0
$\begingroup$

The vacuum functional for the fermionic oscillator is given by

$$ Z[0] = N\int\mathcal{D}\overline{\psi}\mathcal{D}\psi \exp\left(i\int_0^Tdt\left(i\overline{\psi}\psi-w\overline{\psi}\psi \right)\right). \tag{5.80}$$

Using Weyl ordering and discretizing the time integral, it is claimed that this could be written as

$$ Z[0] = \lim_{\epsilon\rightarrow0}\lim_{N\rightarrow\infty}N\int d\overline{\psi}_1\dots d\overline{\psi}_{N-1}d\psi_1\dots d\psi_{N-1}$$ $$\times \exp\left(i\epsilon\sum_{n=1}^N\left(i\overline{\psi}_n\frac{\psi_n-\psi_{n-1}}{\epsilon}-w\overline{\psi}_n\frac{\psi_n+\psi_{n-1}}{2} \right)\right),\tag{5.81} $$

where the mid-point prescription of the Weyl ordering was used.

My question is pretty simple and naive. To write this expression, the author says on the bottom of p.92 that

$\overline{\psi}$ represents the momentum conjugate to $\psi$.

But if I make the computation I get (using left derivatives)

$$ \Pi_\psi = \frac{\partial L}{\partial \dot{\psi}}=-i\overline{\psi}.\tag{5.43} $$

Why is it valid to ignore the $-i$ factor and just to consider $\overline{\psi}$ as the conjugate momentum?

$\endgroup$
  • $\begingroup$ ${}$ Which page? $\endgroup$ – Qmechanic Jun 25 at 4:47
  • $\begingroup$ p92, second edition. Thanks for looking it up! $\endgroup$ – user2820579 Jun 25 at 17:29
  • $\begingroup$ Does it make you equation for $Z[0]$ invalid? $\endgroup$ – Oбжорoв Jun 25 at 18:06
1
$\begingroup$

Ashok Das is not ignoring any factor of $-i$ in formulas. The word "represent" in the sentence $\overline{\psi}$ represents the conjugate to $\psi$ is here used semantically in a weaker sense than "is equal", e.g. "is equal up to a multiplicative constant".

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ General question. What is then the rule in order to apply the mid-point prescription? It seems to me that this is a loosely rule, and I'm worried that different cases yield different answers. Does this only happens for fermions? $\endgroup$ – user2820579 Jun 26 at 18:19

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.