I am reading chapter 3 of Anomalies in quantum field theory by Reinhold Bertlmann and I found one statement that I don't know how to prove. First of all he defined the right derivative on the Grassmann number to satisfy \begin{equation} \frac{\partial^{R}}{\partial \theta_{i}} \theta_{k} \theta_{l} = \theta_{k} \delta_{il} - \delta_{ik}\theta_{l}.\tag{3.195} \end{equation} My understanding is that the derivative will act from the right. So we can first take derivative on $\theta_{l}$ and this will produce $\theta_{k} \delta_{il}$. Note that no minus sign is here since we act the derivative from right. However, to differentiate $\theta_{k}$ from the right, we need to move the right derivative next to $\theta_{k}$, this will produce a minus sign and then we get $-\delta_{ik} \theta_{l}$. Since I can reproduce this formula, I guess I am on the right track.

He then stated that for both the left and right derivatives we have \begin{align} \left\{ \frac{\partial}{\partial \theta_{i}} , \theta_{k} \right\} = \delta_{ik}.\tag{3.197} \end{align} Let us consider the right derivative and let $\left\{ \frac{\partial^{R}}{\partial \theta_{i}} , \theta_{k} \right\}$ acts on a Grassmann function $f$. We will have \begin{align} \left\{ \frac{\partial^{R}}{\partial \theta_{i}} , \theta_{k} \right\}f & = \left( \frac{\partial^{R}}{\partial \theta_{i}} \left[ \theta_{k}f\right] + \theta_{k} \frac{\partial^{R}}{\partial \theta_{i}}f \right) \\ & = \theta_{k} \frac{\partial^{R}}{\partial \theta_{i}}f - \delta_{ik} f + \theta_{k} \frac{\partial^{R}}{\partial \theta_{i}}f \\ & \neq \delta_{ik} f. \end{align} I am confused, what goes wrong with my attempt? I can show that for left derivatives we have $\left\{ \frac{\partial^{L}}{\partial \theta_{i}} , \theta_{k} \right\} = \delta_{ik}$, but it seems to me that when it comes to right derivative we need to be more careful. However, I am not sure what exact thing we should be careful of since I can really get the $\frac{\partial^{R}}{\partial \theta_{i}} \theta_{k} \theta_{l} = \theta_{k} \delta_{il} - \delta_{ik}\theta_{l}$ correct and I just apply the same logic but it fails to prove $\left\{ \frac{\partial^{R}}{\partial \theta_{i}} , \theta_{k} \right\} = \delta_{ik}$.


OP has a point. Ref. 1 did not expect that OP would evaluate eq. (3.197) on a Grassmann-odd function $f$, and thereby effectively insert a Grassmann-odd object in between the $\theta$-derivative and $\theta$.

We can correct eq. (3.197) in at least 2 ways:

  1. Introduce left and right multiplication operators: $$ m^L_f(g)~:=~fg \quad\text{and}\quad m^R_f(g)~:=~gf. $$ Then $$ \left\{\frac{\partial^L}{\partial\theta^i}, m^L_{\theta^k} \right\}_+~=~\delta^k_i \quad\text{and}\quad \left\{\frac{\partial^R}{\partial\theta^i}, m^R_{\theta^k} \right\}_+~=~\delta^k_i.\tag{3.197'}$$

  2. Introduce the notation $$(\frac{\stackrel{\rightarrow}{\partial}}{\partial\theta^i}f) ~=~(\frac{\partial^L}{\partial\theta^i}f)\quad\text{and}\quad (f\frac{\stackrel{\leftarrow}{\partial}}{\partial\theta^i}) ~=~(\frac{\partial^R}{\partial\theta^i}f),$$ where a right derivative stands to the right of its argument. Then $$ \left\{\frac{\stackrel{\rightarrow}{\partial}}{\partial\theta^i}, \theta^k \right\}_+~=~\delta^k_i \quad\text{and}\quad \left\{\frac{\stackrel{\leftarrow}{\partial}}{\partial\theta^i}, \theta^k \right\}_+~=~\delta^k_i.\tag{3.197"}$$


  1. R.A. Bertlmann, Anomalies in QFT, 1996; section 3.3.1 p. 150.
  • $\begingroup$ Thanks a lot, this is so clear! $\endgroup$
    – ocf001497
    Feb 14 at 21:16

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.