if you want to recover the Lagrangian of a real scalar field from that of a complex scalar field, do we need to put the factor of $1/2$ by hand after setting $\phi=\phi^*$? $\endgroup$ Jan 19, 2020 at 20:02
  • $\begingroup$ We redefine in order to keep standard conventions. $\endgroup$
    – Qmechanic
    Jan 19, 2020 at 20:27
  • $\begingroup$ What happens is that you find the Lagrangian for the real field $\phi_1$ without any adjustments, and then you drop the subscript 1. So you forget about the old complex field $\phi$ and instead let $\phi$ denote the old $\phi_1$. $\endgroup$ Jan 20, 2020 at 13:06

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.