Attempting to reconstruct the origin and diversification of human lineages or of human languages is a matter of hypothesizing events within our undocumented prehistory. There are only two ways to do that: we can speculate, suggesting what must have happened on the basis of ‘common sense’, or we can find ways to extrapolate from the observable present into the unobservable past, using some appropriate version of the ‘uniformitarian principle’ (UP), as paleontologists and historical geologists do. The conclusion derived is that when we have a better idea what patterns of linguistic development correspond to particular network patterns, it will be time to tackle the problem.