From time to time I read something that says that the tree model of language history is obsolete and that it has been replaced by something else, but I have not understood what that is.
With the tree model, related languages looked more similar in earlier stages, and there could be suggested reconstructions of common protolanguages that of course never would have been existing since our knowledge is fragmentary, but they gave a hint of what might have been.
The tree model is sometimes contrasted with the wave model, where innovations spread over areas and time, but to me this is rather looking at the tree from another angle, making horizontal slices instead of a large vertical slice, so I do not see a difference there.
I get the impression that today, protolanguages are wrong, and language families are wrong, but what is there instead? Does anyone know of some good sources?