3
$\begingroup$

I am following the algorithm in W-algebras with two and three generators, in order to construct consistent (anti-)commutator relations for a particular W-algebra.

I am considering $W(2,4,4)$ where both dimension four operators are fermionic. I have two questions related to the method introduced in the paper, namely:

  1. They use Jacobi identities (e.g. $\{\Psi_m,\{\Psi_n,\Psi_\ell\}\} + \mathrm{permutations} = 0$) to fix some of the arbitrary constants, but this isn't sufficient to constrain all of them. How do they calculate the rest? (There is some step in the example involving computing determinants but it isn't clear how this relates to the notation of the previous section.)
  2. For fermionic operators, when computing $\{\Psi_m,\{\Psi_n,\Psi_\ell\}\}$, due to the terms that appear in the inner anti-commutator, you encounter stuff like $\{\Psi_m, L_{n+\ell}\}$ but normally, we use a commutator in this case, as one element is even and one is odd. So when computing the Jacobi identity with fermionic operators, should one actually use $[,\}$ as opposed to $\{,\}$?
$\endgroup$
7
  • 1
    $\begingroup$ Permalink: dx.doi.org/10.1016/0550-3213(91)90624-7 $\endgroup$
    – Qmechanic
    Commented Jan 23, 2020 at 18:21
  • $\begingroup$ Jacobi identities are not the most efficient way to find $W$ algebras. This is more easily done with OPEs and the conformal bootstrap method. But maybe more relevant: why would you do this approx 25 years after most of this has been done already? $\endgroup$ Commented Jan 23, 2020 at 18:31
  • $\begingroup$ @Oбжорoв I can't find a paper with the full OPEs of $W(2,4,4)$ (with the operators fermionic). Some papers comment on properties of the algebra but no explicit OPEs. $\endgroup$
    – JamalS
    Commented Jan 23, 2020 at 18:52
  • 1
    $\begingroup$ I would be surprised in anyone has ever written down such an algebra explicitly. But hold on. Spin 4 fermionic operators? $\endgroup$ Commented Jan 23, 2020 at 20:04
  • 1
    $\begingroup$ Good luck indeed! The motivations for doing such horrible calculations are obscure to me. In particular, high level null states are complicated and not very useful, even for the Virasoro algebra. $\endgroup$ Commented Jan 24, 2020 at 8:13

1 Answer 1

0
$\begingroup$

As it turns out, even though the authors of the paper I linked claim there are benefits to working with the commutators rather than the OPEs, I found using the OPEs much simpler, and didn't need the machinery provided in the algorithm in the paper.

In my case, the two fermionic generators carry an $\mathfrak{sl}(2)$ charge which we require the OPE to respect, and so I simply wrote down all possible terms that may arise order by order, i.e.

$$\Psi^+(z) \Psi^{-}(z) = \frac{a_1}{z^8}\mathbb{I} + \frac{a_2}{z^6}T + \frac{a_3}{z^5}\partial T + \frac{a_4}{z^4}\partial^2 T + \frac{a_5}{z^4}T^2 + \dots$$

using combinations of derivatives and $T$ with arbitrary constants $a_i$. Imposing the Jacobi results in some trivial constraints already satisfied (e.g. taking $TTT$) but others (e.g. $\Psi^+ \Psi^{-}T$) led to constraints on the $a_i$ up to null states.

A useful paper for carrying this out is An Algorithmic Approach to Operator Product Expansions, W-Algebras and W-Strings.

$\endgroup$

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.