I am studying Weyl and Dirac spinors. Suppose we have two Weyl fermions $\eta, \chi$ transforming under $(1/2,0)$ representation of the Lorentz group. I learned that to construct Lorentz invariant term in the Lagrangian, we have terms like $i\eta^\dagger\bar\sigma^\mu\partial_\mu\chi$ or $i\eta\sigma^\mu\partial_\mu\chi^\dagger$. I understand that $\sigma^\mu$ maps left-handed spinors to right-handed and it is Lorentz invariant, therefore sandwiched between the two spinors, but I am still not quite sure how can I justify that $\partial^\mu$ is needed, and we don't need higher-derivative terms?

Also, why for a vector field $A_\mu$, the possible interaction is similar to the derivative term: $i\eta^\dagger\bar\sigma^\mu A_\mu\chi$ ?


1 Answer 1


Index notation is rather useful here.

Left-handed spinors have $\psi_\alpha$ and right-handed spinors have ${\bar \psi}_{\dot \alpha}$ (or the other way around, I can never remember what the conventions are).

If you want a Lagrangian with only left-handed spinors, we need to write a term of the form $\psi_\alpha D^{\alpha\beta} \psi_\beta$. The only tensorial object with this tensor structure is $( \sigma^{\mu\nu} )^{\alpha\beta}$. This has extra $\mu\nu$ indices so we need to contract them. However, it is anti-symmetric in $\mu\nu$ and there is no other such tensor quantity we can contract with. So a kinetic term involving two $\psi_\alpha$'s is impossible.

If you follow the same sort of argument with ${\bar \psi}_{\dot \alpha} D^{{\dot \alpha} \beta} \psi_\beta$ then you will see that at the lowest derivative order, the Dirac Lagrangian is the only possibility. The same argument follows for the gauge field as well.

  • $\begingroup$ Thanks for the answer! Could there be a kinetic term looks like $\psi^\gamma(\sigma^{\mu\nu})^{\alpha}_\gamma(\sigma_{\mu\nu})_{\alpha}^\delta\psi_\delta$ which involves two left-hand spinors? $\endgroup$
    – IGY
    Feb 26 at 2:19
  • $\begingroup$ Or that term looks more like an interaction term instead? $\endgroup$
    – IGY
    Feb 26 at 2:32
  • $\begingroup$ Firstly, that term has no derivatives so it cannot be a kinetic term. That term simplifies to $-3\psi^\alpha \psi_\alpha$ and such a term is totally allowed. $\endgroup$
    – Prahar
    Feb 26 at 10:55

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.