I am trying to make sense of this simple case in my book, but I am still baffled by the notation that is used in the indices with the commas and semicolons; I also do not understand how these are equivalent.


  • 1
    $\begingroup$ Please explain better the notation, or add a reference in which this notation is used, some (like me) may be not familiar with this notation $\endgroup$ – Fabio Di Nocera Mar 9 '20 at 14:03
  • $\begingroup$ Which book? Which page? $\endgroup$ – Qmechanic Mar 9 '20 at 14:08
  • 2
    $\begingroup$ The connection is presumably torsionfree. $\endgroup$ – Qmechanic Mar 9 '20 at 14:10
  • $\begingroup$ Sean Carroll-Spacetime and Geometry: An Introduction to General Relativity, 2019 edition, Chapter 3, Page 152. $\endgroup$ – UF6 Mar 9 '20 at 14:43
  • $\begingroup$ I think that the book probably explains the comma and semicolon notations somewhere, before page 152. $\endgroup$ – G. Smith Mar 9 '20 at 17:17

By definition: \begin{align}\tag{1} A_{\mu,\nu} &\equiv \frac{\partial}{\partial x^{\nu}} \, A_{\mu} \equiv \partial_{\nu} \, A_{\mu}, \\[1ex] \tag{2} A_{\mu;\nu} &\equiv \frac{\partial}{\partial x^{\nu}} \, A_{\mu} - \Gamma_{\nu \mu}^{\lambda} \, A_{\lambda} \equiv \nabla_{\nu} \, A_{\mu},\end{align} If the connection is symetric: $\Gamma_{\mu \nu}^{\lambda} = \Gamma_{\nu \mu}^{\lambda}$ (Levi-Civita connection AKA Christoffel symbols, which means no torsion), then the antisymetric expression $A_{\mu;\nu} - A_{\nu;\mu}$ gives the same as $A_{\mu,\nu} - A_{\nu,\mu}$.

  • 5
    $\begingroup$ Just to be more precise, please don't take it as a critic, I completely understand the spirit of your answer. But the connection and the Christoffel symbols are two completely different objects, what you meant is that if you choose the Levi-Civita connection (as it is usually done in GR) then the Christoffel symbols are symmetric in the two "down" indices. $\endgroup$ – Fabio Di Nocera Mar 9 '20 at 14:17

The semikolon indicates a covariant derivative, in General Relativity only covariant derivatives transform like a tensor, therefore it is so important.

The simple comma in difference indicates a simple derivative.

the difference between covariant derivative and simple derivative is:

$$A_{a;b} = A_{a,b} - \Gamma^c_{ba} A_c$$

where $\Gamma$ stands for the Christoffel-symbols. See a.o.: https://en.wikipedia.org/wiki/Christoffel_symbols

On the basis of this formula it can be easily shown that

$$A_{a;b} -A_{b;a} = A_{a,b} - A_{b,a}$$

because the Christoffel-symbols are symmetric on the 2. and 3. index (if the torsion is zero). Actually, in order to define a covariant derivative on a (pseudo)-Riemannian manifold an additional structure is required, the "connection". The Christoffel-symbols represent here the connection. However, for the computation of an exterior derivative of a one-form $A$ a connection is not required. Actually $ A_{a,b} - A_{b,a}$ represent the result of the exterior derivative of a one-form $A$ in components. The equality $A_{a;b} -A_{b;a} = A_{a,b} - A_{b,a}$ shows that an exterior derivative can be defined on a (pseudo)-Riemannian manifold without further structure introduced.


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.