A phylogenetic analysis will be just as speculative and inconclusive as any other method if the analyser weights cognitives (as has been done here), using a selective and by definition subjective method to isolate the cognitives concerned. In semantics this approach would almost be deemed heretical in that there is a huge assumption being made regarding the solidity of meaning which, in that discipline, is the very first thing one has to jettison before one can even take up the study.
What I really don't like about the way these linguistic theories are presented, and in fact even the naming conventions used, is the other huge assumption of a primal base from which later languages (also given geographical or quasi-geographical names) then emerge and split. Linguistic evolution of course involves diversification and regional attachment, but it also entails other very important characteristics such as invention, extinction, near extinction followed by re-ignition, and most importantly assimilation and utilisation by people not always travelling westwards! I have no doubt that many of the cognitives can indeed be traced back to Anatolia (this is not a new theory by any means), just as they can with an equal degree of certitude be traced back to the Steppes. This is only an apparently contradictory set of claims if one is hideboundedly devoted to a rather linear and uncomplicated model of human expansion. If on the other hand one is open to the apparent evidence presented anthropologically that human interaction and movement often defies general assumption, then one is obliged to afford the same latitude to the development and movement of the languages they spoke.
I assume the New Zealand researchers won't care too much either if their bald assertion of conclusivity is rejected by their peers in any case. It appears to me that the whole thing is a rather poorly disguised attempt to promote phylogenetic analysis as applicable in disciplines outside of the study of viral transmission for which it was developed programatically. Like other methods of analysis so completely dependent on computer-generated modelling (climate change has been an obvious recent example), the golden rule in assessing its worth - often despite the claims of its practitioners - is to question the validity of the selection of the data which was inputted to generate the model concerned.
In this case (as is indeed so often the case) the answer to this question reveals, it appears yet again, major flaws in terms of objectivity. There is an old-style piece of computer jargon which applies here and has always applied in processes dependent on artificial calculation - GIGO.
Garbage In, Garbage Out.
Home 
