Parallel transport, as one means of quantifying the curvature of a coordinate space, enables changes in a vector's components, when it is carried around variously oriented loops in that space, to be properly measured, i.e. by comparisons made at the same location. Those changes which are independent of the size of the loop are measurable manifestations of local curvature, and can be coded into components of the Riemann tensor. Have I got this right? Now spacetime has four dimensions, three of space and one of time. Transporting anything around a loop takes time, so a one-way leg along the time dimension must in principle be part of any loop. It is therefore never quite possible --- especially in cosmology! --- to compare the original vector with its parallel-transported version at the same location in spacetime, as is possible with a loop on the 2-D Earth's surface (sometimes used to explain how parallel transport measures curvature). How could the curvature of spacetime on say, a cosmological scale then be measured, even in thought experiments? And is the separation of spacetime curvature into that of space sections and of time thus moot?