I've been reading various texts on tensor algebra and calculus in preparation for applications of it, and I find myself continuously having issues with the index calculus. I've been seeing:
$$ \partial^{\mu}T^{\nu} = g^{\rho\mu} \frac{\partial T^{\nu}}{\partial x^{\rho}} $$
ie the metric appears explicitly with a dummy index to match $ \partial x^{\rho} $. Everything appears to be contravariant but the appearance of the metric with a dummy index matching the denominator leads me to believe that indices in the denominator of differentiation are of opposite character than what their index states. My question arises because I don't see the metric when taking a derivative with respect to a covariant index and the denominator still appears as contravariant. Am I correct in my assumption?
$\partial^\mu$ is often a shortcut for $\partial/\partial x_\mu$ and $\partial_\mu$ for $\partial/\partial x^\mu$ because of this. See here for example.
$$ g^{\rho\mu} \frac{\partial}{\partial x^\rho} = g^{\rho\mu} \partial_\rho = \partial^\mu $$