I am comfortable doing the following calculation, Derivation of the adjoint of Dirac equation, notably — going from the standard Dirac equation to the adjoint Dirac equation via using Dirac conjugation.

I am however not comfortable deriving the adjoint Dirac equation from the Euler Lagrange equation (for $\psi$, the EL eq for $\overline{\psi}$ leads to the standard Dirac equation) of the Dirac Lagrangian

$$\mathcal{L} =\overline{\psi}(i\gamma^\mu \partial_\mu -m)\psi $$.

My problem boils down to the following term,

$$ \partial_\mu (i\overline{\psi}\gamma^\mu)$$

How can I get this to yield

$$ -i\gamma^\mu \partial_\mu \overline{\psi}? $$

My only thought is to write out Dirac conjugate, write out the Einstein summation, and use the properties of the gamma matrices ($\gamma^0\gamma^0=$ the identity matrix, and $\gamma^0\gamma^i=-\gamma^i\gamma^0$), but it doesn't seem to yield anything fruitful.




1 Answer 1


The Euler-Lagrange-Equation is given by:

$$\frac{\partial\mathcal{L}}{\partial {\psi}} - \partial_{\mu}\frac{\partial \mathcal{L}}{\partial(\partial_{\mu}\psi)} = 0$$

Let us take both derivatives separately. We treat $\psi$ and $\bar{\psi}$ as independent fields. This gives

$$\frac{\partial}{\partial\psi}\bar{\psi}(i\gamma^\mu\partial_{\mu}-m)\psi = -m \bar{\psi}$$


$$\partial_{\mu}\frac{\partial}{\partial(\partial_{\mu}\psi)}\bar{\psi}(i\gamma^\mu\partial_{\mu}-m)\psi = i\partial_{\mu}\bar{\psi}\gamma^\mu$$

So by plugging this into the first equation we get:

$$-m\bar{\psi} - i\partial_{\mu}\bar{\psi}\gamma^\mu = 0 $$

Now, to make this look nicer (and bring this in the usual form) we say, that the differential operator operates to the left:

$$\bar{\psi}(i\gamma^{\mu} \overleftarrow{\partial}_{\mu}+m) = 0$$

Your proposed solution $i \partial_\mu \gamma^\mu \bar{\psi}$ doesn't work dimensional wise. Since $\bar{\psi}$ is a row vector, it needs to be left of the $\gamma^\mu$, which is a $4\times 4$ square matrix.

  • $\begingroup$ "$\bar{\psi}(i\gamma^{\mu} \overleftarrow{\partial}_{\mu}-m) = 0$", pls check the sign. $\endgroup$
    – MadMax
    Commented Mar 10, 2020 at 15:00
  • $\begingroup$ Whoops, thank you $\endgroup$ Commented Mar 10, 2020 at 15:50
  • 1
    $\begingroup$ How is it that you pulled out $\bar\Psi$ from $\partial_\mu\bar\Psi\gamma^\mu$? If $\gamma^\mu$ is a matrix, it will not in general commute. $\endgroup$
    – user400188
    Commented May 13, 2020 at 7:43
  • 1
    $\begingroup$ I didn't do this anywhere? $\endgroup$ Commented May 13, 2020 at 8:10
  • 1
    $\begingroup$ Yes. Since $\gamma^\mu$ does not depend on $x$, it commutes with $\partial_\mu$. $\endgroup$ Commented May 13, 2020 at 9:06

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.