That's been my problem with this approach too. It's an arbitrary model, fitted onto very limited data that doesn't generalize, under the very assumption that th

That's been my problem with this approach too. It's an arbitrary model, fitted onto very limited data that doesn't generalize, under the very assumption that they all derive from a common root. I'm totally willing to accept that the language and DNA models can be thought of as complementing each other, but any conclusion stronger than "perhaps (almost) all European languages have a common origin" remains speculative. And that's not a very interesting conclusion.