Consider a projective representation of $\mathbb Z^2$ with $U(1)$ coefficients. I would like to find the covering group corresponding to this representation. For this, one needs to find the appropriate central extension by $\mathbb Z^2$ of the Schur multiplier $H^2(\mathbb Z^2, C^\times)$. I have some confusion in calculating this object. If I use the definition of the Schur multiplier given in terms of the homology group $H_2(\mathbb Z^2,\mathbb Z)$, I obtain the group $\mathbb Z$. This is in contradiction to the result $H^2(\mathbb Z^2, C^\times) = U(1)$ that I have seen quoted elsewhere, and which I expect to obtain on certain physical grounds (there is a separate physical interpretation of this problem). I would like to know which result is correct.

  • $\begingroup$ People use "Schur multiplier" to refer to both $H_2(G, \mathbb{Z})$ and $H^2(G, \mathbb{C}^{\times})$ for $G$ a finite group because they are (noncanonically) isomorphic in that case, which they aren't in general. Also, the universal central extension is an extension of $G$ by $H_2(G, \mathbb{Z})$ and it only exists in general if $G$ is perfect, which $\mathbb{Z}^2$ isn't. $\endgroup$ – Qiaochu Yuan Jun 26 '18 at 23:19
  • $\begingroup$ @QiaochuYuan Thanks. I didn't know the finiteness condition. I am trying to interpret 'central extension' as the group of symmetries on a two-dimensional lattice in the presence of a magnetic field (for a fixed field, the translation operators determine a projective representation of $\mathbb Z^2$). So different extensions would correspond to different magnetic fields, and I would not expect there to be a universal central extension for this problem. $\endgroup$ – Naren Manjunath Jun 27 '18 at 7:29

I think that we have

$$H_2(\mathbf{Z} \times \mathbf{Z}, \, \mathbf{Z}) \simeq \mathbf{Z}, \quad H^2(\mathbf{Z} \times \mathbf{Z}, \, \mathbf{C}^{\times})\simeq \mathbf{C}^{\times}.$$

In general, by the Universal Coefficient Theorem there is an isomorphism $$H^2(G, \, \mathbf{C}^{\times})=H_2(G, \, \mathbf{Z})^*$$ Now, it can happen that an infinite group is not isomorphic to its dual. For instance, in your case we have $$\mathbf{C}^{\times}=H^2(\mathbf{Z} \times \mathbf{Z}, \, \mathbf{C}^{\times})=H_2(\mathbf{Z} \times \mathbf{Z}, \, \mathbf{Z})^*=\mathrm{Hom}_{\mathbf{Z}}(\mathbf{Z}, \, \mathbf{C}^{\times}).$$

The isomorphism $\mathrm{Hom}_{\mathbf{Z}}(\mathbf{Z}, \, \mathbf{C}^{\times}) \simeq \mathbf{C}^{\times}$ is explicitly given by associating to every element $h \in \mathrm{Hom}_{\mathbf{Z}}(\mathbf{Z}, \, \mathbf{C}^{\times})$ its value in $1$, namely $h \mapsto h(1)$.


Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.