path integral quantization
From TheTangentBundle
We would like to perform the path integral
,
however, since the action is diffeomorphism and Weyl invariant, the integral over
overcounts equivalent configurations and is indeed divergent. We would be better off to fix the gauge to (say) the conformal gauge
or the unit gauge
. Furthermore, to ensure that such a change of metric leaves the path integral unchanged, we need to be aware of any anomalies that may arise.

