10

The Riemann tensor has its component expression: $R^{\mu}_{\nu\rho\sigma}=\partial_{\rho}\Gamma^{\mu}_{\sigma\nu}-\partial_{\sigma}\Gamma^{\mu}_{\rho\nu}+\Gamma^{\mu}_{\rho\lambda}\Gamma^{\lambda}_{\sigma\nu}-\Gamma^{\mu}_{\sigma\lambda}\Gamma^{\lambda}_{\rho\nu}.$

It is straight forward to prove the antisymmetry of $R$ in the last two indices; but how to prove the antisymmetry in the first two ones without assuming symmetric connection/torsion-free metric?

Ernesto Lopez Fune
  • 450
  • 7
    Riemann tensor can be equivalently viewed as curvature 2-form $\Omega$ with values in a Lie algebra $\mathfrak g$ of group $G = SO(n)$. The antisymmetry in one pair comes from being a 2-form, the antisymmetry in the other pair comes from the antisymmetry of ${\mathfrak so}(n)$. This holds even when the connection has torsion. But it won't neccesarily hold for non-metric connections (such as symplectic connections). http://en.wikipedia.org/wiki/Curvature_form – Marek Dec 12 '13 at 17:28
  • Doesn't this just fall from definition? – qqo Nov 21 '18 at 05:24
  • $R(X,Y)Z=\nabla_Y \nabla_X Z - \nabla_X \nabla_Y Z+ \nabla_{[X,Y]}Z.$

    Switching $X,Y$

    $R(Y,X)Z=\nabla_X \nabla_Y Z - \nabla_Y \nabla_X Z+ \nabla_{[Y,X]}Z = -(\nabla_Y \nabla_X Z - \nabla_X \nabla_Y Z - \nabla_{[X,Y]}Z) = -R(X,Y)Z$

     – qqo Nov 21 '18 at 05:29

1 Answers1

1

This simple change in definition,

$$\nabla_a\nabla_b-\nabla_b\nabla_a$$

for

$$\nabla_a\nabla_b-\nabla_b\nabla_a+T^d_{\quad{ab}}\nabla_d$$

However, that the Riemann tensor with torsion is no longer symmetric under exchange of the first pair of indices with the second.

jimbo
  • 2,204