Out of all of my time learning General relativity, this is the one identity that I cannot get around. $$ \Gamma_{\alpha \beta}^{\alpha} = \partial_{\beta}\ln\sqrt{-g} \tag{1}$$ where $g$ is the determinant of the metric tensor $g_{\alpha \beta}$.

With the Christoffel symbol, we start by contracting

$$ \begin{align} \Gamma_{\alpha \beta}^{\alpha} &= \frac{1}{2} g^{\alpha\gamma} (\partial_{\alpha} g_{\beta\gamma} + \partial_{\beta} g_{\alpha\gamma} - \partial_{\gamma} g_{\alpha\beta} ) \\ &= \frac{1}{2} g^{\alpha\alpha} ( \partial_{\beta} g_{\alpha\alpha}) \\ &= \frac{1}{2g_{\alpha\alpha}} ( \partial_{\beta} g_{\alpha\alpha}) \end{align}\tag{2}$$

where I took $\gamma \rightarrow \alpha$ and $g^{\alpha\alpha} = 1/g_{\alpha\alpha}$.

The next steps to take now, I have no clue. MTW gives a hint by saying to use the results from some exercise, which are,

$$\det A = \det||A^{\lambda}_{\ \ \rho}|| = \tilde{\epsilon}^{\alpha\beta\gamma\delta}A^{0}_{\ \ \alpha}A^{1}_{\ \ \beta}A^{2}_{\ \ \gamma}A^{3}_{\ \ \delta} $$

$$(A^{-1})^{\mu}_{\ \ \alpha}(\det A) = \frac{1}{3!}\delta_{\alpha\beta\gamma\delta}^{\mu\nu\rho\sigma} A^{\beta}_{\ \ \nu} A^{\gamma}_{\ \ \rho}A^{\delta}_{\ \ \sigma} $$

$$ \mathbf{d}\ln|\det A| = \mathrm{trace}(A^{-1}\mathbf{d}A) ,\tag{3}$$ where $\mathbf{d}A$ is the matrix $||\mathbf{d}A^{\alpha}_{\ \ \mu}||$ whose entries are one-forms.

I fail to reason why the metric turns into the determinant from what I have done and then becomes the result at the top.

  • 3
    $\begingroup$ Comment to the post (v2): Giving four indices in eq. (2) the same name $\alpha$ is not legitimate, or at best: asking for trouble. $\endgroup$
    – Qmechanic
    Commented Feb 3, 2017 at 16:05

3 Answers 3


Recall the matrix identity $$\tag{1}\log\det M=\operatorname{tr}\log M.$$ If $M=M(\lambda)$ is differentiable in $\lambda$, then $$\tag{2}\frac{d}{d\lambda}\log\det M=\operatorname{tr}\left(M^{-1}\frac{d}{d\lambda} M\right).$$ The proof of $(1)$ for symmetric matrices follows from the usual formulae for the trace and determinant in terms of eigenvalues$^{1}$.

As for the Christoffels, we have $$\Gamma^i{}_{ij}=\frac{1}{2}g^{ik}(\partial_i g_{jk}+\partial_j g_{ik}-\partial_k g_{ij})=\frac{1}{2}g^{ik}\partial_j g_{ik}=\frac{1}{2}\operatorname{tr}(g^{-1} \partial_j g).$$ The last equality is just what the contraction of indices means for the (symmetric!) matrix $g=(g_{ij})$, and there is an error in the indices in OP's post. Now, using $(2)$ we have $$\Gamma^i{}_{ij}=\frac{1}{2}\partial_j\log \det g.$$ This can be brought into the form $$\Gamma^i{}_{ij}=\partial_j \log\sqrt{|\det g|}$$ by the usual rules of calculus.

$^{1}$ For symmetric matrices, such as $g$, it is easy because $g$ can be diagonalized. For other matrices you might need a Jordan normal form to compute $\log M$.

  • 1
    $\begingroup$ I know this is an old question but I am not seeing how it follows that $\frac{1}{2}g^{ik}(\partial_i g_{jk}+\partial_j g_{ik}-\partial_k g_{ij})=\frac{1}{2}g^{ik}\partial_j g_{ik}$ $\endgroup$
    – Jepsilon
    Commented Nov 18, 2018 at 9:29
  • $\begingroup$ @Jepsilon The first and third terms cancel each other. $\endgroup$
    – Ryan Unger
    Commented Nov 23, 2018 at 15:29
  • $\begingroup$ This is now an even older post, but it looks to a beginner like me like $g^{ik}\partial_jg_{ik}$ ought to be a sum over all $i$ and $k$, e.g. $g^{00}\partial_jg_{00} + g^{01}\partial_jg_{01} + \cdots + g^{10}\partial_jg_{10} + \cdots$, not just the sum over the diagonal. Can you explain why this is not true? Thanks. $\endgroup$ Commented Oct 5, 2019 at 13:43
  • 2
    $\begingroup$ @flevinBombastus It's a sum over all indices. If $A$ and $B$ are (symmetric) matrices, then $trace(AB)=\sum_{ij}A_{ij}B_{ij}$. $\endgroup$
    – Ryan Unger
    Commented Oct 8, 2019 at 1:30

The original and the most general definition of determinant is given by Gauss . For the determinant of metric tensor we write \begin{eqnarray} g&:=& \frac{1}{4!}\varepsilon^{{\alpha\beta}{\gamma\delta}}\varepsilon^{{\mu\nu}{\rho\sigma}}g_{\alpha\mu}g_{\beta\nu}g_{\gamma\rho}g_{\delta\sigma}.\\ \therefore \delta g &=& \frac{1}{3!}\varepsilon^{{\alpha\beta}{\gamma\delta}}\varepsilon^{{\mu\nu}{\rho\sigma}}g_{\alpha\mu}g_{\beta\nu}g_{\gamma\rho}\delta g_{\delta\sigma},\\ &=&\frac{-g}{3!}\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon^{{\mu\nu}{\rho\sigma}} g_{\alpha\mu}g_{\beta\nu}g_{\gamma\rho}\delta g_{\delta\sigma},\\ &=&\frac{-g}{3!}\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon_{{\alpha\beta}\gamma}{}^\sigma \delta g_{\delta\sigma},\\ &=&g \, g^{\delta\sigma}\delta g_{\delta\sigma}. \end{eqnarray} (Equivalently, $$ \delta(\ln |\det g|)=Tr (g^{-1}\delta g)= Tr ( \delta \ln g )$$) By using this result we the have $$ \frac{1}{g} \partial_\beta g =g^{\delta\sigma}\partial_\beta g_{\delta\sigma} $$

Note: You may need some basic manipulation of these quantities, $\varepsilon$ is Levi-Civita symbol, $\epsilon$ is Levi-Civita tensor.

$$\epsilon^{{\alpha\beta}{\gamma\delta}} = -\frac{1}{\sqrt{-g}}\varepsilon^{{\alpha\beta}{\gamma\delta}}$$ $$\epsilon_{{\alpha\beta}{\gamma\delta}} = \sqrt{-g}\varepsilon_{{\alpha\beta}{\gamma\delta}}$$

$$\varepsilon_{{\alpha\beta}{\gamma\delta}}=\varepsilon^{{\alpha\beta}{\gamma\delta}}= \delta^{[\alpha}_0 \delta^\beta_1 \delta^\gamma_2 \delta^{\delta]}_3$$

$$\varepsilon^{{\alpha\beta}{\gamma\delta}}\varepsilon_{{\mu\nu}{\rho\sigma}} =-\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon_{{\mu\nu}{\rho\sigma}}= 4! \delta^{[\alpha}_\mu \delta^\beta_\nu \delta^\gamma_\rho \delta^{\delta]}_\sigma \equiv \delta^{\alpha\beta\gamma\delta}_{\mu\nu\rho\sigma}\equiv \left| \begin{matrix} \delta^\alpha_\mu & \delta^\alpha_\nu & \delta^\alpha_\rho & \delta^\alpha_\sigma \\ \delta^\beta_\mu & \delta^\beta_\nu & \delta^\beta_\rho & \delta^\beta_\sigma \\ \delta^\gamma_\mu & \delta^\gamma_\nu & \delta^\gamma_\rho & \delta^\gamma_\sigma \\ \delta^\delta_\mu & \delta^\delta_\nu & \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right|$$

$$\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon_{{\alpha\nu}{\rho\sigma}} = -\left| \begin{matrix} \delta^\alpha_\alpha & \delta^\alpha_\nu & \delta^\alpha_\rho & \delta^\alpha_\sigma \\ \delta^\beta_\alpha & \delta^\beta_\nu & \delta^\beta_\rho & \delta^\beta_\sigma \\ \delta^\gamma_\alpha & \delta^\gamma_\nu & \delta^\gamma_\rho & \delta^\gamma_\sigma \\ \delta^\delta_\alpha & \delta^\delta_\nu & \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right| =-\left| \begin{matrix} 1 & 0 & 0 & 0 \\ 0 & \delta^\beta_\nu & \delta^\beta_\rho & \delta^\beta_\sigma \\ 0 & \delta^\gamma_\nu & \delta^\gamma_\rho & \delta^\gamma_\sigma \\ 0 & \delta^\delta_\nu & \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right| =-\left| \begin{matrix} \delta^\beta_\nu & \delta^\beta_\rho & \delta^\beta_\sigma \\ \delta^\gamma_\nu & \delta^\gamma_\rho & \delta^\gamma_\sigma \\ \delta^\delta_\nu & \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right|$$

$$\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon_{{\alpha\beta}{\rho\sigma}} = -\left| \begin{matrix} \delta^\beta_\beta & \delta^\beta_\rho & \delta^\beta_\sigma \\ \delta^\gamma_\beta & \delta^\gamma_\rho & \delta^\gamma_\sigma \\ \delta^\delta_\beta & \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right|=-\left| \begin{matrix} 2 & 0 & 0 \\ 0 & \delta^\gamma_\rho & \delta^\gamma_\sigma \\ 0 & \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right|=-2 \left| \begin{matrix} \delta^\gamma_\rho & \delta^\gamma_\sigma \\ \delta^\delta_\rho & \delta^\delta_\sigma \end{matrix} \right| $$

$$\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon_{{\alpha\beta}{\gamma\sigma}} =-2 \left| \begin{matrix} \delta^\gamma_\gamma & \delta^\gamma_\sigma \\ \delta^\delta_\gamma & \delta^\delta_\sigma \end{matrix} \right| =-2 \left| \begin{matrix} 3 & 0 \\ 0 & \delta^\delta_\sigma \end{matrix} \right| = -3! \delta^\delta_\sigma $$

$$\epsilon^{{\alpha\beta}{\gamma\delta}}\epsilon_{{\alpha\beta}{\gamma\delta}} =-3! \delta^\delta_\delta = -4!$$

Note2: \begin{eqnarray} g^{-1}&:=& \frac{1}{4!}\varepsilon_{{\alpha\beta}{\gamma\delta}}\varepsilon_{{\mu\nu}{\rho\sigma}}g^{\alpha\mu}g^{\beta\nu}g^{\gamma\rho}g^{\delta\sigma}.\\ \therefore \delta g^{-1} &=& \frac{1}{3!}\varepsilon_{{\alpha\beta}{\gamma\delta}}\varepsilon_{{\mu\nu}{\rho\sigma}}g^{\alpha\mu}g^{\beta\nu}g^{\gamma\rho}\delta g^{\delta\sigma},\\ &=&\frac{-g^{-1}}{3!}\epsilon_{{\alpha\beta}{\gamma\delta}}\epsilon_{{\mu\nu}{\rho\sigma}} g^{\alpha\mu}g^{\beta\nu}g^{\gamma\rho}\delta g^{\delta\sigma},\\ &=&\frac{-g^{-1}}{3!}\epsilon_{{\alpha\beta}{\gamma\delta}}\epsilon^{{\alpha\beta}\gamma}{}_\sigma \delta g^{\delta\sigma},\\ &=&g^{-1} \, g_{\delta\sigma}\delta g^{\delta\sigma}.\\ \therefore -g^{-2}\delta g &=&g^{-1} \, g_{\delta\sigma}\delta g^{\delta\sigma}.\\ \delta g &=& -g\,g_{\delta\sigma}\delta g^{\delta\sigma}. \end{eqnarray} By comparing this $\delta g$ to the first result, we have $$\delta g_{\alpha\beta}=- g_{\alpha\mu}g_{\beta\nu} \delta g^{\mu\nu}$$


Here's a direct/theoretical way to arrive at the formula for $\Gamma^{\nu}_{\mu \nu}$ in terms of the square root of the determinant of the metric from first principles. Note $\sqrt{-g}$ appears in covariant four-dimensional integrals, so you can expect the formula to be derived using integrals.

Using $$\nabla g_{\mu \nu} = \nabla g = 0 \ \ \ , \ \ \ \nabla^2 = g^{\mu \nu} \nabla_{\mu} \nabla_{\nu} \ \ , $$ we have (ignoring integration by parts terms assumed to vanish) $$\int d^4 x \sqrt{-g} (\nabla^2 \phi) \phi = - \int d^4 x \sqrt{-g} g^{\mu \nu} \nabla_{\mu} \phi \nabla_{\nu} \phi = - \int d^4 x \sqrt{-g} g^{\mu \nu} \partial_{\mu} \phi \partial_{\nu} \phi = \int d^4 x \sqrt{-g} \frac{1}{\sqrt{-g}} \partial_{\mu} (\sqrt{-g} g^{\mu \nu} \partial_{\nu} \phi) \phi $$ giving the curvilinear form of the Laplacian $$\nabla^2 \phi = \frac{1}{\sqrt{-g}} \partial_{\mu} (\sqrt{-g} g^{\mu \nu} \partial_{\nu} \phi) $$ so that on expanding it directly $$\nabla_{\mu} (\nabla^{\mu} \phi) = \partial_{\mu} (\partial^{\mu} \phi) + \Gamma^{\mu}_{\mu \nu} (\partial^{\nu} \phi) = \frac{1}{\sqrt{-g}} \partial_{\mu} (\sqrt{-g} g^{\mu \nu} \partial_{\nu} \phi) = \partial_{\mu} \partial^{\mu} \phi + (\frac{1}{\sqrt{-g}} \partial_{\nu} \sqrt{-g}) \partial^{\nu} \phi$$ we find the following expression for the contracted Christoffel symbols $$\Gamma^{\mu}_{\mu \nu} = \frac{1}{\sqrt{-g}} \partial_{\nu} \sqrt{-g} = \partial_{\nu} \ln \sqrt{-g} .$$


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.