2
$\begingroup$

Consider a time dependent Grassmann field i.e. $\theta(t)$. Now, consider the following Berezin integral $$\int [\mathcal{D}\theta] ~\prod_{t}\dot{\theta}\tag{1}$$ where $\dot{\theta}$ time derivative of $\theta$ and $[\mathcal{D}\theta]$ is a functional measure over the Grassmann field, something like a path integral measure and a product integral of $\dot{\theta}$. Now, what does the above integral evaluate to? One way I can think of evaluating the integral is that by recognizing that $\dot{\theta} = \delta(\dot{\theta})$ as $\dot{\theta}^2 = 0$. Therefore, we have $$\int [\mathcal{D}\theta] ~\prod_{t}~\dot{\theta} = \int [\mathcal{D}\theta]~\prod_{t} ~\delta(\dot{\theta}) = \det({\partial_t}).\tag{2}$$ Is my evaluation of the integral correct?

$\endgroup$
2
  • 3
    $\begingroup$ If you have $\theta (t)$ then you have infinitlely many Grassmann variables (one for each $t$) and so the product of infinitelly many in $D[\theta]$. You will threfore need infinitely many of them in the integrand to get a non-zero answer. $\endgroup$
    – mike stone
    Sep 23, 2021 at 11:25
  • $\begingroup$ @mikestone oops, good point. Fixed that. $\endgroup$
    – user44690
    Sep 23, 2021 at 16:42

1 Answer 1

1
$\begingroup$

TL;DR: Yes, OP is correct.

Let's discretize time $t$. In other words, assume that we have $n$ Grassmann-odd variables $\theta^1, \ldots, \theta^n$. In this language OP's differential operator $\partial_t$ is replaced with some Grassmann-even matrix $A^i{}_j$. Define $$\theta^{\prime i}= \sum_{j=1}^nA^i{}_j \theta^j,$$ so that $$\frac{\partial}{\partial\theta^j} ~=~\sum_{i=1}^n A^i{}_j\frac{\partial}{\partial\theta^{\prime i}}.$$ Then a straightforward calculation shows that OP's path-integral (1) is replaced with$^1$ $$\begin{align} \int\!d\theta^n \ldots d\theta^1~\theta^{\prime 1}\ldots \theta^{\prime n}~=~&\ldots~=~ \int\! d\theta^n \ldots d\theta^1~\delta(\theta^{\prime 1})\ldots \delta(\theta^{\prime n})\cr ~=~&\ldots~=~\det(A),\end{align} $$ which indeed is a discrete version of OP's claim (2).

--

$^1$Recall that Berezin integration $$ \int\!d\theta^n \ldots d\theta^1 ~=~\frac{\partial}{\partial\theta^n}\ldots \frac{\partial}{\partial\theta^1}$$ is the same as differentiation!

$\endgroup$

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.