Ok for real photons there is the formula when summing over the polarizations:

$$ \sum_{\lambda=\pm}\epsilon^{*\mu}_\lambda\epsilon^\nu_\lambda = -\eta^{\mu\nu}$$

But if I have a matrix element of the form: $$\epsilon^{*\mu}M_\mu\qquad \epsilon^\mu\bar{M}_\mu$$

So when I take the absoulte squared of that I have: $$|M|^2 = \epsilon^{*\mu}M_\mu \epsilon^\nu\bar{M}_\nu= -M_\mu\eta^{\mu\nu}M_\nu = -M \bar{M}$$

But know the absolute squared is negative. So I know I have a heavy mistake somewhere here. Can someone help me understand? Thanks a lot.

  • $\begingroup$ You are using the $(1,-1,-1,-1)$ signature right? $\endgroup$ – MannyC Apr 19 at 13:08
  • $\begingroup$ Yes. Sorry I didn't mention. $\endgroup$ – higgshunter Apr 19 at 13:16
  • $\begingroup$ Why must $M_\mu M^{*\mu}>0$? $\endgroup$ – WAH Apr 19 at 14:35
  • 1
    $\begingroup$ Well I thought that now we have $|M|^2 = -M_\mu M^\mu$ and the total amplitude should be real of course and positive? Or am I missing something? $\endgroup$ – higgshunter Apr 19 at 14:41

$$-M_\mu\eta^{\mu\nu}M_\nu=-M_0^2+M_1^2+M_2^2+M_3^2$$ Since due to Ward identity $M_0=M_3$ (choosing $x^3$ as the direction of motion), the above result is positive.

  • $\begingroup$ Seems logical. Since it should be positive and also in my actual calculations if I evaluate the resulting Trace I get something positive. I was just looking for a general explanation to see why this should be the case. $\endgroup$ – higgshunter Apr 20 at 9:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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